INTRODUCTION TO THE UBC WATERSHED AND FLOW MODELS

WHAT THE WATERSHED MODEL DOES

The Watershed Model generates a computational representation of watershed behavior. Using

daily maximum and minimum temperatures and precipitation data entered by the user, the
program calculates daily watershed outflow resulting from glacier melt, snowmelt and rainfall.
In addition to streamflow, the model also provides information on the accumulation and
depletion of snowpacks, soil moisture budgets, soil and groundwater storage values,
contributions to runoff from various portions of the watershed, and surface and sub-surface
components of runoff. These data can be viewed on the screen in graphics or text form, printed,
or saved to a file.
1) Given continuous meteorological input data, the model will operate continuously,
accumulating and depleting the snowpack and producing estimates of stream flow.
For calibration and verification purposes, the model uses historical meteorological
and streamflow records as reference data and calculates performance statistics on
volume and hydrograph shape reconstitution.

THE UBC FLOW MODEL

This model is used to represent the channel, lake and reservoir network. The Watershed Model
is used to generate the tributary inflows from all sub-watersheds into this total river network, and
then the Flow Model calculates the channel, lake or reservoir routing of these various tributary
inflows to any required downstream point in the network.
The channel routing is carried out using a modified kinematic wave process, which also accounts
for wave dispersion. The great advantage of this method is that the routing constants can be
determined from river gauging data, namely the stage-discharge and the velocity-discharge data.
Provided the river gauging stations are reasonably representative of the channel reach, no
calibration is usually required.

USES OF THE WATERSHED MODEL

The Watershed Model was designed primarily for short term river flow forecasting. The
accuracy of this forecast depends upon:

. the current assessment of the snowpack, soil, and groundwater storage,

. the representativeness of the meteorological data to date, and

. the accuracy of the meteorological forecast.

Because the Watershed Model operates continuously if given continuous meteorological input, it
is possible to use the model for longer term forecasting using projected patterns of weather. This
capability can assess the range of possible outcomes for a whole season of snowmelt and rain
runoff once an initial watershed status has been specified. Such seasonal forecasts can be
upgraded with the recorded data as the season advances and the possible range of outcomes will
gradually narrow. An extension of this mode of operation has been used to estimate complete
years of data when the streamflow records have not existed, but for which meteorological data
has been available.

DESIGN CONSTRAINTS
The Watershed Model was developed to describe and forecast watershed behavior in
mountainous areas. This introduced several important design constraints because data in such
regions are usually scarce, particularly at higher elevations. A major design consideration
resulting from this was to provide the model with the ability to interpret meteorological data at a
point in terms of basin-wide conditions. This requires assumptions to be made concerning
temperature and precipitation distribution by elevation and by area. Appendix A describes the
algorithms that carry out these functions. An important part of model calibration (see Chapter 5)
is the determination of suitable control parameter values to describe these processes.
Another constraint is the availability of streamflow reference data used in the calibration process.
For headwaters and major tributaries, independent streamflow data are usually available. For
many smaller tributaries and for watersheds downstream of the headwaters no independent
streamflow records may exist. To calibrate such watersheds it is necessary to use a streamflow
routing model, such as the UBC FLOW MODEL, to deduce the streamflow contributions from
each sub-area. These derived streamflow records will generally be lower in quality because
distributed lateral inflows to a channel system may have to be treated as a lumped inflow and
also measurement errors at the two reference stations will tend to interact to produce an
enhanced random error component.
A further design consideration was to provide the model with the ability to estimate the
snowpack. Two data sources are available, precipitation measured at meteorological stations
which are usually valley sites, and snowcourse measurements which are available at various
elevations, from valley floor to higher snow fields. Snowcourse data are usually available only
for certain dates during the snowmelt season, perhaps at two-week intervals. In addition,
snowcourse data cannot be considered an absolute measure of snow at that elevation. Rather,
they tend to be indices of snow accumulation. On the other hand, precipitation measurements
are available on a daily basis at all seasons of the year. For this reason, the Watershed Model
uses daily precipitation inputs to estimate snow accumulation and rainfall amounts by elevation.
The snowcourse data is used, external to the model, as calibration reference data. Plots for
snowcourse data on semi-log paper show a reasonably consistent pattern from year to year, not in
absolute but in relative amounts, and also show consistent patterns with computed snowpacks.

Recent Developments
Recent improvements to the model have proceeded on two related fronts. Firstly the
hydrologic models, the UBC Watershed Model and the UBC Flow Model have undergone
considerable improvement to their structure and to the user interface. Secondly mountain
precipitation processes and mountain hydrological flood response have been investigated,
resulting in a better understanding of precipitation distribution in the mountains and of the
hydrologic runoff processes, and these findings have been incorporated into the computer
models. These complimentary studies have been reported in a series of publications by Loukas
and Quick, references (11 to 14).
The hydrologic model work has been strongly supported by B.C. Hydro and a full time
professional programmer, Edmond Yu, and graduate students, Loukas, Millar, Nixon and Assaf
have worked under my supervision to develop improved watershed and flow models with
excellent user interfaces in an MS Windows environment. These models are now powerful
hydrologic analysis models which can be easily calibrated to a given watershed, and are
available for use by Engineering Consultants or other researchers.
A recent Masters thesis, Zoran Micovic, September 1998, has shown that the model can be
accurately calibrated to a large number of watersheds in different climatic zones in British
Columbia. A common criticism of such models is the large number of parameters and their
variability, but Micovic has shown that the UBC Model parameters are very stable and almost
constant, except for watershed impermeability. He then finds that this impermeable parameter
can be determined from surficial geologic information. The other requirement is a good
meteorological data input for each watershed. This work indicates that the model structure
appears to be a stable and accurate representation of the mountain runoff processes, and can
therefore be applied to ungauged watersheds or can be used for regional flood studies, provided
there is accurate meteorological data and surficial geology data. The results of this work show
that very high Nash-Sutcliffe efficiencies of the order of 80 to 95% were achieved for both
calibration and verification periods, (32).
Further validation of the watershed model structure has been provided by a water chemistry
study, Hudson and Quick (31). The model makes separate estimates of surface, or near surface
fast runoff, slower interflow runoff, upper zone, and deep zone groundwater runoff. These flow
components have been used to estimate water chemistry of the watershed outflow, recognizing
that the different residence times and flow paths give rise to different chemical signatures.
Naturally occuring minerals, such as silica, calcium, manganese etc. are used as tracers to
identify the various flow components. This work will be exploited in further forest hydrology
studies of water quality. Other related work is set out in (3, 7, and 28,
Work for the Department of Fisheries and Oceans has resulted in the development of good
predictive methods for estimating river water temperatures. This work has been carried out in
conjunction with M. Foreman of the Institute of Ocean Sciences and is being used to forecast
water temperatures throughout the Fraser River and Thompson River system. This research has
already been in use for four years for real-time forecasting of both river flows and river
temperatures on a grid of about 90 river reaches of 10 km each. The results are used to make
decisions on fishing and allowable fish escapements to the river systems, Foreman et al (6)
Quick, Yu and Martin (26)
This work was particularly critical and useful this year, 1999, when there was a huge
snowmelt flood potential in the Fraser River System, and when salmon stocks were so much at
risk.

Real-time forecast updating

The objective of this work is to provide real-time forecasters with a fully automatic method
which will update the UBC Watershed Model so that it agrees with the latest recorded flows at
the start of the forecast period. For a well calibrated model, the main sources of error are (1)
precipitation measurements are missing or are not representative of the total watershed values,
(2) snowpack has been underestimated by the model during the winter accumulation period,
leading to reduced snowmelt and groundwater flows, (3) temperature inversions have led to
basin temperatures being underestimated, so that precipitation estimated to be snow actually fell
as rain, which would produce immediate runoff, etc.. Each of these error sources has a
characteristic error signal.
 Criteria have been developed by which the model will automatically determine the error
source. The model already identifies the components of the total runoff by its source, namely
rain, snowmelt and glacier melt, and by its time signature, fast (direct), medium (interflow) and
slow (groundwater), so that the percentage from any source is known. The intention is to use
this model information so that the model can examine the error signature to recognize the most
likely source of error, and then to back-calculate the necessary change , or update, in the rain,
temperature or snowpack.

Development and Uses of the UBC Watershed Model

This model has been developed at UBC, starting with early versions by Quick and Pipes in the
70’s, and later versions by Quick and a number of co-workers, culminating in the latest windows
version, developed with funding from the BC Disaster Relief Fund, the BC Science Council, and
BC Hydro. The model has proved to be a good and reliable representation of the mountain
snowmelt, glacier and rainfall runoff processes, and has been widely used by various agencies
and engineering consultants. Some recent examples are given below

1) Real-time forecasting. The model is used by B.C. Hydro for operation of the following
systems:
a) Mica Dam on the Upper Columbia River.
This is the major storage facility on the Upper Columbia and is operated as part of
the US/Canada Columbia Treaty for power supply and flood control.
b) Portage Mountain Dam (Bennett Dam and Williston Lake) on the Upper Peace
River.
c) Campbell River Project (Vancouver Island).
This coastal project is subject to large rain floods. The model is used in hourly
mode to assist in flood reduction.
B.C. Hydro are currently applying the model to all other smaller projects in their system,
as a replacement of their previous Flowcast system.

Various graduate students have worked on aspects of model development for these and other
applications, and several are now working as engineers for BC Hydro: Hamed Assaf,
PhD., Dan Nixon, MASc. and Murray Kroeker, MASc.
2) Flood forecasting. Early versions of the model were used at UBC in the 1970’s by Quick
and Pipes to provide forecasts of the Fraser River at the request of the B.C. Ministry of
Environment. Since that time, the Ministry have formed their own forecasting group,
headed by R. McNeil and R. Wyman, with responsibility for the Fraser, Skeena, Lillooet,
Elk and Similkameen river systems, and have used various versions of the UBC
Watershed Model for this work.

3) International Work.
a) Pakistan. The UBC Model was used as part of a glacier and snowmelt study of
the Upper Indus in Pakistan (IDRC funding with Wilfred Laurier University
group).
 Based on this work, the World Meteorological Organisation gave Quick and Pipes
a contract to develop a forecasting system for the Mangla Project on the Indus
system.
This work formed the basis for an IDRC project in which B.C. Hydro became the
Contractor and at UBC, Quick et al supplied the forecasting technology (UBC
Watershed and Flow models in updated form) and trained three Pakistani
engineers who completed the UBC Masters program in Civil Engineering. This
work has provided the Pakistan Water and Power Development Authority with a
forecasting system for the whole of the Upper Indus, including a high elevation
meteorological data network system. The expertise has been transferred to the
WAPDA personnel by the UBC training and by on-site training in Pakistan. Dr.
H. Assaf, BC Hydro, A. Loukas and R. Millar assisted in this modelling work (All
three were former graduate students, and now Loukas and Millar are faculty
members, one at Thesaloniki and one at UBC.

b) India. The watershed modelling work has also been applied to the Bhakra Dam -
Satluj River system in India. In 1993, UNDP funded myself to work with the
Indian National Institute of Hydrology at Roorkee. Dr. Pratap Singh also spent
time at UBC learning the technology. The model is now fully available to the
National Institute.

4) Department of Fisheries and Oceans. Collaborative work has been carried out for the last
4 years to forecast Fraser River water temperatures in real-time for management of the
salmon fisheries resource. At UBC Edmond Yu and Quick, assisted also by Violeta
Martin, graduate student, provide forecasts of 13 sub-watersheds in the Fraser system and
then process these inputs with the UBC Flow Model to provide flow estimates throughout
the Fraser-Thompson system. These flow estimates are then used by Dr. Foreman at the
Institute of Ocean Sciences to generate water temperature estimates for the next 10 days.
This process is updated each week during the June to September period. It has been
particularly critical during the last 2 years when the returning fish have faced either very
high flows or high temperatures.

Hydrological Studies using the UBC Hydrologic Models

Various real-time forecasting systems have been developed and design studies have been carried
out with the model. Some of the more major studies are:

North Saskatchewan River: Forecasting model development, Prairie Provinces Water

Board, 1972-74.
Fraser River Forecasting System: Continuing development and use since 1972 to
present, for BC Ministry of the Environment.
Columbia River: Design flood estimation for the Revelstoke Project: PMF and
Diversion Floods, BC Hydro, 1974
Columbia River: Mica Dam. Development of Operational Procedure. (with A. Pipes
and with W. Goines and D. Druce of BC Hydro.), 1976
 Peace River: Design Flood studies for possible downstream developments, BC Hydro,
1973 and 1975.
Columbia River: Murphy Creek Design Flood Study, BC Hydro, 1976-81.
Upper Indus River: Development of forecasting system, World Meteorolocical
Organisation, 1987, and IDRC , with BC Hydro, 1988 to 1996, and ongoing
Campbell River: Hourly Rainflood Forecasting System, for BC Hydro, 1992-93, and
ongoing.
Satluj River, India: Development of Forecasting System, for Indian Institute of
Hydrology, United Nations Development Program, 1993.
Kootenay River. Consultant to Acres International. Derivation of PMF for the Brilliant
Dam, 1998

List of Users of the UBC Watershed Model

B.C. Hydro Power Authority

Indian National Institute of Hydrology

B.C. Ministry of Environment: River forecast Group

Department of Fisheries and Oceans: Institute of Ocean Sciences

Acres International

Pakistan Water and Power Development Authority

Peter Ward and Associates Ltd.

Dayton and Knight Ltd.

Knight and Piesold Ltd.

Klohn-Crippen Consultants Ltd.

Wilfred Laurier University: Snow and Ice Hydrology Group.

(Note that there are also various researchers at universities and institutes that are using the
model.)
 SELECTED JOURNAL PUBLICATIONS

1. Micovic, Z. and Quick, M. C., “A Rainfall and Snowmelt Runoff Modelling Approach to
Flow Estimation at Ungauged Sites in British Columbia.”, Journal of Hydrology, 226,
December 1999, pp.101-120.
2. Loukas, A and Quick, M.C., “The Effect of Climate Change on Floods in British
Columbia”, Nordic Hydrology, 30 (3), 1999.
3. Hudson, R.O. and Quick, M.C. “A Component Based Water Chemistry Simulator for Small
Subalpine Watersheds”, Canadian Water Resources Journal, Vol. 22, No. 3, 1997, pp. 299-
325
4. Loukas, A. and Quick, M.C., "Spatial and Temporal Distribution of Storm Precipitation in
Southwestern British Columbia," J. Hydrology, 174/1-2, January 1996, pp. 37-56.
5. Loukas, A. & Quick M.C., “The Effect of Climate Change on the Hydrological Regime of
Two Climatically Different Watersheds” J. Hydrologic Eng., American Society of Civil
Engineers, Vol. 1, No. 2, April 1996, pp. 77-87
6. Loukas, A. and Quick, M.C., "Physically-Based Estimation of Lag Time for Forested
Mountainous Watersheds," IAHS Hydrological Sciences J., 41(1), Feb. 1996, pp. 1-19.
7. Loukas, A. and Quick, M.C., "A 24-Hour Design Storm for Coastal British Columbia," J.
Hydr. Eng., ASCE, Vol. 121, No. 12, Dec. 1995, pp. 889-899.
8. Quick, M.C., "Channel Routing," Section of the Can. Flood Guide, National Research
Council of Canada, ISBN 0-6660-12876-4, NRCC No. 29734, 1989.
9. Quick, M.C. and Pipes, A., "UBC Watershed Model," Hydrological Sciences Bulletin
XXI,1,3, 1977, pp. 285-295.
10. Quick, M.C. and Pipes, A., "A Combined Snowmelt and Rainfall Runoff Model," Can. J.
Civil Eng., Vol. 3, No. 3, Sept. 1976, pp. 449-460.
11. Quick, M.C. and Pipes, A., "Nonlinear Channel Routing by Computer," J. Hydr. Div.,
ASCE, Vol. 101, No. HY6, Sept. 1976, pp. 651-664.
12. Quick, M.C., "A Comparison of Measured and Theoretical Snowpack Temperatures," J.
Hydrology, Vol. 5, No. 1, Mar. 1967, pp. 1-20.
13. Quick, M.C., "River Flood Flows," J. Hydr. Div., ASCE, No. HY3, Vol. 9, May 1965, pp.
1-18.
 Precipitation Temperature

Rainfall Snowfall

Melt Computation
Infiltration
Control

Flash Behavior from High Snowmelt and Glacier melt

Intensity Rainfall

Soil Moisture Control

Impermeable Fraction Permeable Fraction

Fast Runoff
Evapotranspiration Losses

Percolation to Ground Water

Slow Runoff Very Slow Runoff

Excess Water

Medium Runoff

Reservoir Routing Control

Surface Runoff Interflow Upper Zone GW Deep Zone GW

Generated Stream Flow

 CHAPTER 1: INTRODUCTION TO THE WATERSHED MODEL1

1.1. WHAT THE WATERSHED MODEL DOES

The Watershed Model generates a computational representation of watershed behavior.

Using daily maximum and minimum temperatures and precipitation data entered by the
user, the program calculates daily watershed outflow resulting from snowmelt and
rainfall. If data is available, the model can also be run in hourly mode. In addition to
streamflow figures, the model also provides information on the accumulation and
depletion of snowpack, soil moisture budget, soil and groundwater storage values,
contributions to runoff from various portions of the watershed, and surface and sub-
surface components of runoff. These data can be viewed on the screen in graphics form,
printed, or saved to a file.

1) Given continuous meteorological input data, the model will operate continuously,
accumulating and depleting the snowpack and producing estimates of stream flow.
For calibration and verification purposes, the model uses historical meteorological
and streamflow records as reference data and calculates performance statistics on
volume and hydrograph shape reconstitution.

1.2 USES OF THE WATERSHED MODEL

The Watershed Model, although designed primarily for short term river flow forecasting,
can also be used for seasonal forecasting, for design flood estimation and for other water
supply and design flow estimation

. The accuracy of forecast depends upon:

. The current assessment of the snowpack, soil, and groundwater storage,

. The representativeness of the meteorological data to date, and

. The accuracy of the meteorological forecast.

Because the Watershed Model operates continuously if given continuous meteorological

input, it is possible to use the model for longer term forecasting using projected patterns
of weather. This capability can assess the range of possible outcomes for a whole season
of snowmelt and rain runoff once an initial watershed status has been specified. Such
seasonal forecasts can be upgraded with the recorded data as the season advances and the
possible range of outcomes will gradually narrow. An extension of this mode of
operation has been used to estimate complete years of data when the streamflow records
have not existed, but for which meteorological data has been available.

1.3 DESIGN CONSTRAINTS

The Watershed Model was developed to describe and forecast watershed behavior in
mountainous areas. This introduced several important design constraints because data in
such regions are usually scarce, particularly at higher elevations. A major design
consideration resulting from this was to provide the model with the ability to interpret
meteorological data at a point in terms of basin-wide conditions. This requires
 1-2

assumptions to be made concerning temperature and precipitation distribution by altitude

and by area. Appendix A describes the algorithms that carry out these functions.

The temperature lapse rate algorithm has been pre-calibrated using high quality data in
the Rocky Mountain and Selkirk Mountain region. These parameters have been widely
tested in British Columbia and other parts of the World, including the Himalayas, and
therefore should not be adjusted.

The precipitation distribution by altitude is an important part of model calibration and

will be discussed more fully, (see Chapter 5).

Another constraint is the availability of streamflow reference data used in the calibration
process. For headwaters and major tributaries, independent streamflow data are usually
available. For many smaller tributaries, and for tributary regions downstream of the
headwaters, no independent streamflow records may exist which allow the breakdown of
a total river basin into sub-watersheds. To calculate these streamflow contributions, the
mainstem flow measurements may have to be subtracted to give an approximate estimate
of these lateral inflows. To make a more accurate estimate of these lateral inflows, it may
be necessary to use a streamflow routing model, such as the UBC FLOW MODEL, to
deduce the streamflow contributions from each sub-area. These derived streamflow
records will generally be lower in quality because distributed lateral inflows to a channel
system may have to be treated as a lumped inflow. Also measurement errors at the two
reference stations will tend to interact to produce an enhanced random error component.

A further design consideration was to provide the model with the ability to estimate the
snowpack. Two data sources are available, precipitation measured at meteorological
stations, which are usually valley sites, and snowcourse measurements, which are
available at various elevations, from valley floor to higher snowfields. Snowcourse data
are usually available only for certain dates during the snowmelt season, perhaps at two-
week intervals. In addition, snowcourse data cannot be considered an absolute measure
of snow at that elevation. Rather, they tend to be indices of snow accumulation. On the
other hand, precipitation measurements are available on a daily basis at all seasons of the
year. For this reason, the Watershed Model uses daily precipitation inputs to estimate
snow accumulation and rainfall amounts by elevation, and it has been found that these
estimates can be quite reliable, provided that the meteorological data is of reliable quality.

The snowcourse data is used, external to the model, as calibration reference data. Plots of
snowcourse data compared with calculated snowpack estimates for the same altitude
show a reasonably consistent pattern from year to year, not in absolute but in relative
amounts. These historical comparisons can be used to give guidance on whether to
update snowpack estimates when forecasting. Generally, the model snowpack estimates
are reliable, and updating may not improve the forecast.

1.4 OVERVIEW OF SETUP PROCEDURES FOR THE WATERSHED MODEL

Before streamflow forecasting can take place three preliminary steps must be completed.
These are:
 1-3

. Obtain historical streamflow and/or meteorological data for the region and
enter them into the computer in the appropriate file format,

. Create a description of the watershed and enter it into the computer (in the
parameter file TEMPLATE.WAT), and

. Calibrate system parameters to reflect the conditions existing in the

watershed so that streamflow forecasts produced by the system using the
description are consistent with historical data.

File formats are detailed in Chapter 4. The file system has been set up so that data can be
entered into a spreadsheet, such as EXCEL, and then exported as a comma delimited file.

To develop a description of a watershed, the user must first examine topographic maps
and aerial photographs of the region in order to determine the number and area of
elevation bands, north-south orientation, percentage of forest cover, and so on. This
information may be available in Geographic Information Systems for the study area.
Once such issues are decided, the information is entered into the .WAT file. These steps
are explained in detail in Chapter 3, as is the other information contained in the .WAT
file.

Within the Watershed Model, streamflow information is contained in a file having the
extension .ADQ, which is then converted into a binary file *.BDQ. The file name is
entered by the user and should identify the watershed with which it is associated.
Meteorological information, obtained from weather stations in the watershed, is stored in
files with the extension .ADM, an ASCII file which is then converted to a binary file,
*.BDM. Again, the file name is entered by the user and should distinguish one station
from the other. The Watershed Model will handle data sets from a maximum of five
meteorological (ADM) stations and one streamflow (ADQ). The procedure for creating
these files is discussed in Chapter 4.

The process for calibrating the system to reflect a given watershed involves comparing
streamflow forecasts produced from the watershed description in the .WAT file and from
the .ADM files, with actual streamflow recorded in the .ADQ file. If forecast results are
not consistent with historical information, the semi-automatic calibration routine is used
to adjust the values of relevant parameters in the .WAT file. The process of refinement
continues until the forecast results closely reflect recorded streamflow. The calibration
process is described in Chapter 5. Once it is complete, the system is ready to accept
current meteorological data.

1.5 HARDWARE REQUIREMENTS

The Watershed Model has been developed for use on the IBM PC family of computers
and compatibles, running the MS-DOS operating system. The following is a list of the
minimum equipment required.
Computer: Pentium II or better with 450 MgHz to handle 32-bit application.
Lesser-equipped machines may still run, but the model will run slowly.
 1-4

Operating System: Windows 95 or 98, or Windows NT or 2000. The

system must be specified when ordering the Model, so that the correct
installation package can be provided.

Floppy Drive: 1.5 MB high density.

CD ROM Drive: The system can be installed from a CD, or from Emailed
files. A CD Writer is also desirable for storing output and transmitting
results.

Hard Drive: Recommended minimum 2 Gigabytes hard drive. Installation

will check available storage.

Printer: All graphics and printing is handled using the Windows Operating
System

1.6 ABOUT THE MANUAL

The Watershed Model User Manual is designed to explain the model and how to use it.
Chapter 2 describes the structure of the system, including data and executable files,
output files and so on, and how to install the model. It also describes the user interfaces,
including data input screens, special keys and related features. Chapter 3 examines the
.WAT file in detail, explaining the various sections and their function within the
Watershed System, along with key parameters. Chapter 4 explains how to set up the
.ADQ and .ADM files containing the data used during system calibration. These files are
then converted into binary files, *.BDQ and *.BDM, using the utility program provided.
Chapter 5 explains how to calibrate a watershed and the parameters critical to the process.
Chapter 6 gives an example of a watershed calibration. In addition to these chapters,
appendices are included which provide the user with detailed reference material about the
system, and, in particular, Appendix A gives information about the model algorithms
used to define the various watershed processes.
 2-1

CHAPTER 2: SYSTEM STRUCTURE AND USER INTERFACE1

2.1 HOW TO INSTALL AND RUN THE UBC WATERSHED MODEL

To install the UBC Watershed Model do the following:

 Double click setup.exe from the installation package.

 Enter the directory for installation when the setup program prompts.

To run the UBC Watershed Model, go to Start -> Programs -> UBC Watershed Model

2.2 STRUCTURE OF THE UBC WATERSHED SYSTEM

The UBC Watershed Model consists of a number of program files containing coded
instructions to the computer and a series of data files containing watershed information.
The program files are recognizable by their three-letter file extension, EXE, for example
STATS.EXE. (Note: a file is generally identified by a file name, consisting of up to 8
characters, followed by a period, followed by a three-letter extension.) There are two
types of data files: input files which contain historical meteorological and stream flow
data as well as that which contains the description of the watershed being modeled, and
output files created by the program as it runs and containing information derived from the
input files. The extensions of the input data files are Watershed parameter (WAT),
Binary Daily Meteorological (BDM) and Binary Daily Flow (BDQ), respectively. The
user supplies the filename for these, e.g. STRMFLOW.BDQ. Output data files are given
the extension Binary Daily Calibrated (BDC) or Statistics (STA). The filenames of these
output files are usually the same as the watershed (WAT) file from which the output file
is derived, unless the user chooses to assign another name. BDC and STA files are used
by the graphics and statistics programs to provide information to the user.

When running the STATISTICS option from the RUN of the top bar menu, an output
file with the name of the .BDC file used for the process is generated and stored on the
hard drive. The extension for these files is .STA, e.g., TEMPLATE.STA.

2.3 WATERSHED MODEL USER INTERFACE

User Interface is the term applied to the parts of a software package that enable the user
to interact with the program. The UBC Watershed Model utilizes screens that appear
while the program is running and mouse that is used to move around on the screens, enter
data, and select from lists and menus. Two types of windows are included, there are input
 2-2

windows that enable you to provide instructions and information to the program, and
output windows that display information generated from the Watershed Model.

In the UBC Watershed Model, the input windows have a title at the top to indicate which
program it is a part of the system. The main menu screen (Figure 2-1) comes up when
you click UBC watershed model from Programs. It has a top bar menu that lists the
available Watershed Model operations.

Figure 2-1: UBC Watershed Model Main Menu Screen

Figure 2-2: WAT File Parameter Input and Edit Screen

 2-3

Input screens, such as those found in the EDIT WAT function (see Figure 2-2), allow
user to enter and edit the model parameters. They consist of a number of data entry fields
and identifiers (such as parameter names). To enter or edit information on these screens,
use the mouse to move the cursor to the grid you wish to edit, left click the mouse button
to activate the field, type the required information, then press ENTER to confirm it. For a
complete explanation of how to edit the .WAT file, see the section entitled Edit the
WAT File in Chapter 3.

Output screens, like those shown in Figure 2-3, are displayed by double click the BDC
file from the file list shown in the left side of the window environment. This file list
displays all the relevant files included both the input and output files for the selected
WAT file. The windows graphically show calculated or measured data, such as Observed
and Calculated Flow. The user can tile the graphics windows either in vertical or in
horizontal mode. (see Figure 2-4) The number of graphics to display on screen can be as
many as the user needs provided the system resource is available. Once the graphics
windows are tiled, the user can move all the windows synchronize by clicking the
synchronized icons on the top bar menu of window. The user can click on any point of
the graphics to display the actual value for a particular component on a particular date.

Figure 2-3: Graphics Display Screen

 2-4

Figure 2-4: Vertical tile of Graphics

The user can select different components of hydrograph to display in graphic windows.
As shown in Figure 2-5, the default components for BDC or BHC file are observed and
calculated flows. The default components for BDM or BHM are maximum temperature,
minimum temperature and precipitation. For BDQ or BHQ file, the default component is
observed flows.

Figure 2-5: Select Components

 2-5

Figure 2-5: Optimization

The optimization program can help the user to speed up the calibration process. It is a
semi-automatic process, requiring some guidance from the user, who must specify ranges
for the parameters being optimized. The automation will then run automatically and will
update the WAT parameter file with the best set of parameters for that particular search,
and will also display the model performance statistics for the top ten sets of parameter
values. As shown in Figure 2-5, the WAT parameters are classified into three main
groups – Precipitation, Water and Routing. Each group has a set of parameters for user to
select in the optimization process. The defaults for those parameters are not used in the
process. The user has to click the Selected (Yes/No) grid cell to select the particular
parameter. The user has different options to specify the statistics for the specified run
period. The default statistics specification is the whole specified run period.
 3-1

CHAPTER 3: TO SET UP AND MODIFY THE WAT FILE1

3.1 EXPLANATION OF THE WAT FILE

The .WAT file describes an individual watershed and governs the execution of the
Watershed Model. A separate .WAT file must be prepared for each watershed before
starting the calibration procedure. (The calibration of a watershed is described in Chapter
5.) The user is assisted through the .WAT file preparation process by the EDIT WAT
option on the main menu.

The .WAT file is a parameter file which contains descriptions and values for each
watershed parameter. A parameter is a variable that restricts or describes some aspect of
the event or object it characterizes. For example, it can be a value that controls the
execution of the program such as the start and end dates for a forecast period; or, it can
represent a physical characteristic of the watershed such as the number of elevation
bands. A parameter can also provide a value for some aspect of a physical process, such
as the recessional constant for the albedo decay of fresh snow. Together, these
parameters provide the Watershed Model with run control instructions and a physical
description of a watershed which determines how it responds to temperature and
precipitation inputs.

During the process of describing and calibrating a new watershed the user edits parameter
values, using the EDIT WAT option on the main menu, until the .WAT file accurately
reflects the hydrological response of the watershed being modeled.

3.2 STRUCTURE OF THE WAT FILE

The .WAT file is divided into groups of parameters, each dealing with a particular aspect
of the modeling process. These groups, listed on the EDIT WAT screen, are:

. Time and date run control

. Meteorological and flow data
. Elevations and parameters for AES stations
. Description of the watershed
. Distribution of meteorological variables
. Snowmelt function
. Water distribution
. Initial conditions
. Initial values of outflows from routing storages
. Monthly parameters
. Elevation band specification
. .WAT file descriptor

In the TEMPLATE.WAT file, the parameters are assigned to default values. Some of
these, such as the physical description of the watershed, must be modified for each
watershed. Others, such as the snowmelt function variables, will rarely, if ever, be
changed. The process of editing parameter values is explained in the section entitled,
 3-2

Edit the .WAT File. The various parameter groups are explained below. The
Parameter Directory (Appendix B) provides a list of the parameters and their respective
descriptions, default values, and ranges.

Time and Date Run Control

This section contains parameters which control the start and end dates of runs, the control
of detailed output, and latitude information.

Meteorological and Flow Data

This section of the .WAT file provides information about the names and number of
meteorological (AES) stations, the name of the stream flow station (WSC), the frequency
with which data is recorded (hourly or daily), temperature lapse rate and precipitation
gradient indices, and the snow capping temperature of the upper AES station. This
information is unique to a given watershed and the initial default parameter values must
be edited by the user. The maximum number of AES stations allowed is 5. Only one
streamflow station is permitted.

It should be noted that if you have two meteorological data stations, one in the higher part
of the watershed and one in the lower part of the watershed, then it is often best to set the
gradient indices (LAPSER and IGRADP) to 2. Note that the precipitation interpolation
will only operate when rainfall occurs at both stations. If the upper station temperature
drops below TSNCAP (+2oC), there is a danger of snow-capping on the precipitation
gauge. In this case, the model will automatically revert to using precipitation data from
the lower station only, and it will lapse this data through to the upper elevations. If you
wish to make use of three or more stations, you must set the gradient indices to 1.

Elevations and Parameters for AES Stations

This group of parameters defines how the watershed responds to different meteorological
conditions. For each station these parameters provide information about the station
elevation, precipitation adjustment values, maximum temperature range, the temperature
ranges at which the sky is clear or cloud-covered, and evapotranspiration. Of these, the
parameters representing snowfall and rainfall adjustment factors, P0SREP and P0RREP,
are particularly important. They are primary precipitation adjustment parameters and are
used to adjust snowfall and rainfall at each AES station. Determining suitable values for
these parameters is part of the calibration process discussed in Chapter 5.

Watershed Description

This section of the .WAT file describes the physical characteristics of the watershed. The
information is specific to each watershed and it is usually derived from the analysis of
topographical maps and aerial photographs. For each elevation band these parameters
give information about the mean elevation, mean area, forested fraction of the band, the
density of the forest canopy, north-south orientation, glaciated area of the band, the
fraction of any glacier with south orientation, the impermeable fraction of band,
precipitation adjustment, and station temperature, precipitation and evapotranspiration
indices. Once entered for a watershed, these parameters are usually not modified.
 3-3

Distribution of Meteorological Variables

This group of parameters is subdivided into five smaller groups:

. Temperature distribution (lapse rates)

. Precipitation distribution
. Interception
. Form of precipitation
. Evapotranspiration

The default values in these groups do not often need to be changed, except for the parameters in
Precipitation distribution. These parameters govern the rate of change of precipitation with
elevation and are an important part of the calibration process.

Snowmelt Function

The UBC Watershed Model uses an energy method of melting snow and glaciers. Most of the
parameters governing this melt process are preset and should not be altered. However, be sure to
set R0GLAC equal to one if a glacier exists.

Water Distribution

The water distribution parameters consist of those dealing with water budget allocations,
the number of routing reservoirs, and the routing time constants and lags. The number of
fast rain reservoirs (N0FASR), fast snow reservoirs (N0FASS), and fast glacial reservoirs
(N0GLAC) need to be entered for each new watershed. P0PERC, P0DZSH, V0FLAS
and the time constants are important parameters in the calibration process.

Initial Conditions

When a new .WAT file is set up these parameters should take the values given in the
TEMPLATE.WAT file which should correspond to the recommended default values
given in Appendix B. After the watershed is calibrated and the user wishes to forecast
with the model, a FORECAST.WAT file can be generated. The FORECAST.WAT
will now contain initial conditions corresponding to the values of the initial parameters on
the day the FORECAST.WAT file is generated. After generating the
FORECAST.WAT file, it should be renamed and the its run dates should be reset.

The Initial Conditions parameters include snowpack water equivalent storage, snowpack
cold content storage, snowpack holding deficits, cumulative snowmelt, soil moisture
storage deficits, and albedo values. It should be noted that at the beginning of each
season the initial cumulative melt values are reset to zero. For latitudes north of the
equator this happens every October 1st. For latitudes south of the equator, this happens
every April 1st.

Generally, the model is run from October, when the watershed is usually snow free and
hydrologically inactive, through to September. During October the initial conditions can
be set to the TEMPLATE.WAT initial values. The most important initial condition is
the groundwater outflow. At other times of the year, even in the middle of the most
 3-4

active runoff, it is possible to generate a new .WAT file which will contain starting
values for snowpack, runoff storages, soil moisture status, etc., so that a calculation can
be re-started from that date. This is especially useful for forecasting and avoids much
repetitive calculation of antecedent conditions.

Initial Values of Outflows from Routing Storages

The Initial values of outflows from routing storages are treated in the same way as the
Initial Conditions parameters in that they are calculated by the Watershed Model during
the annual run. The parameters include glacial storage, rainfall and snowmelt storage for
interflow, rainfall and snowmelt storages for fast flow, and upper and lower groundwater
values.

Monthly Parameters
The monthly parameters allow the user to specify monthly adjustment factors that are
used by the program when calculating the amount of solar radiation on north and south-
facing bands, evapotranspiration rate, rainfall and snowfall. These values have been
determined from the analysis of long term temporal data and should not be altered.

Elevation Band Specifications

This parameter is used to specify the number of elevation bands in a watershed. When
you are setting up a watershed, you should enter this value before you enter any other
details of the watershed description. If you change the number of elevation bands, you
must adjust the parameters in the Watershed Description section accordingly. If you
change the number of bands while editing the .WAT file, you are automatically returned
to the Watershed Description parameter screen. The maximum number of bands
allowed is 12.

WAT File Descriptor

This information is unique for each .WAT file and is entered by the user when the file is
first set up. TITLE consists of the name and brief geographic description of the
watershed, total area, total area of glaciers, and any basin identifiers. LAT is the
approximate mean latitude of the watershed in degrees.

3.3 TO EDIT THE WAT FILE

This section explains how to edit and save the .WAT file. The editing function is the
same whether you are setting up a new file, or modifying an existing one during the
calibration process. If you are setting up a new watershed, read the section entitled
Develop a Watershed Description, later in this chapter, then follow the procedure
described in the section Set Up a New Watershed, to ensure that the description is
entered correctly.
To edit parameter values in the .WAT file, do the following:
 3-5

o On the MENU BAR, click on FILES and then open the required WAT FILE, or
TEMPLATE.WAT, if starting a new watershed.

o On the MENU BAR, click on TOOLS and then on WAT EDIT.

o Click on the required parameter set and enter the required values. Note that,
normally, it is best to use the CALIBRATION routine to arrive at optimized
parameter values.

. If you are saving the file, the current file name is displayed. You may save the
modified .WAT file under this name, or you may enter another name, then confirm it.

3.4 TO SET UP A NEW WATERSHED

Six parameter groups at the WAT EDIT screen are involved in setting up a .WAT file
for the new watershed. These groups, in the order in which they should be edited, are:
. .WAT file descriptor

. Time, date and run control

. Elevation band specification

. Meteorological and flow data

. Elevations and parameters for AES stations

. Description of the watershed

Not all of the parameter values in each group need to be modified. The sections that
follow explain which parameters should be edited and in which order.

WAT File Descriptor

The TITLE parameter is the description of the watershed. You may enter a maximum of
65 characters to identify the watershed. In most cases this should consist of the name of
the watershed, the total area in square kilometers, the glaciated area and identifiers for the
meteorological stations. The LAT parameter is to identify the approximate latitude (in
degrees) of the middle of the watershed.

Time, Date and Run Control

P0LATS - enter the approximate latitude (in degrees) of the middle of the
watershed, positive for north of the equator and negative for south

J0TIMA - enter the date on which the calibration or forecasting period is to

start in the form YYMMDD.
J0TIMZ - enter the end date for the calibration or forecasting period.
 3-6

J0PARR - enter the date on which the annual .WAT file generation should
take place. The file generation takes place automatically on the date
entered here. The file generated has the name FORECAST.WAT.

J0RLTA - enter the start date for detailed output for each band.

J0RLTZ - enter the end date for detailed output.

Elevation Band Specification

This must be entered before you enter values for the Watershed Description parameters
because EDIT WAT automatically adjusts the number of fields in the .WAT file to
accommodate the number of bands specified here. Memory use increases quickly with
the number of bands chosen. An adequate calibration can usually be obtained for most
watersheds with the number of bands chosen between 4 and 8.
N0BANS - enter a number from 1 to 12.

Meteorological and Flow Data

The parameter values for this section must be entered before Elevation and Parameter
data.

N0AESS - enter the number of AES (meteorological stations) from which

data is collected. This creates the necessary fields for each meteorological
station.

WSCNAME - enter the name of the WSC (stream flow data) file. A
maximum of 8 characters may be used. Enter VOID if no streamflow data
are available.

TICAES - specify the frequency with which meteorological data is recorded.

Any time step (divisor) may be used as long as it produces a whole number
as a dividend of 24. For example, 24 corresponds to daily data and 1
corresponds to hourly data.

TICWCS - specify the frequency with which streamflow data is recorded.

As with TICAES, any time step (divisor) may be used as long as it
produces a whole number as a dividend of 24.

Elevation and Parameters for AES Stations

Up to 5 stations can be specified for each watershed. When the number of stations
changes, the model automatically adjusts the number of fields available for data input, but
the user needs to enter the new information. For each AES station in the watershed
specify the following:
AESNAME - the name of the station, using a maximum of 8 characters.
 3-7

C0ELPT - the elevation of the station in meters.

P0SREP and P0RREP are modified during the calibration process. The remainder of the
parameters in this section are rarely changed.

Description of the Watershed

Here a column is displayed for each band specified in the preceding section (Elevation
Band Specification). Since the screen can only accommodate 6 fields, values for bands 7
through 12 are entered in the line below. Use the Tab key to move between the fields.
Press F10 to confirm when the values for all bands have been entered for a particular
parameter.

C0ELEM - enter the mean elevation of the band (m).

C0ALEM - enter the area of the band (km2).

C0TREE - enter the forested area of the band as a decimal fraction of the
total band.

C0CANY - specify the density of the forested area. This is an estimate of the
fraction of area that is shaded by the forest. For instance, coniferous will
provide more shade than deciduous in the winter.

C0RIEN - specify the relative north/south orientation of the band: 0 indicates

a northerly orientation, 1 indicates a southerly orientation, 0.2 represents a
20% southerly orientation. When calculating the fraction of "South facing"
slope, you should also include East and West facing slopes. These east and
west facing slopes receive sunshine early or late in the day and on average
receive very similar amounts of radiations as the south slopes. North,
North-East and North-West slopes should all be included in the "North
facing " category.

C0AGLA - enter the amount of glaciated area (km2) of each band .

C0AGOR - if the band contains a glacier, enter the decimal fraction of the
glacier that is south facing.
C0IMPA - enter the decimal fraction of the band that is impermeable.

I0TSTA - this index directs which meteorological station the band will use
for temperature calculations.

I0PSTA - this index directs which meteorological station the band will use
for precipitation calculations.

P0PADJ - enter a real number which will adjust the precipitation at each
band. 0.5 will increase the precipitation by 50%, 1.0 will double it.

I0ESTA - this index directs which meteorological station the band will use
for evapotranspiration calculations.
 3-8

3.5 DEVELOPING A WATERSHED DESCRIPTION

To develop a watershed description it is necessary to decide how many elevation bands it

comprises and the values the eight parameters should have for each elevation band. This
is done by surveying topographical maps and aerial photographs of the watershed in
question. The first step is to develop a hypsometric curve for the watershed.

Once you have determined the number of elevation bands into which the watershed will
be divided, the general procedure is to do the following for each band:

1. From the maps, determine the elevation of the mid-point of the band.

2. Calculate the band's area using a planimeter or use the hypsometric curve.

3. Examine aerial photographs to determine the fraction of the band area that is
forested, and the density of the forest canopy.
4. Establish the relative north/south orientation of the band.

5. Determine whether or not a glacier is present in the band. If a glacier is present,

estimate the area of the band that is covered by glaciers, and the fraction of the
glaciated area that is south-facing.

Steps 1 and 2 in this procedure are essential to setting up a new watershed. Steps 3 and 4
should at least be estimated.

Once you have compiled all of the necessary information, you are ready to set up your
.WAT file.

3.6 DISTRIBUTION OF METEOROLOGICAL VARIABLES

To accurately model a watershed in terms of its response to precipitation and temperature

changes, it is essential that the meteorological variables, measured at a point (i.e the
weather station), be correctly distributed over the watershed. The section of the .WAT
file entitled, Distribution of Meteorological Variables, deals with this problem.

As mentioned previously, most of the values in this section should rarely, if ever, be
changed. These include temperature distribution, interception, form of precipitation and
evapotranspiration. The parameters in these groups have been pre calibrated using
extensive field testing and data and have proven to remain constant for most watersheds.

The parameters in the precipitation distribution group do vary from one watershed to
another. These deal with precipitation gradient factors, and elevations of precipitation
gradient changes. Their names are: P0GRADL, P0GRADM, P0GRADU, E0LMID,
E0LHI and A0STAB. Values for these are determined during the calibration process,
described in Chapter 5.

3.7 SNOWMELT FUNCTION

The parameters in this section of the .WAT file determine the accumulation of snowpack
and its depletion through snowmelt. As with most of the parameter values in the section
 3-9

dealing with the distribution of meteorological variables, the parameter values for snow
modeling have been pre calibrated and, normally, will not be altered.

3.8 NEW WAT FILE STRUCTURE

Versions' 4.0 and more of the UBC Watershed Model use an updated .WAT file structure
different from previous versions. This new .WAT file structure is not readable by the
versions of the Watershed Model less than 4.0. And, the .WAT file structure of versions
less than 4.0 cannot be read by version 4.0. However, there is a conversion program
provided with version 4.0 under UTILITY, WAT file reformat, which can be used to
convert older .WAT files to the new format read by version 4.0 of the Watershed Model.

Certain parameters which are no longer used by the Watershed Model have been
removed, and other newer parameters have been added to the .WAT file. These include a
latitude parameter, P0LATS, a barrier height parameter, P0BAHT, an atmospheric
turbidity factor, P0CLAR, and a cold content decay factor for glacial areas. These are
described in Appendix B.
 4-1

CHAPTER 4: HOW TO SET UP WATERSHED DATA FILES1

4.1 OVERVIEW OF HISTORICAL DATA

The UBC Watershed Model uses recorded meteorological and streamflow data as inputs
to calibrate the model when setting up a new watershed. The objective is to simulate a
watershed hydrograph. Once the model is calibrated for a particular watershed (see
Chapter 5), meteorological forecasts are used to predict streamflow. The accuracy and
representativeness of the recorded data are crucial to the performance of the model. Data
can be obtained from various sources, such as the Atmospheric Environment Service
(AES), Water Survey Canada (WSC), B.C. Hydro, or other independent data sources.

The model uses daily meteorological data in ADM format (see Figure 4-10), and daily
stream flow data in ADQ format (see Figure 4-11). All hourly meteorological data and
stream flow data must be reformatted to a customized file structure unique to the UBC
Watershed Model (see Figure 4-12 and Figure 4-13). This can be done using the Convert
Standard UBC DOS to Windows Structure option under the TOOLS menu and will
be explained later.

If your data are from another source, but in computer files, compare them with the
formats described in this chapter to insure that they are usable. Computer files from other
sources which are not in the formats described below, as well as data from books or other
print sources, will need to be formatted correctly by manually entering the information
using the Excel spreadsheet. The Convert Standard UBC DOS to Windows Structure
program will check for missing dates and 9999 entries and give a warning if the file is not
usable. Any missing data must be entered by the user. Once the Convert Standard UBC
DOS to Windows Structure program has checked the data files, it will rename them with
extensions .ADM and .BDM for the meteorological files and .ADQ and BDQ for the
river flow data. These files are now recognizable by the main program.

4.2 FORMATS FOR METEOROLOGICAL AND FLOW DATA FILES

There are four meteorological data file formats recognized by the Utility: two for daily
readings and two for hourly readings. For streamflow data there are five file formats:
three for daily readings and two for hourly readings. These are discussed in the sections
below.

Meteorological Data Formats

A meteorological reading consists of a maximum and minimum temperature in degrees

Celsius and precipitation in millimetres. A file of meteorological data consists of a series
of these readings taken hourly or daily over a specific period of time. In order for a
meteorological data file to be recognized as such by the utility it must be identified by a
name and the extension .ADM. Once converted, the file name is entered in the WAT file
in the section Elevations and Parameters for AES Stations, under the parameter
 4-2

AESNAMES. A name is entered for each station from which meteorological data is to be
used.

Of the four allowable file formats there are daily and hourly formats used by the
Atmospheric Environment Service. The details for requesting data files from the
Atmospheric Environment Service are given at the end of this chapter.

The following is the procedure for extracting Climate Data from the Environment Canada
CD and converting to UBC format.

To start extracting the data from the Climate data CD, insert the CD into the drive and
view using Windows Explorer. Then click on CDEX.EXE.

Select the river basin, e.g. "Okanagan"

Select the Station, e.g. "Mt Kobau"

Set Conversion Type: Archive Format (4 Digit Year)

Elements to convert: press enter and then toggle with space bar to select Max. Temp.,
001; Min. Temp., 002 and One Day Precip., 012.

Select years to be converted.

Specify path.

Press F10 to download as *.all file.

Rename file to *.aes, and then use dos conversion program, CON_AES.EXE, to give
UBC DOS format.

 if there is any missing element in the file, a user should get a warning message on the
bottom of the screen.

 The detail of missing period can be viewed in the file called missing.txt

 all those missing elements have to be fixed first in order to convert to the UBC file
structure
Convert Dos to Windows format by using the Convert Standard UBC DOS to
Windows Structure option in the TOOLS MENU of the UBC Windows Model.

When you set up the WAT file, in addition to entering the AES file names, you must also
set the parameter TICAES to either 1 (if the readings are hourly) or 24 (if the readings are
daily). This parameter is in the WAT file section titled Meteorological and Flow Data.
This is discussed more in Chapter 3.

Streamflow Data Formats

As with meteorological readings, streamflow readings can be either hourly or daily. A

streamflow reading consists of the mean volume of streamflow for the specified time
period, measured in cubic metres per second.
 4-3

The following is the steps to download Hydat flow data and convert to UBC Watershed
Model DOS Format.

 Insert the Hydat CD and click on Hydat.exe.

 Click “display”and request “summary data”. Then search for required station
using the “station lookup”

 Click on the required station in the displayed list, and it will be displayed in the
lower window. Click on it and then click the tick symbol, which will cause
several more buttons to appear. (or click the OK button.)

 Click on “display” button at bottom of screen and Monthly Summary data will be
displayed. (You can print or save this summary, if you wish)

 Click on “export” at top of page, in header task bar and click on “comma
delimited” because there is no other reasonable choice (you can change to “card
format” on the next screen.).

 Check the information on the display. Change the report type to “Daily”, then
change Format to “card format”(you have to click the up-arrow to find Card
Format!!) and make sure it is set to download “flow”. Enter the years required.
The available years are displayed near bottom of screen.

 Click “Export” button on right of screen. (And not the header task).

 Specify the file name and the folder to save in.

 Click “Export” button on right of screen. (And not the header task).

 A message will tell you that it has been successfully exported.

 When you have exported all the files you require, close Hydat.

Re-formatting Hydat data to UBC format:

 Find the saved file and rename it to “Name”.wsc, where “name” must not exceed
8 characters.
 Open the file in Notebook or WordPad. (It is clearer in WordPad and easier to
edit.)

 Add a new line to the top of the file as follows:

3 Name of Station

The 3 must be in the first column. The name can contain any information you
find useful! But do not exceed the one line.

 Save the file into your Data Folder, (e.g. Basin Name). This folder must also
contain the conversion files, CON_WSC.EXE for the Flow Files, and
CON_AES.EXE for the Climate files.

 Double click on the program “CON_WSC.EXE”

 4-4

 The program will display the available *.WSC files. Select the file you have just
saved and press “ENTER”.

 The program may convert the file, or it may tell you that there is missing data.
This missing data information will be saved in a text file to assist you to edit it.

 All missing data must be replaced, either with zeros or with data from another
year. The date information must be corrected to run in sequence.

(For example, if winter flows are missing for several months for a station in which the
winter flows are very small, then it is reasonable to replace these months from another
year. But be careful to change the year value in each line to the correct year. Also be
careful about LEAP YEARS, when February must have 29 days of data. Perhaps select
from another leap year, if possible.)

 Make sure you have not left a blank line at the top of the file, and make sure your
inserted top line is immediately above the data, with no blank line.
 Now when you call CON_WSC.EXE again and click on the required file, it
should tell you that conversion is successful.

Conversion to Windows format can now be done using the Convert Standard UBC
DOS to Windows Structure option in Tools menu of UBC watershed model in
windows.

Once you have determined the frequency of the streamflow readings enter this
information into the .WAT file. To do this, simply use EDIT WAT parameters
submenu under Tools menu, as described in Chapter 3, to set the parameter TICWSC to
either 24, for daily readings, or 1 for hourly readings and enter the WSC file names under
the parameter WSCNAME$ in the .WAT file under the section entitled Meteorological
and Flow Data.

4.3 DATA REFORMAT OPTION

The data reformat option is used to transform the data formats into the standard formats
read by the UBC Watershed Model.
 The ADM option reformats daily or hourly meteorological data.

 The ADQ option reformats the daily streamflow data.

 The QHR option reformats the Water Survey Canada hourly #4 format.

Once formatting is complete, the old file is renamed with the extension reversed (e.g.
<filename>.AES becomes <filename>.SEA) and the new format takes the original name
and extension. This allows the user to check the data conversion.

4.4 MANUAL DATA ENTRY

If your streamflow or meteorological data are not in the formats described above, you
may have to enter them manually using the Excel spreadsheet. Enter data in the same
sequences as shown in Figure 4-10 or Figure 4-11 to the corresponding columns of
 4-5

spreadsheet, and then EXPORT the file to a comma delimit format. Rename the extension
of the saved file to ADM or ADQ, respectively. The plain text ADM or ADQ file can be
converted to binary format by running the INPUT DATA CONVERSION UTILITIES
submenu under TOOLS menu of UBC watershed model.

4.5 HOW TO ORDER AES FILES

Meteorological data may be requested from the Atmospheric Environment Service in

either daily or hourly format. You will need to specify file format, station identifier and
time period required. Station identifiers, locations, elevations, and record periods can be
found in the Atmospheric Environmental Service Climatological Station Catalogue.

For daily data, request compressed format GRF207. To define the time period you
require provide a starting and an ending month and year. These dates should be
expressed as YYMM to YYMM, (e.g., 9010 to 9310). The information being requested
(temperature and precipitation readings) must be in a particular order for the Watershed
Model. To insure this, request file elements in the following order: 001, 002 and 012.
These elements are:

001 - the daily maximum temperature (TX)

002 - the daily minimum temperature (TN)

012 - daily precipitation (PP)

For hourly data, request compressed format GRF124. The time period is expressed as
YYMMDD to YYMMDD, and the file's elements should be in the order 078 and 123.
These are:

078 - the mean hourly temperature (T)

123 - the hourly precipitation (PP)

In B.C. contact:

Atmospheric Environment Service

Pacific Regional Office
Weather Service Directorate
1200 - W. 73 Avenue
Vancouver B.C. V6P 6H9

Elsewhere contact either the B.C. office or your most convenient Atmospheric
Environmental Service office.

4.6 HOW TO ORDER WSC FILES

Streamflow data can be obtained from Water Survey Canada by specifying the type of
data (hourly or daily), CARD format, station identifier and time period expressed as
YYMMDD to YYMMDD. The Historical Streamflow Summary, also available from
Water Survey, lists gauge location, identifier, longitude, latitude and basin area, as well as
monthly and annual mean discharges.
 4-6

In B.C. contact:

Environmental Conservation Directorate

Pacific and Yukon Region
224 W. Esplanade
North Vancouver, B.C. V7M 3H7

Elsewhere contact your nearest Inland Water Directorate office.

4.7 SNOWCOURSE DATA

It is often helpful to compare the simulated snowpack properties with measured ones. In
the GRAPHICS window the UBC Watershed model will display the snowpack water
equivalent (at the same elevation as the corresponding snowcourse station) together with
a plot of a given year's snowcourse water equivalent and the mean snowcourse water
equivalent. This information is useful to determine the watershed model's distribution of
precipitation as snowfall and to determine which distribution parameters (i.e., P0GRADL,
P0GRADM, P0GRADU, P0SREP and P0RREP) to adjust. It is important to remember
that the snowcourse data is only an indicator and may not be representative of the entire
basin.

To plot the snowcourse data the model requires an .SNC file and not less than two band
files (BAND#.CAL) where BAND# is the band number. These can be created by
entering the appropriate dates for detailed band output in the .WAT file or by choosing
"Yes" when prompted to create band files in the snowcourse program. See sections 4.9
and 4.10

4.8 WHERE TO OBTAIN SNOWCOURSE DATA

In British Columbia the Ministry of Environment's Water Management Branch is

responsible for the collection of snowcourse data. Annual records have been published
since 1935 and presently there are more than 270 stations in British Columbia from
which data can be obtained. This data is available in print or on diskette. To assist the
user to extract specific snow survey measurements from the diskette the UBC Watershed
Model contains a special utility that is discussed in section 4.9 HOW TO SET UP .SNC
FILES.
To order Snow Survey Measurements Summary contact:

Hydrology Section
Water Management Branch
Ministry of the Environment
Parliament Buildings,
Victoria, B.C. V8W 9Z9

4.9 HOW TO SET UP .SNC FILES

The UBC Watershed Model provides the user with a utility to extract data from the data
file provided by the Ministry of Environment and enter it into the model as an .SNC file.
This option is found in the UTILITY directory under the Create .SNC option.
 4-7

To set up the .SNC file do the following:

1. Copy the contents of the data disk into the same directory which contains
the UBC Watershed Model.

2. While running UBCMENU select Create .SNC from the UTILITY

directory.

3. When you are prompted for a Course? type the 4 or 5 digit station
identifier, then press Enter. This station identifier can be found in the
Snow Survey Bulletin published by the Ministry of Environment, Water
Management Branch.

4. The program tells you that it has summarized the specified station data and
prompts you to enter the next station. Continue until you have entered all
the stations for your watershed.
5. After you have entered all the stations, press Enter at the Course? prompt
to exit the snowcourse program.

6. Enter a name for the .SNC file, such as the name of the watershed you are
working with, and then press Enter to confirm. The .SNC file is now
created.

4.9.1 HOW TO SET UP .SNC FILES MANUALLY

If you are using data in a format different from the Ministry of Environment's, you will
have to use the SNC Data Entry option to create a correctly formatted file. To do this
enter the UTILITY directory and select the Data Entry option. From this option select
the SNC option. Then do the following:

1. Enter an .SNC file name to be created.

2. Enter the alpha-numeric code that describes the station. Usually this is the
station number. This information is not required to run the model but it is
useful to identify the station.
3. Indicate whether the station is currently active or whether it is inactive.

4. Enter a title of your choice that describes the station's data.

5. Enter the elevation of the station and its coordinates in latitude and
longitude.

6. Indicate the period for which data exists for this station.

7. Press F10 to confirm.

8. The next window will ask for the snow depth and the equivalent water
depth for any 8 months of the year.
9. After you have entered the data for all the years you can enter another
snow survey station at the prompt. When all the data for each year is
 4-8

entered you can exit the editor by answering "No" to "Input another snow
course?"

You have now created a .SNC file by hand.

4.10 HOW TO CREATE A BAND#.CAL FILE

The BAND#.CAL file is a file that contains information for a given elevation band. This
file is necessary to run the snowcourse plotting option. To create the BAND#.CAL file
do the following:

1. From UBCMENU enter the EDIT WAT directory and select the Time and Date Run
Control option. Edit the limit J0RLTA and J0RLTZ which define the parameters of
interest. J0RLTA is the beginning of the period and J0RLTZ is the end of the period.

2. After editing, exit the EDIT WAT by pressing ESC and make sure you save your
changes.

3. Run the UBC Watershed Model by entering the UBC MODEL directory and selecting
Run UBC watershed model.

4.11 PLOTTING SNOWCOURSE DATA

Now that you have created the .SNC file and the BAND#.CAL file you can use the
Snowcourse option under GRAPHICS to plot the snowcourse data.

Simply enter the GRAPHICS directory from UBCMENU and select Plotting
Snowcourse. Then do the following:

1. You will be asked to enter an .SNC file name. Select this file from the list shown.

2. Then select the corresponding .WAT file.

3. Select a snowcourse number or station.

4. Enter the water year of interest. For example, to view October 1975- May 1976 you
would enter the water year as 1976.

5. Repeat step 4 for any other water years.

 4-10

1 GOAT R nr ERICKSON 08nh004 (1973-1977)

Q08HB048973 11 .028 .028 .028 .028 .028 .028 .028 .028 .028 .028 31
Q08HB048973 12 .028 .028 .028 .028 .028 .028 .028 .028 .028 .028
Q08HB048973 13 .028 .028 .028 .028 .028 .028 .028 .028 .028 .028 .028
Q08HB048973 21 0 .028 .028 .028 .028 .028 .028 .028 .028 .028 28
Q08HB048973 22 .028 .028 .028 .028 .028 .028 .028 .028 .028 .028
Q08HB048973 23 .028 .028 .028 .028 .028 .028 .028 .028 -1111 -1111 -1111
Q08HB048973 31 0 .028 .028 .028 .028 .028 .028 .028 .028 .028 31
Q08HB048973 32 .028 .028 .028 .028 .028 .028 .028 .028 .028 .028
Q08HB048973 33 .028 .028 .028 .028 .028 .028 .028 .028 .028 .028 .028
Q08HB048973 41 0 .028 .028 .028 .028 .028 .028 .028 .028 .028 30
Q08HB048973 42 .028 .028 .028 .028 .028 .311 .028 .028 .028 .028
Q08HB048973 43 .028 .028 .028 .028 .028 .028 .028 .028 .028 .028 -1111

Figure 4 - 5: Water Survey Division WSC File Format (daily)

 4-11
4 - illecillewaet river at Greely
Q08ND01376 129 1 8.910 8.900 8.880 8.860 8.810 8.770
Q08ND01376 129 2 8.720 8.680 8.680 8.690 8.690 8.670
Q08ND01376 129 3 8.660 8.650 8.630 8.610 8.590 8.560
Q08ND01376 129 4 8.540 8.510 8.500 8.500 8.500 8.500
Q08ND01376 130 1 8.500 8.510 8.510 8.510 8.510 8.510
Q08ND01376 130 2 8.510 8.510 8.500 8.480 8.470 8.450
Q08ND01376 130 3 8.430 8.410 8.380 8.370 8.360 8.360
Q08ND01376 130 4 8.350 8.340 8.330 8.330 8.320 8.310
Q08ND01376 131 1 8.300 8.290 8.290 8.280 8.270 8.260
Q08ND01376 131 2 8.230 8.220 8.210 8.200 8.190 8.110
Q08ND01376 131 3 8.020 7.960 7.940 7.940 7.930 7.930
Q08ND01376 131 4 7.930 7.930 7.930 7.930 7.930 7.920
Q08ND01376 2 1 1 7.930 7.940 7.940 7.930 7.930 7.930
Q08ND01376 2 1 2 7.930 7.900 7.830 7.730 7.620 7.440
Q08ND01376 2 1 3 7.360 7.360 7.350 7.330 7.280 7.290
Q08ND01376 2 1 4 7.300 7.320 7.340 7.350 7.530 7.550
Q08ND01376 2 2 1 7.570 7.580 7.790 8.110 8.170 8.160
Q08ND01376 2 2 2 8.160 8.150 8.140 8.140 8.130 8.120
Q08ND01376 2 2 3 8.120 8.110 8.110 8.090 8.080 8.070
Q08ND01376 2 2 4 8.050 8.040 8.030 8.060 8.100 8.090
Q08ND01376 2 3 1 8.170 8.290 8.410 8.450 8.440 8.440
Q08ND01376 2 3 2 8.440 8.430 8.430 8.430 8.420 8.420
Q08ND01376 2 3 3 8.410 8.200 7.880 7.850 7.820 7.850
Q08ND01376 2 3 4 7.960 8.040 8.090 8.120 8.130 8.120
Q08ND01376 2 4 1 8.110 8.110 8.100 8.100 8.090 8.090
Q08ND01376 2 4 2 7.910 7.730 7.570 7.480 7.450 7.440
Q08ND01376 2 4 3 7.300 7.010 6.870 6.860 6.860 6.860
Q08ND01376 2 4 4 6.860 6.870 6.900 6.900 6.910 6.900
Q08ND01376 2 5 1 6.930 7.040 7.100 7.130 7.130 7.120
Q08ND01376 2 5 2 7.120 7.110 7.110 7.120 7.380 7.550
Q08ND01376 2 5 3 7.540 7.470 7.390 7.380 7.410 7.520
Q08ND01376 2 5 4 7.660 8.310 13.300 11.900 10.900 10.800
Q08ND01376 2 6 1 10.800 10.800 10.800 10.800 10.800 10.800
Q08ND01376 2 6 2 10.800 10.800 10.800 10.800 10.900 10.900
Q08ND01376 2 6 3 10.900 10.700 10.600 10.500 10.600 10.600
Q08ND01376 2 6 4 10.800 11.000 11.200 11.500 11.800 12.100
Q08ND01376 2 7 1 12.300 12.400 12.500 12.600 12.600 12.700
Q08ND01376 2 7 2 12.800 12.900 12.900 13.000 12.800 12.000
Q08ND01376 2 7 3 11.600 11.300 11.300 11.300 11.300 11.300
Q08ND01376 2 7 4 11.500 11.800 12.000 12.200 12.300 12.300
Q08ND01376 2 8 1 12.300 12.300 12.300 12.300 12.300 12.300
Q08ND01376 2 8 2 12.300 12.300 12.300 12.300 12.300 12.300
Q08ND01376 2 8 3 12.300 12.300 12.300 12.300 12.300 12.300
Q08ND01376 2 8 4 12.300 12.300 12.300 12.300 12.400 12.400

Figure 4 - 6: Water Survey Division WSC File Format (hourly)

 4-12
2 CAMPBELL LK INFLOWS (M3/S) (1980 - 1991)
QSCAADJ 980 1131 74.2 89.4 70.8 64.4 43.4 42.3 33.5 27.2
QSCAADJ 980 1231 43.2 31.9 65.0 54.4 40.6 41.8 73.3 52.4
QSCAADJ 980 1331 31.3 32.1 27.7 16.4 8.6 14.5 32.8 42.2
QSCAADJ 980 1431 43.8 45.7 25.0 41.1 15.2 19.6 40.1
QSCAADJ 980 2128 29.7 114.2 129.3 61.1 62.5 44.5 42.8 72.8
QSCAADJ 980 2228 51.7 46.2 57.8 43.5 41.0 26.9 31.3 43.4
QSCAADJ 980 2328 31.8 17.3 39.1 39.3 17.7 35.8 19.0 33.0
QSCAADJ 980 2428 58.3 260.9 224.6 251.2 177.6
QSCAADJ 980 3131 143.2 104.7 83.4 79.7 79.9 64.3 54.0 54.1
QSCAADJ 980 3231 48.0 61.1 38.5 24.6 73.3 22.7 31.5 30.7
QSCAADJ 980 3331 113.1 74.6 41.0 33.8 36.6 41.0 33.5 28.5
QSCAADJ 980 3431 46.4 39.2 28.3 34.1 45.1 36.0 9.5
QSCAADJ 980 4130 47.3 40.9 20.4 20.5 40.0 47.9 25.0 53.8
QSCAADJ 980 4230 86.1 32.3 30.3 40.0 52.7 86.5 86.0 80.2
QSCAADJ 980 4330 83.2 135.8 293.3 143.6 93.5 74.8 70.4 63.0
QSCAADJ 980 4430 74.4 54.6 57.0 80.7 67.5 85.3

Figure 4 - 7: B.C. Hydro WSC File Format (daily)

1990
901021 SCA H INF 000 O 25 S N 0 110.91
131.61 131.61 131.61 131.61 131.61
131.61 103.70 103.70 103.70 103.70
103.70 103.70 131.90 131.90 131.90
131.90 131.90 131.90 76.42 76.42
76.42 76.42 76.42 76.42
901022 SCA H INF 000 O 25 S N 0 76.67
103.99 103.99 103.99 103.99 103.99
103.99 48.88 48.88 48.88 48.88
48.88 48.88 104.62 104.62 104.62
104.62 104.62 104.62 49.20 49.20
49.20 49.20 49.20 49.20

Figure 4 - 8: B.C. Hydro WSC File Format (hourly)

 4-13
3 illecillewaet River at Greely (1963 - 1990)
Q08ND013963112 19.0X 18.5X 18.0X 17.5X 17.0A 15.0A 15.0 13.9 13.9 11.6
Q08ND013963113 12.3 14.4 15.0 15.0 13.3 21.3 24.4 18.3 15.8 14.8 -1111
Q08ND013963121 13.0 13.1 12.5 13.3 15.3 15.3B 14.7B 13.0B 11.3B 10.2B 31
Q08ND013963122 9.91B 10.3B 11.3B 12.2B 13.0 12.5 13.9 16.5 18.5 22.7
Q08ND013963123 21.1 12.0 10.1 7.48 10.6 12.0 10.9 10.3 10.1 9.49 9.63
Q08ND013964011 9.63 9.34 8.95 9.06 8.83 8.69 8.55 7.93 8.55 9.49 31
Q08ND013964012 10.6 10.8 11.6 10.3 9.20 8.33 9.49 9.34 9.34 8.69
Q08ND013964013 8.33 9.49 9.77 9.91 9.49E 8.78E 8.21E 8.07E 7.99E 7.93E .93E
Q08ND013964021 7.79E 7.59E 7.33E 7.25A 7.14 6.80 6.68 6.91 6.91 6.80 29
Q08ND013964022 6.80 6.23 6.68 6.80 6.46 6.57 6.57 6.57 6.46 6.00
Q08ND013964023 6.34 6.00 5.78 6.23 5.32 7.02 6.80 7.02 6.57 -1111 -1111
Q08ND013964031 6.12 6.23 6.00 5.89 5.44 5.66 4.87 5.89 6.00 5.66 31
Q08ND013964032 5.66 5.44 5.44 5.55 5.55 5.66 5.66 5.66 5.21 5.55
Q08ND013964033 5.32 5.32 4.53 3.94 5.89 5.66 5.10 5.32 5.89 6.80 8.55
Q08ND013964041 11.2 12.0 11.3 11.5 11.0 11.8 13.1 14.1 15.5 17.2 30
Q08ND013964042 17.7 16.3 15.0 15.0 14.4 13.8 13.0 12.5 12.2 12.7
Q08ND013964043 13.3 13.9 13.4 14.1 16.8 17.0 17.0 16.7 18.8 21.1 -1111
Q08ND013964051 21.5 22.1 23.4 31.1 42.8 42.8 39.9 47.6 57.2 60.9 31
Q08ND013964052 56.6 53.8 61.2 55.5 50.7 53.2 66.5 80.1 106 129
Q08ND013964053 147 103 79.0 69.9 59.7 57.2 68.8 89.2 115 133 145
Q08ND013964061 161 171 180 171 169 180 161 158 172 180 30
Q08ND013964062 191 204 200 212 202 203 203 181 163 155
Q08ND013964063 150 152 174 192 168 158 176 158 123 115 -1111
Q08ND013964071 130 157 190 197 203 193 205 246 314 241 31
Q08ND013964072 212 237 244 236 244 204 151 145 138 129
Q08ND013964073 141 133 113 101 103 111 113 115 133 132 151
Q08ND013964081 107 98.8 89.8 141 127 100 96.8 100 115 101 31
Q08ND013964082 95.7 105 133 108 93.2 88.9 87.8 94.3 85.0 73.6
Q08ND013964083 69.7 69.7 67.1 59.7 62.9 72.8 67.1 62.6 56.6 52.7 47.0
Q08ND013964091 44.2 43.9 43.6 41.1 40.5 37.9 35.7 36.8 36.5 34.8 30
Q08ND013964092 31.7 30.3 29.7 29.2 30.6 39.9 109 63.1 50.4 47.9
Q08ND013964093 43.9 45.0 42.2 70.8 126 73.3 59.2 51.0 56.6 82.4 -1111
Q08ND013964101 75.0 73.1 66.8 58.3 59.5 57.2 60.0 61.7 60.9 62.0 31
Q08ND013964102 56.1 53.2 53.2 72.5 59.7 52.1 48.1 43.6 41.6 40.2
Q08ND013964103 38.2 36.2 34.5 33.4 32.0 30.3 28.0 27.0 28.9 33.1 28.9
Q08ND013964111 46.7 46.4 37.9 37.9 35.4 33.4 32.0 29.2 28.6 28.6 30
Q08ND013964112 28.0 26.2 25.0 23.8 24.0 22.7 22.7 22.1 21.5 21.1
Q08ND013964113 20.3 19.8 19.8 20.1 19.4 19.0 18.7E 18.1E 17.3E 16.4E-1111
Q08ND013964121 19.2A 18.6 17.4 16.3 15.8 15.5 14.8 15.6 15.5 15.1 31
Q08ND013964122 14.8 14.1 13.6B 13.2B 12.7B 12.5B 11.9B 11.5B 11.0B 10.8B
Q08ND013964123 10.3B 10.2B 9.91B 9.85B 9.77B 9.77B 9.77B 9.91B 10.1B 10.6B
Q08ND013965011 10.6B 10.8B 10.9B 11.0B 11.0B 11.0B 11.0B 11.0B 11.0B 1.0B31
Q08ND013965012 10.9B 10.8B 10.8B 10.6B 10.6B 10.5B 10.3B 10.2B 10.1B 9.91B
Q08ND013965013 9.77B 9.63B 9.49B 9.34B 9.34B 9.34B 9.34B 9.40B 9.06B8.64B

Figure 4 - 9: Water Survey Division WSC File Format (CARD - daily)

 4-14

Figure 4-10: ASCII DAILY METEOROLOGICAL (ADM) STRUCTURE

The first line of the ADM file must contain the start and end dates for meteorological data, as in
the format as shown. The first column of the ADM file can include a station identifier which is an
optional, but if left blank, the column must have at least one blank space and a comma before the
date column. The date should be constructed in yyyy/mm/dd format. From the second line
onwards, the data line includes station identifier, date, maximum temperature, minimum
temperature and precipitation. All data are delimited by commas.
 4-15

Figure 4-11: ASCII DAILY FLOW (ADQ) STRUCTURE

The ADQ file structure is similar to the ADM file structure. The first line of the file contains the
same content as ADM file. From the second line onwards, the data line includes station identifier,
date and recorded flow data. All data are delimited by commas.
 4-16

Figure 4-12: ASCII HOURLY METEOROLOGICAL (AHM) Structure

 4-17

Figure 4-13: ASCII HOURLY STREAMFLOW (AHQ) Structure

 5-1

CHAPTER 5: CALIBRATION OF THE WATERSHED 1

The calibration of a watershed involves adjusting parameter values in a newly set up

.WAT file until the annual and monthly volumes and patterns of runoff derived from this
file approximate historical records. This procedure does require some understanding of
hydrological processes, and in some cases a more detailed understanding of the
Watershed Model. We are always available to help. If you would like our assistance
with calibration, send us some data. (A year is often sufficient.)

The .WAT file consists of a large number of parameters which have been discussed in
Chapter 3. However, only a small number of these parameters are necessary for the
calibration of a watershed.

It is recommended that you begin with the TEMPLATE.WAT file provided. The
majority of the parameters have default values which should not be changed. For
example, the snowmelt parameters have been pre-calibrated; similarly, the temperature
lapse rate parameters rarely need to be adjusted.

The main calibration parameters are discussed in this chapter. These should be calibrated
using the OPTIMIZATION ROUTINE. While using the optimization routine, it is best
to test a limited number of parameters at a time.

The OPTIMIZATION ROUTINE IS SEMI-AUTOMATIC and is found under the

TOOLS on the MENU BAR. The routine makes a series of model runs for the specified
calibration period, and each run is evaluated statistically. The best 10 results are
automatically ranked and the best set of parameters are written into the new *.WAT file.

 There are three groups of parameters, PRECIPITATION, WATER, AND ROUTING,

which are used in the calibration process. The user can select which parameters will be
used, and can specify ranges for each parameter. The default values are shown as a guide,
and should be used as a mean value of the selected range.

 The first group of parameters, PRECIPITATION, contain the very important parameters,
such as POSREP etc. and POGRADL etc., which control the amount and distribution of
precipitation, both rain and snow. These parameters have the biggest influence on the
volume of runoff, and is the most important step in achieving a calibration. Try this step
first and you will see that the statistical results will improve considerably.

 The results of this first calibration are viewed from the VIEW tab on the MENU BAR in
the OPTIMIZATION REPORT.

 The second group of parameters are in the WATER group. They control the soil
behaviour, which controls the subdivision of runoff into FAST, MEDIUM AND SLOW
components of runoff. This represents the main runoff response of the system, and
controls the non-linear behaviour of the watershed. COIMPA is the most important
parameter which controls the degree of impermeability, which determines the fast runoff
response of the system. POPERC controls the infiltration to GROUNDWATER, which
 5-2

is then divided into fast and slow groundwater by PODZSH. Select this group of
parameters, specify ranges for each, and carry out the next stage of optimization.

 The third group of parameters, ROUTING, control the time distribution of runoff. The
default values that have been determined from studying many watersheds usually give
reasonable results. Test a moderate range of values, based on the suggested default
values, for each parameter and carry out the optimization.

 This optimization process can be carried out for all three parameter groups, and they will
be run automatically and sequentially. However, when first carrying out an optimization
for a new watershed, it may be best to do one step at a time and examine the results both
statistically and graphically.

Evaluate the results of the run using the GRAPHICS and STATISTICS options to
determine the degree to which the estimated hydrograph agrees with recorded streamflow
behaviour. Repeat the process until suitable parameter values are established.

 The procedure is carried out in three stages, each dealing with different groups of
parameters, but following the same iterative pattern of modification and evaluation.
Stage one is concerned with the meteorological distribution parameters and is carried out
first in a very simplified way to determine a working basis for further calibration, and
later to achieve greater parameter refinement. Stage two deals with the time distribution
of runoff and stage three assesses gradients of behaviour in the watershed. Stages two
and three are refining processes, which should not be attempted until the annual and
monthly volumes of runoff, calculated in stage one, correspond closely to historical data.
After completing stages two and three, the user may return to the first stage to make any
final parameter corrections.

 In the early stages of calibration, the seasonal and annual output statistics can adequately
monitor the improvements in the state of the calibration. They cannot, however,
represent individual events such as severe one or two day rains which become significant
during the latter stages of calibration. Visual comparison of estimated and observed
flows are useful especially when calibrating the model for runoff timing.

5.1 SIMPLIFIED STAGE ONE CALIBRATION

The parameters important to stage one calibration are E0LMID, E0LHI, P0GRADL,
P0GRADM and P0GRADU found in the section Distribution of meteorological
variables in the .WAT file. P0SREP and P0RREP, found in the AES Station elevations
and parameters section, are often considered, as well.

P0GRADL is the precipitation gradient for elevations below E0LMID. P0GRADM is the
precipitation gradient for elevations above E0LMID and below E0LHI. P0GRADU is the
precipitation gradient for elevations above E0LHI. E0LMID is often taken as the
elevation of the middle of the barrier height (1/2 H). E0LHI is usually estimated as two-
thirds of the barrier height.

For simplified stage one calibration the only parameter considered is P0GRADL. The
remaining stage one parameters are either left at 0 (P0SREP & P0RREP) or given values
greater than the maximum values for the watershed (E0LMID & E0LHI), thus rendering
other parameters inoperative (P0GRADM & P0GRADU).
 5-3

As mentioned earlier, P0SREP and P0RREP are precipitation representation factors for
snow and rain, respectively. Default values of 0 are assigned automatically to these
parameters for each AES station in the watershed when the .WAT file is set up.

The procedure for simplified stage one calibration is as follows:

 Set E0LMID and E0LHI to values greater than the maximum

elevation by entering these values in the Distribution of
meteorological variables section of the .WAT file, using WAT
EDIT, described in Chapter 3. P0GRADL is automatically set to the
default value of 5 and should not be changed until a preliminary run
has been performed.

 Carry out the OPTIMIZATION and then examine the optimization

report. More detailed statistics can be generated fro the final result by
running STATISTICS from the RUN menu.
 View the statistics report from the STATISTICS menu.

 In the statistics report, the column entitled Tot Qobs contains the
monthly and annual runoff figures taken from the .WSC file. The
column entitled Tot Qest contains the figures calculated by UBC
Model. Comparison of Tot Qobs and Tot Qest indicates how well
the precipitation gradients distribute the precipitation.

Notes on some of the factors.

To evaluate the elevation distribution of precipitation in the watershed P0GRADL should

be tested. For example, P0GRADL is given a value which is a percentage of precipitation
change in the elevation. A typical value might be 4, which is a 4% increase per 100
meters. Various trial values should be used and volumes of calculated annual and
monthly runoff should be compared with measured values. As the calibration procedure
develops, evidence from snowcourse and seasonal runoff distribution can be used to
determine the best-fit value for P0GRADL.

As stage one calibration proceeds, evidence may indicate that refinement of P0SREP and
P0RREP is necessary. Normally, in watersheds where the available data for the basin is
reliable and elevation-representative, these parameters can be left at 0, which means that
the measured precipitation data is not changed. In some situations, there may be strong
evidence that the measured precipitation should be increased or decreased to match the
calculated watershed values. For example a value of -0.15 for P0SREP would decrease
the recorded snowfall by 15%.

At a later stage in calibration it may become apparent from the hydrologic response or
from snowcourse information, or a combination of these, that precipitation gradients are
higher or lower at various elevations. To adjust for this, P0GRADM can be activated by
setting E0LMID to an elevation value below which P0GRADL will operate and above
which P0GRADM will govern. Similarly P0GRADU can be activated by setting E0LHI.

Once annual and monthly volumes are close to the desired levels, the next step is to
determine if the time distribution of runoff is the same for the results generated by the
UBC Model as for historical data. The GRAPHICS option on the main menu provides a
visual representation of the shape of the runoff. Using the display options, the user can
observe the characteristics of the estimated and observed flows.
 5-4

The key factors in stage two calibration are C0IMPA, P0PERC, P0DZSH, P0UGTK, and
P0DZTK. C0IMPA is the fraction of impermeable area in a band and it controls the
amount of water entering the sub-surface; P0PERC allocates how much of this volume of
water entering the sub-surface can be stored in the groundwater (any excess goes to the
interflow components), P0DZSH divides the groundwater component into an upper and
deep zone component, and P0UGTK and P0DZTK are the routing time constants for
these two zones, respectively. See Appendix B for suggested starting values for these
parameters.

The volumes of runoff and the shapes of the groundwater recession flows are then
examined, particularly at the end of the summer for P0UGTK and through the winter for
the longer recession P0DZTK.

The semi-automatic optimization will usually yield a good calibration. However, the
results can be evaluated, and perhaps improved, by examining the graphical output and
observing the calculated and recorded hydrographs as follows:
. Determine the recessional constant (P0DZTK) for deep-zone groundwater
reservoir from recorded winter flows (January - March).

. Determine required seasonal allocation to deep zone groundwater storage

(equals P0PERC*P0DZSH).

. Determine recessional constant (P0UGTK) for upper zone groundwater

reservoir from late fall or early winter flows (October - January).

. Adjust groundwater percolation P0PERC to calibrate allocation to upper

groundwater storage (equals P0PERC*(1-P0DZSH)).

. Adjust C0IMPA with standard fast runoff shape parameters to determine

surface runoff allocation.

. Adjust soil moisture deficit production parameter P0EGEN and the

impermeable area recessional parameter P0AGEN to reconstitute summer
rainfall runoff.

. Adjust recessional parameter for interflow to account for time distribution

of residual runoff components inputs.
. Repeat any of the above steps, as required.

Time distributions of runoff calculations divide rain and snowmelt into fast, medium and
slow runoff components. The parameters in the Water distribution section of the .WAT
file are the ones involved in these calculations. To deal with these components the user
should divide the total annual period into sub-periods that reflect most clearly the separate
influence of each. Groundwater components should be analyzed first, then fast surface
runoff components, and finally the interflow component. As a general rule, fast runoff
must all occur within a period of a week or slightly more, medium runoff (interflow)
generally occurs over a period of several weeks or months, and only the slow
(groundwater) components can contribute significantly to flows many months after
recharge occurs.
During the winter, the only significant runoff contributions that normally occur result
from the runoff components with longer time recessions. In the more active summer
 5-5

runoff period, the composite effect of all runoff components cannot be broken down into
its constituent parts.

5.3 STAGE THREE CALIBRATION

Stage three of the calibration process deals with gradients of behaviour in the watershed
(C0IMPA). It is a fine-tuning process that can only be carried out when stages one and
two are at a high level of model performance. Gradients of behaviour can be studied and
assessed because the snowmelt process tends to be most active in a few elevation bands at
a time, starting at low elevations early in the season and progressing upwards as the
snowline recedes during the mid and late snowmelt season.

5.4 HOW TO USE THE STATISTICS AND GRAPHICS FUNCTIONS

The STATISTICS and GRAPHICS options on the main menu of the UBC Watershed
Model allow the user to see the results of running UBC Model for a particular watershed.
You must know how to use them in order to track the progress of watershed calibration.

Generate and Review Watershed Statistics

The STATISTICS option produces various statistical data for each month from the
.CAL file and the results are stored in a data file with the extension .STA. This file can
be printed as a report using the Print option on the main menu or displayed on the screen
using the View option. Statistics include the mean observed flow averaged over the
stated period, the mean estimated flow averaged over the stated period, the total observed
flow, the total estimated flow, the difference between the observed flow and the estimate
flow, the coefficient of efficiency and the coefficient of determination.

The coefficient of efficiency, e!, relates how well the estimate hydrograph compares in
shape and volume to the observed hydrograph, and it is calculated as follows:

 Qobs  Qesti 
2
i
e! 1  i 1

 Qobs 
n
2
i  Qobs
i 1

where,

1
Qobs   Qobs
n 

n = the number of days for daily runs or hours for hourly runs

Qobsi = the observed flow on day (hour) i

Qesti= the estimated (calculated) flow on day (hour) i

 5-6

An efficiency of 1.0 means a perfect fit, i.e. that the two hydrographs are identical. An
efficiency of less than 1.0 means imperfect shape, volume, or timing. Efficiencies can be
negative. The efficiency result is more sensitive to large peaks than small underlying
flows.

The coefficient of determination, d! or R2, is a factor that only relates how well the shape
of the estimated hydrograph corresponds to the shape of the observed hydrograph, and it
is independent of volume. However, timing does affect the results of this statistic. For
example, if there were two identical hydrographs one over top of the other, the coefficient
of determination would be 1.0. If one of these hydrographs was shifted slightly to the left
then the coefficient of determination would decrease. The coefficient of determination is
calculated as follows,
n

 (Qobs  (b Qest  a)) i i

2

d! 1  i 1
n

 (Qobs Qobs)
i 1
i
2

where,

1 n n

a  Qobsi b Qest i 
n  i 1 i 1 

n n n

 Qobsi Qesti    Qest i  Qobsi

1
n
b i 1 i 1 i 1
2
n
1 n

 Qest     Qesti 
2
i
i 1 n  i 1 

To generate a statistics report, do the following:

. Select STATISTICS from the main menu, then select Run Statistics.

.may press Escape to return to the STATISTICS menu.

. To review the statistics report on the screen, select View Statistics

Report, then select the desired .STA file from the list that is displayed.
 5-7

Figure 5-2: Screen Display of Statistics Report

Using the Statistics Report

The Statistics Report (Figure 5-2) provides the user with a summary for the period
selected (last line) and the monthly values to indicate the model's performance monthly.
Comparison of Tot Qobs and Tot Qest on the last line indicates if the estimated total
volume resembles the actual total volume. The coefficient of efficiency is a measure of
how well the estimated hydrograph compares to the actual hydrograph. It is sensitive to
both the shape of the hydrograph and the volume of runoff, as well as the phase of the
estimated hydrograph to the observed hydrograph. Efficiency varies from negative values
to plus one with one being the highest efficiency. The coefficient of Determination is a
measure of how well the regression of Qest and Qobs compares to the actual flow. This
factor is only sensitive to shape and phase. When in phase with the observed
hydrograph, the coefficient of determination will tell you how well the shape of the
calculated hydrograph compares.

View Watershed Graphics

The GRAPHICS function displays information about watershed behaviour in the form of
graphs that can be viewed on the screen, printed, or sent to a file. The items available to
be displayed and their units of measurement are:
. Observed flow (measured in cumecs)
. Calculated flow (cumecs)
 5-8

. Difference between observed and calculated flow (cumecs)

. Snowmelt outflow (cumecs)
. Glacial contribution (cumecs)
. Rainfall outflow (cumecs)
. Groundwater (cumecs)
. Solar radiation (calculated - Langleys/day)
. Albedo (calculated as percentage)
. Area of snowcover (percentage of total area)
. Evapotranspiration loss potential (millimeters/day)
. Loss to evapotranspiration and soil (cumecs/day)
. Cloud cover (calculated - percentage)
. Snowpack water equivalent (millimeters)
. Temperature - maximum (oC *10)
. Temperature - minimum (oC *10)
. Interception (precipitation * 10 mm)
. Snowfall (millimeters * 10)
. Rainfall (millimeters * 10)
. Snowmelt (millimeters * 10)

When you select items to display, the scale used is the same as the units of the top item
tagged. For example, if you select Observed and Calculated flow and Maximum and
Minimum temperatures, the scale units are cumecs. However, the values in the scale will
be correct for the tagged items.

To select and display watershed information using GRAPHICS, do the following:

. Select GRAPHICS from the main menu.

. The list of items available for plotting is displayed be used) or

type another title, then press Enter.

5.5 OPTIMIZATION : A Brief Summary.

The optimization option is designed to assist the user in the calibration of a watershed by
removing the tedium of having to vary parameters and re-run the model. It is still helpful
if the user has some knowledge of the hydrological processes involved in watershed
modeling. Through experience and judgment the user can define reasonable ranges for
the parameters and let the computer do the work of finding the best values for them.

The optimization module optimizes three groups of parameters separately. The first
group, precipitation distribution, adjusts the precipitation gradients. This optimization
is usually done until the estimated volumes begin to compare with the observed values.
Once the annual volumes of observed and estimated flow are reasonably close, the second
group, water distribution, distributes the rain and the snowmelt to groundwater,
interflow, and surface runoff through the soil moisture budget. The third group of
parameters, the routing constants, adjusts the time constants for each component of flow
controlling the length of time taken to pass through the watershed.
 5-9

Once a range is defined for each parameter, the optimization routine randomly selects
values for the set of parameters tagged in each group, uses these values to execute the
model, and then calculates the coefficient of efficiency, the coefficient of efficiency
compensated for volume error (EOPT!), the coefficient of determination, and the
estimated volume for the period. This procedure can be run for any number of iterations,
at the end of which the ten best efficiencies and the corresponding parameters are saved
to a .OPT file. The .WAT file is then updated with the parameter values giving the best
efficiency. This updated .WAT file is then used to optimize the next optimization group.
When all the selected groups are processed the resulting .WAT file will contain the
parameter values that give the best efficiency for that watershed. The whole process can
be repeated to obtain better efficiencies.

When the optimization process is complete the results of each tagged group are viewed
under the option, View Optimization Report. The ten best efficiencies with
corresponding parameter values are kept for each group along with the .WAT file name,
dates, and max/min values of each parameter.
Note that the best parameters to pass to the next group are selected on the best efficiency
defined as EOPT!.

EOPT! = -ABS( 1 - {Qest/Qobs}) + E!

where: ABS - is the absolute value

Qest - is the estimated flow

Qobs - is the observed flow

E! - is the Nash-Sutcliffe efficiency

5.6 SENSITIVITY ANALYSIS

Under the menu UBC MODEL the Watershed Model includes an option to test the
sensitivity of certain parameters on a selected watershed. This can be helpful in
determining which parameters should be given attention during a manual calibration or an
optimization calibration.
During a sensitivity run, you may select a number of parameters from the pop up menu by
using T to tag. After entering a range for each selected parameter to be tested through,
you will be prompted to enter the number of iterations that you desire. Eleven iterations
for one year of data are usually enough to determine how sensitive a parameter is and the
range in which the optimum value is located. Keeping all other parameters remaining at
their original values as entered into the .WAT file, the program will then run the range of
each parameter through the model and calculate the efficiency (E!), Coefficient of
determination (R2), and volume error (V) for each value of each parameter.

To view the results of this procedure, use the OPTGRAPH option under GRAPHICS
(see Figure 5-5). You can view each successive parameter you tagged by depressing the
space bar.
The more horizontal an optgraph is the less sensitive that parameter will be for the range
specified and the particular watershed being used.
 5-10

The volume error, V, is the absolute value of the difference,

Qobs  Qest
V 
Qobs

Once these components are viewed, the most sensitive range for each parameter can be
determined from the optgraph. The sensitivity analysis can be performed again with these
narrower ranges, or better yet, they can be used in the optimization option.

It should be noted that many of the parameters are interdependent, and therefore changing
the value of one parameter and keeping the other parameters constant may give a different
sensitivity plot if the value of one of the other parameters is changed.

Also, some of the parameters have been generalized so that P0SREP and P0RREP are
generalized to receive the same value for each station, as well as C0IMPA and C0RIEN
which receive the same value for each band.

Experience has shown that P0SREP and P0RREP are usually the most sensitive
parameters, and hence are excellent calibration parameters. These two parameters mostly
affect the volume of flow. Other volume sensitive parameters include P0GRADL,
P0GRADM, and P0GRADU used in conjunction with E0LMID. Parameters which affect
the shape of the hydrograph include the time constants, C0IMPA, as well as P0SREP and
P0RREP. Parameters which affect the phase of the hydrograph are C0RIEN, C0IMPA
used in conjunction with P0PERC, and the time constants.

5.7 VIEWING MELT COMPONENTS

Sometimes it is helpful to know which melt components are contributing at what time.
Aside from rain melt, there are four melt components: short wave melt, long wave melt,
convective melt, and condensation or advective melt. How these are calculated is
discussed in Appendix A. To view the melt components, run the UBC Watershed Model
and when prompted to plot the energy components enter Y. You will then be asked to
specify which band you wish to see the melt/energy components for. Choose this
carefully as some bands do not contribute very much to snowmelt.
After the program is finished running, you may go the GRAPHICS menu and select Plot
energy graph from the menu. Please enter the start date and end date from when you
wish to view the energy/melt components. Type the space bar to view each melt
successive melt component.
 Appendix A - 1

APPENDIX A: WATERSHED MODEL ALGORITHMS1

A.1 SUMMARY OF THE WATERSHED MODEL STRUCTURE

The Watershed Model is designed to operate on daily or hourly meteorological data

inputs of maximum and minimum temperatures and precipitation. From these inputs the
Watershed Model calculates estimates of daily watershed outflows. Additional
information such as snowpack, soil moisture deficit, and groundwater storage values is
also computed.

The watershed model is designed primarily for mountainous watersheds and calculates
the total contribution from both snow and glacial melt as well as rainfall runoff. These
are calculated on an elevation band by band basis. The model has been used for
watersheds ranging from a few square kilometers up to areas of several thousand square
kilometers. The factors influencing the choice of watershed size are the available
streamflow reference data for calibration and the available meteorological data base.

In general, the meteorological data base is sparse for most of the mountainous regions
modeled. In the majority of situations the meteorological data is from valley stations. As
a result of these data constraints, an important aspect of the Watershed Model is the
elevation distribution of data. Functional relationships are specified describing the
variability of temperature lapse rates. The temperature lapse rate is a key relationship
because it influences the precipitation distribution, and it is also very significant in
determining snowmelt rates at various elevations. Precipitation inputs are made
functionally dependent on elevation and on temperature regime. This functional variation
of precipitation automatically recognizes that precipitation undergoes greater orographic
enhancement during the winter than it does during warm summer rainstorms. The
importance of these temperature and precipitation gradients is illustrated in the watershed
calibration discussed in Chapter 5. The general structure of the Watershed Model is
indicated in the flow chart, Figure A-1.

The response of the watershed to snowmelt and rainfall is controlled by a soil moisture
model. The soil moisture status of each area-elevation band controls the subdivision of
the total snowmelt and rain input into the various components of watershed runoff
response. These components of runoff can be characterized as fast, medium, slow and
very slow runoff, and they may be conceptually thought of as representing surface runoff,
interflow, and superficial and deep groundwater components. The total snowmelt and
rain input to each watershed band is subdivided on a priority basis, for example, a first
priority is the satisfaction of any soil moisture deficit which arises continuously because
of evaporative demand. The various priorities of the components of runoff are described
in the detailed outline of model behaviour.
 Appendix A - 2

Each component of runoff undergoes delay before reaching the outflow point of the
watershed. These delays, or time distribution runoff, are achieved by using unity
hydrograph convolution. As explained later, the various delay processes, or time
distribution processes, can be thought of in terms of cascades of linear reservoirs.

The following section outlines the various algorithms used to describe the processes
involved in the runoff process.

A.2 METEOROLOGICAL DATA DISTRIBUTION ALGORITHMS

Meteorological data is available as point values at given elevations in a watershed. The

model can accept as many as five meteorological stations. These meteorological data are
distributed over the mid-elevation points of each elevation band.

Temperature Lapse Rate

Lapse rates are known to be quite variable, ranging from high values of about the dry
adiabatic lapse rate (10 oC/ 1000 m) to low values representing inversion conditions. A
complete and detailed representation of lapse rate variability is not possible within the
model, but the main features of lapse rate can be represented as a function of daily
temperature range.

The following major features of lapse rate variation are recognized by the temperature
lapse rate algorithm:

1. During continuous rainstorm conditions the lapse rate will approximate the
saturated adiabatic rate. Under these conditions the daily temperature range will
tend to collapse to zero .

2. Under clear sky and dry weather conditions, the lapse rate during the warm part of
the day will tend to the dry adiabatic rate. During the night, under these clear sky
conditions, radiation cooling will cause the temperatures to fall to the dew point
temperature, and this is particularly true for a moist air mass. As a result, night-
time lapse rates under clear skies will tend to be quite low, and at times even zero
lapse rates will occur.

Based on these considerations, two lapse rates are specified in the model, one for the
maximum temperature and one for the minimum temperature. The lapse rate is
calculated for each day using the daily temperature range (diurnal range) as an index. The
functional relationships are shown in Figure A -2. The general form of the equations
used in the program are given below.
 Appendix A - 3

Maximum Temperature Lapse Rate - TXLAPS

TXLAPS = TZLAPS + (TLXM - TZLAPS) * TD/A0TERM (A-1)1

Minimum Temperature Lapse Rate - TNLAPS

TNLAPS = TZLAPS - (TZLAPS - TLMN) * TD/A0TERM (A-2)

where TD = daily temperature range (TX-TN)

and TZLAPS = TZ - (PP/PPM) * (TZ - TZP)

for the above,

PP = daily precipitation

TLXM = 10oC/km

TLMN = 0.5oC/km

TZ = 6.4oC/km, that is, a reference lapse rate for rain-free conditions

TZP = 6.4oC/km, that is, a reference lapse rate when PP > PPM

PPM = 5 mm/day

and the calibration parameter A0TERM equals the maximum temperature range
under open sky conditions (selected from the data set (TX-TN)).

Precipitation Elevation Gradients

The enhancement of precipitation as a moist air mass is driven by wind across mountain
barriers is an important aspect of mountain watershed modeling and is referred to as
orographic precipitation.

Orographic precipitation is influenced by three main factors, the slope of the

mountainside, the mountain barrier height and thirdly, the stability of the air mass. To
generalize, the steeper the mountain side, the more rapid the increase in orographic
precipitation. However, this increase in precipitation with elevation does not continue
indefinitely, but tends to decrease substantially at about half of the maximum mountain
barrier height and shows no further increase above about two-thirds of the maximum
mountain height.

1)Some variable names in Appendix A carry different spelling from those

used in the source code for the model.
 Appendix A - 4

The stability of the air mass depends on the relative values of the saturated adiabatic lapse
rate and the prevailing lapse rate of the surrounding air mass. A useful index of stability
can be developed by comparing the values of the saturated and dry adiabatic lapse rates.
The saturated adiabatic lapse rate is dependent on temperatures (Chow) being
considerably less than the dry adiabatic rate when temperatures are high, but increasing to
similar values as the dry adiabatic rates when temperatures are low. This variation of
saturated adiabatic lapse rate will be used to account for the variation in orographic
precipitation influences between winter snowfall, which shows a high orographic
increase, and warm summer rain, which is relatively insensitive to orographic effects.

The algorithms which describe the variation of precipitation with elevation are
subdivided into two aspects to give the final orographic effect. The first algorithm
describes the basic enhancement of precipitation with elevation barrier height if the
temperature is 32oF and the second algorithm modifies this basic distribution for
variations in temperature.

Orographic Enhancement as Function of Elevation, Barrier Height and Temperature

The precipitation in any elevation band is calculated from the precipitation in the band
immediately above or below using the equation

PI,J,L+1 = PI,J,L*(1 + a)  elev/100 (A-3)

where PI,J,L is the precipitation from meteorological station I for day J and
elevation band L.

a = precipitation gradient (%)

elev = difference in elevation between the AES stations and the
band

The 1+a multiplier produces a logarithmic increase in precipitation with elevation. The
enhancement factor a is separately defined above and below a certain elevation which
may be specified (E0LMID).

Three precipitation gradients are possible: P0GRADL, for bands below E0LMID;
P0GRADM, for bands between E0LMID and E0LHI; and P0GRADU, for bands above
E0LHI.

Modification of a With Temperature

The second algorithm modifies a for temperature other than a base elevation
band temperature of 0oC.

a = a - S * Tavg (A-4)

S is the model parameter A0STAB (0.01). The rationale for the use of the
parameter S is derived as follows.
 Appendix A - 5

A measure of the stability of the air mass is (LD - LS)/ LD where LD is the dry adiabatic
lapse rate. At higher air temperatures the greater difference between LD and LS indicates
that warm moist air is much more unstable than moist cold air. Warm moist air is
therefore likely to produce convective precipitation and does not rely so much on
orographic lifting. In contrast, cold moist air has very little difference between LS and
LD, so that convective rain is rather unlikely. Consequently, a cold moist air mass is very
dependent on orographic lifting to produce precipitation. This argument leads to the
defining of an orographic opportunity factor,

Orographic opportunity = 1 - ((LD - LS) / LD)= LS / LD (A-5)

If there is already considerable instability of the air mass, then there is a much reduced
opportunity for precipitation to be enhanced by orographic lifting. The factor LS/LD will
be proportional to the vertical flow rate induced by the orography and it will also be a
measure of the condensation rate per unit volume of air. The reduction in orographic
effect with temperature, S, will therefore depend on (LS/LD)2. A graph of (LS/LD)2
versus temperature, reveals an almost linear variation between -30oC and +20oC.

Graphs in Figures A-3 and A-4 show (LS/LD)2 as a function of temperature and the
impact of A0STAB on the orographic precipitation factors. In Figure A-4 an orographic
factor of 7,000 at 0oC changes to 5,000 at -15oC and 20,000 at 27oC and, because the
orographic effect is inversely dependent on the orographic factor, these values indicate a
much higher orographic effect in winter and much lower effect in summer.

Form of Precipitation

The model must distinguish between precipitation in the form of snow and precipitation
falling as rain and this distinction must be made for each elevation band. Snow is stored
until melted, whereas rain is immediately processed by the soil moisture model.

The form of precipitation is controlled by three logical statements and the temperature, T,
used in these statements is normally the mean daily temperature in each band but it can be
specified to be the maximum or the minimum daily temperature in each band.

If T < 0oC all precipitation is SNOW

If T >A0FORM all precipitation is RAIN.

A0FORM is specified in the parameter list. If it is not specified or if it is set equal to or

below 0oC, it will take a default value of 2oC.

Between 0oC and A0FORM a proportion of the precipitation will be specified as rain,
and this proportion is defined by FORMPP where,

FORMPP = T 0 < FORMPP < 1 (A-6)

A0FORM
 Appendix A - 6

Then rain, RN = PP*FORMPP

and snow, SN = PP* (1-FORMPP)

Precipitation Representation Factors

Each meteorological station has two precipitation representation factors associated with
it, one for snow, P0SREP, and one for rain, P0RREP. These factors are introduced
because precipitation measured at a point is not always representative of the areal
distribution of precipitation. For example, a meteorological station may be in a rain
shadow situation, or it may be in a narrow valley where it is receiving precipitation which
is more representative of the mountain side some hundreds of meters higher than the
station. These representation factors can be determined by comparing long term volumes
of runoff with computed values.

In general, snow measurements are more likely to be distorted by local exposure and
orography. As temperatures rise, rain values tend to be more representative because
warm rain tends to be more convective in origin. When temperatures rise above 18oC,
the local convective nature of the rain makes it more reasonable to introduce an area
reduction factor. The program logic is:

If TX < 0 use P0SREP

If 0 < TX < A0FORM, use linear interpolation between P0SREP (A-7)
and P0RREP.

Precipitation Adjustment Factor

P0PADJ can be used to increase the amount of precipitation in a band as an alternative to

adjusting the amount of precipitation at the station. It is useful sometimes to assist the
model in building glaciers or modeling avalanches. However, this parameter is rarely
used for routine calibration.

A P0PADJ value of 0.5 increases the precipitation by 50%, and a value of 1.0 doubles the
amount of precipitation.

Evaporation

Evaporation estimation can be subdivided into three processes. In the first process,
estimates are made of the daily potential evapo-transpiration for the reference
meteorological station in the watershed (EVAP). In the second process, this EVAP value
is distributed to each elevation mid-band level and is designated by PET. In the third
Process, these PET values are used in conjunction with the calculated soil moisture
 Appendix A - 7

deficit values to yield an actual evapo-transpiration value for each band (AET) and will
be discussed in the soil moisture section.

EVAP = A0EDDF * V0EMOF * TX (A-8)

where
A0EDDF is an evaporation constant = 0.133
V0EMOF is specified as a monthly factor. It accounts for the seasonal
variation of EVAP

PE(L) = EVAP - A0PELA *  elev/1000 (A-9a)

where A0PELA = 0.9 and is the lapse rate for A0EDDF (mm/km)

 elev is the elevation range between the EVAP elevation and the PET elevation
in metres.

Tree cover may also be included in the formulation, as follows:

PET(L) = PE(L) * (A0PEFO * C0TREE + 1.0 * (1.0 - C0TREE)) (A-9b)

where
C0TREE = fraction of tree-covered area.

A0PEFO = Evapotranspiration multiplier in forested areas

The tree covered area can also be modified by a canopy factor, which represents the
density of the tree cover. This canopy factor, C0CANY, multiplies the tree cover factor,
C0TREE.

A.3 THE SOIL MOISTURE MODEL

The daily amounts of snowmelt and rain are subject to various delays and loss before
eventually appearing as river flow at some watershed outflow point. The Watershed
Model contains certain logical statements which decide how these snowmelt and rain
inputs are subdivided between evaporation loss, and fast, medium, slow and very slow
runoff.

The central control parameter for the subdivision of total watershed input is the SOIL
MOISTURE DEFICIT. Note that rather than attempting to specify a total soil moisture
capacity, the model operates from a lack of soil moisture. When this soil moisture deficit
reaches zero, the watershed reaches its maximum runoff potential, except for FLASH
runoff which will be discussed later.

A diagrammatic representation of the soil moisture model is shown in Figure A-5, and is
described below in terms of the priorities of runoff.
 Appendix A - 8

First Priority: Impermeable Percentage - Fast Runoff Control

Part of each elevation band can be specified to be impermeable (C0IMPA), so that any
input of water to this area will enter the fast runoff component. Such runoff can be
thought of in terms of surface runoff or very superficial percolation through coarse
sediments. Usually such areas must be riparian in nature

The impermeable percentage of the watershed can be made to vary with soil moisture
deficit. The algorithm which describes this process is:

Impermeable fraction = C0IMPA * 10 ** (-S0SOIL/P0AGEN) (A-10)

C0IMPA is the maximum impermeable fraction when the soil is fully saturated.
S0SOIL is the soil moisture deficit in an elevation band.
P0AGEN is a constant which regulates how sensitive the impermeable area is to changes
in soil moisture.

Second Priority: Soil Moisture and Actual Evapotranspiration

Before any further runoff can occur, other than fast runoff, the soil moisture deficit must
be satisfied. While soil moisture deficits are being satisfied by incoming water from
snowmelt and rain, there is also an evaporative demand which is continually building up a
deficit.

Potential evapotranspiration has been described earlier. On any given day, in any given
elevation band, there will exist a specified potential evapotranspiration. The soil
moisture deficit which exists in that band will represent the actual evapotranspiration
capability of that band. The algorithm describing this relationship is:

AET = PET * 10 ** (-S0SOIL/P0EGEN) (A-11)

AET is the actual evapotranspiration.

PET is the potential evapotranspiration.
S0SOIL is the current value of the band soil moisture deficit.
P0EGEN is a specified constant which controls the rate at which S0SOIL
influences PET.

This actual evapotranspiration demand will only influence the area of the watershed
which is not impermeable.

For each day a new value of soil moisture deficit is computed:

New value of S0SOIL = S0SOIL-PRN-BM+AET (A-12)

 Appendix A - 9

where
PRN is rain input
BM is snow and glacial melt input.

If the band soil moisture deficit reaches zero, any excess water inputs can be subjected to
further priorities.

Third Priority: Groundwater Percolation (P0PERC)

Groundwater percolation accepts any water excess up to a fixed limit (P0PERC). Any
excess above this limit goes to the fourth priority, medium runoff.

Water which percolates to groundwater is assumed to be divided into specifiable fractions

which go to the two groundwater components, the upper groundwater and the deep zone
groundwater components. This subdivision of groundwater is controlled by P0DZSH, the
deep zone share.

Upper groundwater zone recharge = (1-P0DZSH) * P0PERC

Deep groundwater zone recharge = P0DZSH * P0PERC (A-13)

Fourth Priority: Medium Runoff

Any excess moisture input now remaining, with the exception of glacial melt, is assigned
to medium runoff, or interflow. Any excess glacial melt is re-routed to contribute to the
fast runoff component of glacial melt. Although this is the lowest priority, it is also
frequently the most significant runoff component during active snowmelt and rain storms.

In the model, interflow is considered to be a large reservoir which receives inflows day by
day during active snowmelt and rain. These inflows are the excesses remaining after
satisfying soil moisture and groundwater abstractions. This reservoir releases a certain
fraction each day, but the volume of water released does not immediately appear in the
downstream channel system. Instead, this released water undergoes a convolution very
similar to the fast component unit hydrograph. This release from an interflow storage
reservoir and convolution before reaching the channel outflow point, produces a much
more sluggish response for this medium runoff component.

A.4 WATERSHED ROUTING

Water allocated to each of the components of runoff, namely fast, medium, slow and very
slow components, are subjected to a routing procedure which produces a time distribution
runoff. The routing procedure for each component is based on the same underlying
concept, namely the linear storage reservoir. The fast and medium components of runoff
are subjected to a cascade of reservoirs which is essentially identical to unit hydrograph
 Appendix A - 10

convolution. The lower components of runoff simply use a single linear reservoir, thus
avoiding the necessity to convolute for the final outflow.

Fast Runoff Routing

The Watershed Model calculates unit hydrograph ordinates for convolution of fast runoff.
The calculation is based on a conceptual model of the runoff process developed by Nash
in which inflows are passed through a cascade of linear reservoirs. The resulting outflow
at a time t from a unit impulse of inflow is:

. .
u(t) = K-n t(n-1) e-t/k / (n-1)! (A-14)

K is the linear storage constant for each of the reservoirs in the cascade.
n is the number of linear reservoirs in the cascade (specified by N0FASR, N0FASS,
N0GLAC).
t is the time after the water input has occurred.

The Watershed Model uses the above equation to calculate unit hydrograph ordinates for
each day. A graphical representation of the equation and the daily ordinates are shown in
Figures A-6 and A-7.

Linear reservoir routing of the components of runoff is computed as follows:

First reservoir: Q1 = Q1 + (1/(1+TK)) * (WB - Q1)

... etc.
mth reservoir: Qm = Qm + (1/(1+TK)) * (Qm-1 - Qm)

where
Qm = outflow from the mth linear reservoir
TK = time constant of the linear reservoir (e.g. P0FSTK etc.)
WB = water input to the first linear reservoir.

Medium Runoff Routing

The medium speed runoff, often thought of as an interflow component, undergoes a two
stage routing process. In the first stage the medium runoff for each day enters a linear
storage reservoir with a release constant INTK. The daily release from this reservoir is
therefore a constant percentage of the total storage on any given day. Such a release
represents an exponential decay which is identical to Nash's result with the number of
reservoirs equal to one.

QI = QI + (1/(1+INTK)) * (WI - QI) (A-15)

 Appendix A - 11

where WI is the water input to the interflow reservoir, and

INTK can be either POISTK for snow runoff or POIRTK for rain.

The daily release QI is then subjected to time distribution by convolution with a unit
hydrograph which is similar to the fast unit hydrograph.

Slow Runoff Routing

The slow components of runoff, usually termed groundwater flow or base flow
components, are divided by the soil moisture model logic into a slow, upper-zone
component and a very slow deep-zone component.

Both of these groundwater components are routed using a single linear reservoir which
accumulates each day's inflow and releases a fixed percentage of the total storage each
day.

The upper groundwater zone release fraction is controlled by P0UGTK.

The deep groundwater release is controlled by P0DZTK.

A.5 SNOWMELT BUDGET

The Watershed Model accumulates precipitation falling as snow and then depletes these
snowpacks according to the calculated melt rate. This snow accumulation and depletion
is carried out separately in each area-elevation band. There are two ways in which this
snow budgeting can be carried out, the first being a simple budget at the mid-band
elevation, referred to as a block budget, and the second being a more complex wedge
computation.

The Block Budget

From a computational viewpoint this is a straightforward calculation procedure in which

snow is accumulated and melted as if it were falling at the mid-elevation of each band.

The Wedge Budget

This analysis option recognizes that the final depletion of snow is a gradual recession of
the snowline from the bottom to the top of a band. In this option, there is a smoother
transition from the bottom of the band to the top of the band. It might be commented that
the advantages of this more complex analysis are not as great as might be thought.
However, if snow gradients are high, as can be the situation under some melt conditions,
then the wedge budget does give a better representation of runoff behaviour.

Essentially the wedge budget still uses the block budget method until the snowpack is
depleted to a value less than or equal to S0PATS, a value defining the millimetres of
 Appendix A - 12

water equivalent that the snowpack begins to appear "patchy." This value is usually in
the range of 100 to 400 mm of water equivalent. At this point the wedge analysis takes
over.

The wedge analysis redistributes the snowpack in such a way that there is more snow in
the higher elevations of the band and less in the lower elevations of the band (hence the
term "wedge"). The snowpack is then melted in such a way that less and less band area is
covered by snow. This is more consistent with natural snow melt processes.

A.6 SNOWMELT

General

The UBC Watershed Model uses an energy budget approach as opposed to a degree-day
method to calculate snow melt. This energy method is simplified for use when there is
only temperature data available, but can be used if more detailed radiation, albedo and
wind data are available. From the consideration of various energy exchange processes
represented in the next section, in dense forests, the longwave heat exchange processes
dominate and consequently a temperature index approach is reasonably valid. However,
in open areas, an energy approach is superior because shortwave radiation and snowpack
albedo are very important. Also the convective and condensation melt components are
increased by turbulent mixing induced by wind and rugged topography. Temperature
alone is not an adequate indication of melt because temperature varies with elevation and
with incoming air mass, so that there is a weak and non-linear relationship between
temperature and shortwave radiation. Albedo varies seasonally, but can rise considerably
during periods of new snowfall, so that net solar radiation is a non-linear function of
albedo. Cloud cover can greatly reduce the shortwave radiation and also has a strong
influence on the longwave radiation balance. This raises the question of how well cloud
cover can be estimated, especially its density, and to what extent the reduction in
shortwave radiation is compensated by the increase in net longwave exchange. The
conclusion from our extensive studies appears to be that open area melt is not a simple
linear function of temperature, but relationships are put forward by which temperature
may be used to estimate the various non-linear inputs which control the following
simplified energy method for the calculation of snowmelt.

The energy exchange at the surface of a snowpack is made up of four major components:

1. The shortwave radiation exchange, which consists of incoming solar radiation,

and the reflected outgoing shortwave radiation. This shortwave component
depends on the time of year, the site exposure, cloud cover and snow albedo.

2. The longwave radiation exchange depends on black body radiation from the snow
itself and from clouds and tree cover, and gray body radiation from the overlying
air mass. Under clear skies, the net longwave is outgoing, or negative, unless air
 Appendix A - 13

temperature exceeds about 20oC. Under cloudy conditions and also under tree
cover, net longwave can be positive at temperatures above freezing.

3. Convective heat transfer is produced by turbulent heat exchange between the air
mass immediately above the snowpack. This heat transfer is dependent on both
wind and air temperature and particularly on the stability of the air mass above the
snowpack. A warm air mass above the cold snow surface tends to be stable,
resisting any downward transport of heat to the snowpack, unless there is enough
wind to produce turbulent mixing. This turbulent heat transfer is governed by the
Richardson number, RI, which is a measure of stability. The bulk Richardson
number, defined below, is essentially the air temperature divided by the wind
speed squared. As air temperature increases, and if the wind is only moderate, the
stability can increase to the point where very little convective heat transfer can
occur. Convective heat transfer is therefore self limiting and becomes quite small
at higher air temperatures, unless there is very strong wind.

4. Advective heat transfer, often termed condensation melt, is caused by the

transport of moisture to and from the snowpack. Whether condensation occurs,
cooling the snowpack, depends on the relative vapour pressures of the air and
snow surface. Wind is once again an important factor and so is stability, as was
discussed for convective transport. Advective heat transport can therefore
produce snowmelt if the dew point temperature is above freezing, but, like the
convective heat transport, becomes limited at higher temperatures by the stability
of the warm air mass.

A simple set of equations has been developed which expresses the various snowmelt
components in terms of millimeters of snowmelt per day, either negative or positive.
Various simplifying assumptions have been used to write these equations, but they have
been tested on high quality data sets from the U.S. Corps studies (1954) and are a
considerable improvement on the calculation of snowmelt and glacier melt, especially for
high elevation, open area regions in the mountains.

A.7 SIMPLIFIED ENERGY COMPONENTS

Net Shortwave - Energy Input

Melt = IS (1 - CL) (1 - AL) mm, (A-16)

Where CL is cloud cover, AL is the albedo of the snowpack and IS is the incident solar
radiation, which varies seasonably and with latitude and with atmospheric conditions. At
35o North, IS can be expressed in terms of millimeters of melt equivalent per day instead
of Langleys per day,
 Appendix A - 14

I S   Wbo cos alpha  e Tba1m!d (A – 17)

The albedo and cloud cover reduce the potential melt values of equation (A-17) to the net
values expressed by equation (A-16).

Longwave Radiation

Stefan's law can be expanded in terms of the temperature above freezing, T, so that the
longwave black body radiant energy, IL, can be expressed as a linear function of
temperature plus small higher order terms,

IL = r (273 + T)4 Langleys (A-18)

= r . 2734 (1 + 4T/273 + 6T2/2732 + ...)
= 661 (1 + 0.015T + 0.0001T2 + ...) Langleys/day
= 82(6) (1 + 0.015T + ...) mm/day

Under clear sky conditions in an open, tree free area, the net longwave radiation received
by the melting snowpack is the difference between the black body radiation of the 0oC
snowpack and the incoming gray body radiation from the clear sky. Various estimated
values for the gray body radiation are available, some of which are quoted in the U.S.
Corps Snowmelt Report (1955). All of the equations show a small dependence on
humidity of the atmosphere, such as equations due to Brunt and Angstrom. The simplest
equation, from the Lake Hefner study, is

ILA = rT4 (0.749 + 0.0049 ea) (A-19)

where ILA is the atmospheric longwave radiation, and ea is the vapour pressure in
millibars. The dependence on vapour pressure is so small that it is reasonable to accept
this value,

ILA = 0.757 rT4 (A-20)

The net clear sky incoming longwave radiation, ILN, can therefore be written, using the
linearization of equation A-20,

ILN = 661((0.757 (1+0.015TA) - (1+0.015TS)) Langleys/day (A-21)

where TA is mean air temperature and TS is the snow surface temperature which is zero
for a melting snowpack.

ILN = 7.51 TA - 161 - 9.92 TS Langleys/day (A-22)

In snowmelt equivalent,

ILN = 0.94 TA - 20.1 - 1.24 TS mm/day (A-23)

 Appendix A - 15

For a melting snowpack, TS is zero, and ILN does not become positive until TA exceeds
21.4oC.

Net Longwave Under Cloudy Conditions

Clouds (temperature, TC) act as black bodies so that under 100% cloud cover the net
incoming longwave, ILNC, is

ILNC = 9.92 (TC-TS) Langleys/day (A-24)

= 1.24 (TC-TS) mm/day

The clear and cloudy sky equations are combined into an expression for the net longwave
exchange, ILNT so that for a partial cloud cover fraction, CL and for a melting snowpack
at 0oC,

ILNT = (-20 + 0.94 TA) (1-CL) + 1.24 TC . CL mm/day (A-25)

Convective and Advective Heat Transfer

The U.S. Corps report (1955) and Anderson (1976) present equations for convective and
advective heat transfer. Anderson refers to the question of air mass stability, but does not
incorporate the results into a final relationship. The following equations are an
approximate estimation of the new heat transfers which have been developed from the
earlier work.

Under stable conditions the convective heat transfer, QC, is approximately,

QC = 0.18 (p/101) TA . V mm/day (A-26)

in which p is the atmospheric pressure in kilopascals for the elevation being considered,
TA is the mean temperature and V is the wind speed in kilometers per hour.

Similarly, the advective, or condensation, melt QA under stable conditions is

approximately,

QA = 0.35 (p/101) Td . V mm/day (A-27)

where Td is the dew point temperature, which can be approximated by the minimum air
temperature.

Both the convective and advective melt rates are reduced by a factor RM which is a
function of the bulk Richardson number RI. A linearization of RM, subject to the
limitations given below, is

RM = 1 - 7.7 RI (A-28)
 Appendix A - 16

where

RI = (2g.ZATA)/[(TA + 273)V2A] = 0.095 TA / V2A (A - 29)

where V is the wind speed at a reference height Z.

This linearization is reasonable for RI between +0.12 and -0.1. For positive air
temperatures, the RM factor is not less than zero, and for negative temperatures, RM
increases to about 1.8 and then slowly goes a little higher.

This RM factor is based on idealized laboratory conditions. In steep mountain terrain, our
own studies of snowmelt indicate that stability is greatly reduced by "terrain mixing"
caused by large scale roughness and slope. Consequently RM may increase by as much as
a factor of 2.5 times, which is an additional "terrain mixing" factor.

It should be noted that if either of these two snowmelt components need to be adjusted,
sometimes increasing the value of the maximum wind speed factor, P0VBMX will allow
more convective and advective melt.

Rainmelt

Snowmelt from rain is computed as follows:

RAINMELT = K * TM * RN (A-30)

where K= a constant, RN = rainfall, and TM = mean air temperature.

This formula assumes that the rain falls at the mean air temperature. The K factor
(mm/oC of rain) represents the heat content of the rain.

A.8 APPLICATION OF THE SIMPLIFIED ENERGY SNOWMELT EQUATIONS

There are advantages to be gained from using energy equations for calculating snowmelt.
The physical basis of the equation makes it possible to estimate snowmelt for forested
and open conditions, for clear or cloudy weather, for various slope and aspects of
mountainous watersheds and for changes in elevation. It is also possible to argue the
impacts of changing forest cover or the snowmelt that would be experienced under
extreme and unusual weather sequences.

For most high mountain regions radiation, cloud cover, albedo, wind, cloud temperature
and dew point temperature are not available. As reasonable approximations, it has been
found possible to represent these various factors with temperate based estimates.
 Appendix A - 17

Cloud Cover

The cloud cover has been assumed to be related to the daily temperature range,

(1 - CL) = (TMAX - TMIN) / DR (A-31)

in which CL is the cloud cover fraction, TMAX, TMIN are the daily maximum and
minimum temperatures, DR is the daily temperature range for open sky conditions at a
certain elevation in the watershed.

Wind Estimate

The wind tends to produce a decrease in daily temperature range, so that cloudy
conditions are also windy conditions,

Vb = P0VBMX - (P0VBMX - 1) * TED / 25 (A-32)

where Vb = the wind speed (km/h)

P0VBMX = the maximum wind speed at elevations less than 2000 m (km/h)

TED = the maximum temperature range for that day

However, experience has shown that the wind speed may increase significantly with
elevation. For elevations greater than 1 km, the wind speed is increased by,

C 0 ELEM i
Vb  Vb (A-33)
1000

where C0ELEMi is the band elevation.

Albedo

The albedo is modeled by three relationships. Fresh snow is assumed to settle and change
its albedo from a starting value of 0.95, which then decays by a factor of 0.9 per day until
it reaches a settled value, ALS, of 0.65, as described by Equation (A-34).

ALS(j+1) = 0.9 ALS(j) (A-34)

From then on, the albedo continues to decrease at an exponential decay rate, which is
controlled by the calculated cumulative melt for the season, given by

ALj = ALS . exp [-RM/KL] (A-35)

 Appendix A - 18

where ALS is the settled value, usually 0.65, RM is the cumulative seasonal melt in
millimeters and KL is a constant approximately of the order of the total seasonal melt,
usually taken as 3500 mm.

When new snow occurs, if it is greater than 15 mm, the albedo is assumed to return to a
value of 0.9. The albedo then decays again at the 0.9 rate, specified by Equation (A-34),
but continues down until it reaches the value computed on the day before the snow
occurred, ALj from Equation (A-35). The underlying recession described by Equation
(A-35) then takes over. This process allows a fairly rapid change of albedo from 0.9 to
0.65, down to the calculated ALj value from Equation (A-35), and then a slow recession
as the season advances to albedo values of 0.4 or even 0.3 for very deep and aged
snowpacks. Snow cover on a glacier is treated in the same manner, but the glacier, when
it becomes free of snow, is assumed to have an albedo of 0.3.

These relationships for cloud cover, wind and albedo have been developed using SIHP
data gathered in the Himalayas and also using U.S. Corps of Engineers data gathered in
great detail for the Central Sierra Snow Studies (1952) as discussed by Quick (1987).

The user must supply the following information:

1. the temperature range which controls cloud cover specified by the parameters
A0FOGY and A0SUNY. A0FOGY is the temperature range below which
complete cloud cover exists, and A0SUNY is the range above which the sky is
assumed clear.

This temperature range depends on the elevation of the meteorological station and
decreases as elevation increases. Typical values are specified in the calibration
parameter files for the various watersheds.

2. the percentage of forested area for each band, C0TREE and C0CANY.

3. a general orientation index for the watershed, 0 for north facing or 1 for south
facing, C0RIEN.

Negative Melt Budget

If temperatures have been below freezing for any length of time, snow and glaciers
become deeply frozen and require ripening before melt can occur. These antecedent
conditions are accounted for by keeping a running sum of antecedent negative melt and
by allowing the total to decay by a certain fraction each day, specified by P0CTK, the
antecedent negative melt time constant. This decaying of the negative melt budget
depends primarily on temperatures that have occurred over the previous 10 days or so.
The memory of earlier cold temperature decays away. The ripening of the snowpack or
glacier takes into account thermal input from rainfall, and further accounts for time delays
induced by the water holding capacity of the snowpack. The ripening of the snowpack or
glacier is computed as follows:
 Appendix A - 19

CC = CC + (1/(1 + P0CTK)) * (CCB - CC) (A-36)

where
CCB = TM + KCC*RN
CC = cold content storage.
KCC = latent heat contribution from freezing of rain falling on
snowpack and glacier

A percentage of melt water can be retained by the snowpack and this water holding
capacity of the snowpack is expressed as a simple fraction of the cumulative snowfall.

A.9 RUNOFF FROM HIGH INTENSITY RAIN

There are two different controlling mechanisms for runoff from a watershed system. The
runoff from moderate intensity rain and snowmelt events can be considered to be
controlled by soil moisture levels and this is the normal operation of the UBC Watershed
Model. The runoff from high intensity events is controlled by the rate at which water can
infiltrate into the soil system and these infiltration rates are relatively independent of soil
moisture levels.

For these high intensity rain events, some of the precipitation infiltrates into the soil
system and is subjected to the normal soil moisture budgeting. The fraction of the rain
which does not infiltrate is considered to contribute directly to the fast runoff component
and is referred to as the FLASH SHARE.

It should be noted that even intense snowmelt rates do not appear to be adequate to
produce flash runoff, even though 75 mm of snow may melt in the day. Apparently, this
snowmelt is released to the soil system steadily over the whole day and the rate of release
is below the infiltration capacity. In contrast, rain events of even less than 50 mm in the
day can exhibit flash behaviour. Examinations of rain intensity data suggests that rainfall
events of this magnitude tend to occur in localized periods, or bursts, of high intensity
rain, intensities which can be higher than the soil infiltration rate.

Comparing a rain event in which say, 50 mm of rain falls in the day with one in which
100 mm might fall, it appears that the greater rainfall may represent similar intensities
occurring more frequently and over a greater period of the day.

These are the basic ideas underlying the FLASH SHARE behaviour of the UBC
Watershed Model. The relationships governing FLASH SHARE behaviour are as
follows:

1. The portion of the total daily precipitation which is subjected to the soil moisture
budget system is now

PMXIMP + (1-PMXIMP) * FMR (A-37)

 Appendix A - 20

where

FMR = [log(RNSM/VOFLAS)] / [log(VOFLAX / VOFLAS)] = flash share fraction

where, 0 <= FMR <= 1

PMXIMP = impermeable percentage of the particular

watershed elevation band from equation A-10
RNSM = total daily rainfall + snowmelt + glacial melt
VOFLAS = threshold value of total precipitation for flash runoff
VOFLAX = maximum value of total precipitation, which limits FMR range

2. The portion of the total daily runoff which flashes off is therefore, by difference,

FMR * (1-PMXIMP) * RNSM (A-38)

3. From the fraction of the watershed which is calibrated to be impermeable, the

total daily precipitation input is assumed to go to fast runoff, and this would be
true with or without the FLASH SHARE mechanism.

4. When high intensity rain occurs at the same time as snowmelt, the snowmelt is
added to the rain and subjected to FLASH SHARE.

5. It should be noted that FLASH SHARE is specified by two parameters: V0FLAS

and V0FLAX. The actual flash response is coupled to the normal watershed
calibration by incorporating C0IMPA, the percentage of the watershed which
behaves as if it were impermeable. With the usual gradient of impermeability
which is observed to exist in mountainous watersheds, the run-off from lower
elevations is slower, whereas for higher elevations C0IMPA is higher and
therefore the run off is faster in these regions.
 Appendix B - 1

APPENDIX B: PARAMETER DIRECTORY1

The term parameter is used to describe all watershed modeling inputs found in the .WAT
file. This includes job control inputs (e.g. the start date of a run) and physical details (e.g.
the band area). A description and a default value and range, where applicable, are given
for each parameter to help the user identify its use in the Watershed Model. Please treat
these default values as "suggested" values - they are not absolute; however, some of the
parameters have been pre-calibrated and have been found to work adequately for most
watersheds.

Temperatures are in degrees Celsius. All depth inputs, such as precipitation, initial soil
moisture, snowpack water equivalents, and so on, are expressed in millimeters of water.
All flows are expressed in cubic meters per second, and are average values for the
specified time step.

NAME DESCRIPTION RANGE DEFAULT

Potential evapotranspiration factor (mm/day) 0.2

A0EDDF 0.0 - .5

Temperature range below which the sky is fully

A0FOGY 0.0 - 5.0 3.0
covered by clouds

A0FORM Temperature above which all precipitation is rain 0.0 - 2.0 2.0

A0PEFO Evapotranspiration multiplier in forested areas 1.0

Lapse rate of potential evapotranspiration factor

A0PELA 0.9
A0EDDF (mm / 1000 m)

Threshold precipitation for temperature lapse rate

A0PPTP 5.0
(mm)

A0STAB Precipitation gradient modification factor 0.0 - 1.0 0.01

A0SUNY Temperature range above which the sky is clear 10 - 12 12

A0TERM Maximum temperature range 12 - 25 12

 Appendix B - 2

Lapse rate for minimum temperatures when the

A0TLNH station elevation is greater than 2000 m, oC / 1000 2.0
m

Lapse rate for minimum temperatures when the

A0TLNM 0.5
station elevation is less than 2000 m,oC / 1000 m

Lapse rate for maximum temperatures when the

A0TLXH station elevation is greater than 2000 m, oC / 1000 6.4
m

Lapse rate for maximum temperatures when the

A0TLXM 10
station elevation is less than 2000 m, oC / 1000 m

Temperature lapse rate when the precipitation is

A0TLZP 6.4
more than A0PPTP, oC / 1000 m

Moist adiabatic temperature lapse rate when the

A0TLZZ 6.4
temperature range is zero, oC / 1000 m

Water holding capacity fraction of snow in forested

A0WEHF 0.05
areas

Water holding capacity fraction of snow in open

A0WEHO 0.05
areas

AESNAME$ Name(s) of the AES file(s), <filename>.AES

Current value of albedo in forested areas (displayed

ALBFOR 0.50
in the FORECAST.WAT file), enter initial value

Current value of albedo in open areas (displayed in

ALBOPN 0.5
the FORECAST.WAT file), enter initial value

C0AGLA Glaciated area for the band, km2 0.00

Fraction of glaciated area(s) with south-facing

C0AGOR 0.0 - 1.0 0.0
orientation

C0ALEM Mean band area(s), km2

C0CANY Density of tree cover (canopy) 0.0 - 1.0 0.0

C0ELEM Mean band elevation(s), m

 Appendix B - 3

C0ELPT Elevation of AES station(s), m

C0IMPA Fraction of impermeable area for the band 0.0 - 1.0 0.2

Orientation index for each band, 0 = north, 1 =

C0RIEN1 south
0.0 - 1.0 1.0

C0TREE Fraction of the band with tree cover 0.0 - 1.0 0

Elevation above which the precipitation gradient

E0LHI P0GRADU applies. Set at approx. 2/3 barrier
height,(m)

Elevation above which precipitation gradient

E0LMID P0GRADM applies. Set at approx. 1/2 barrier
height, (m)

F0ERGY Solar radiation factor for forested areas 0.15

F0WIND Ratio of wind in the forest to wind in open areas 0.70

AES station number from which evapotranspiration

I0ESTA 1-5
data is retrieved. Specify for each band.

AES station number from which precipitation data

I0PSTA 1-5
is retrieved. Specify for each band.

AES station number from which temperature data is

I0TSTA 1-5
retrieved. Specify for each band.

Specify the method of precipitation distribution

IGRADP 1
calculation: 1 = algorithm, 2 = interpolation

Date on which the FORECAST.WAT file is to be

J0PARR! YYMMDD
generated

Start date of generation of BAND#.CAL files.

J0RLTA! YYMMDD
Dates must be within J0TIMA! and J0TIMZ!

End date of generation of BAND#.CAL files. Dates

J0RLTZ! YYMMDD
must be within J0TIMA! and J0TIMZ!

J0TIMA! Start date of run YYMMDD

1In versions earlier than 2.5 (May 1994), the 0 = south and 1 = north
 Appendix B - 4

J0TIMZ! End date of run YYMMDD

LAGR Time step lag in rainfall distribution 0 or 1 0

LAGS Time step lag in snowmelt distribution 0 or 1 0

Method to be used to determine temperature lapse

LAPSER 1
rate, 1 = algorithm, 2 = interpolation

N0AESS Number of AES stations 1-5 2

N0BANS Number of elevation bands 1 - 12 7

N0FASR Number of fast runoff reservoirs for rainfall 1-5 2

N0FASS Number of fast runoff reservoirs for snow melt 1-5 2

N0GLAC Number of fast runoff reservoirs for glacial melt 1-5 2

Fast rainfall runoff storage in the N0FASR

O0FARN 0
reservoirs. Calculated: enter initial values only.

Fast snow melt runoff storage in the N0FASS

O0FASN 0
reservoirs. Calculated: enter initial values only.

Fast glacial melt runoff storage in the N0GLAC

O0GLRO 0
reservoirs. Calculated: enter initial values only.

Deep zone storage of groundwater. Calculated:

O0GWDZ enter initial values only. To view calculated values 0
see the FORECAST.WAT file.

Upper zone storage of groundwater. Calculated:

O0GWUZ enter initial values only. To view calculated values 0
see the FORECAST.WAT file.

Interflow component storage for rainfall runoff.

O0INRN 0
Calculated: enter initial values only.

Interflow component storage for snow melt.

O0INSN 0
Calculated: enter initial values only.
 Appendix B - 5

Impermeable area modification factor. It is

compared with how much moisture has satisfied the
P0AGEN 100
soil demands and used in an exponential decay
function. Units of mm

The value which the albedo initially decays to using

P0ALBASE 0.65
the decay factor P0ALBREC

P0ALBMAX Albedo of fresh snow 0.95

P0ALBMIN Albedo of very aged snowpack 0.30

Compared to the total cumulative melt, S0MEOS or

S0MEFS for a season and is used in an exponential
decay function. It is meant to be of the order of the 1000 -
P0ALBMLX 3500
total snow melt for the season. Chosen so that the 4000
albedo at the end of the snowmelt season is
decreased to ~ 0.4. Units of mm.

Snowfall required to return the albedo to that of

P0ALBSNW 15
new snow, mm

Topographical horizon. Used to compensate for

P0BAHT 20o
shading in valleys.

P0BLUE Longwave multiplier for forested areas 0.4 - 1.25 0.4

P0CAST Fraction of solar radiation penetrating cloud cover 0.25

P0CLAR Atmospheric turbidity (clear = 2; 5 = heavy smog) 2

P0COND Condensation melt multiplier 0.35

P0CONV Convective melt multiplier 0.18

P0CTKF Cold content decay factor for snow in forested areas 0.0 - 20.0 10.0

P0CTKO Cold content decay factor for snow in open areas 0.0 - 20.0 10.0

P0CTKG Cold content decay factor for glaciers 0.0 - 30.0 10.0

P0DZSH Deep zone share (lower fraction) of groundwater 0.0 - 1.0 0.5

P0DZTK Deep zone share (lower groundwater) time constant 100 - 300 150
 Appendix B - 6

Used to determine the actual evapotranspiration

from the potential value. Compared with how much
P0EGEN 100
moisture has satisfied the soil demands and used in
an exponential decay function. Units of mm.

P0FRTK Rainfall fast runoff time constant (days) 0.0 - 2.0 1.0

P0FSTK Snow melt fast runoff time constant (days) 0.0 - 2.0 1.0

P0GLTK Glacial melt fast runoff time constant (days) 0.0 - 2.0 1.0

Precipitation gradient factor for elevations below

P0GRADL 0 - 20 5
E0LMID, %

Precipitation gradient factor for elevations below

P0GRADM 0 - 20 0
E0LHI, %

Precipitation gradient factor for elevations above

P0GRADU 0 - 20 0
E0LHI, %

P0IRTK Rainfall interflow component time constant (days) 1.0 - 10.0 5.0

Snow melt interflow component time constant

P0ISTK 1.0 - 10.0 5.0
(days)

Latitude (in degrees) of the middle of the watershed,

P0LATS 50
(+) for north of the equator and (-) for south

P0LWVF Longwave melt multiplier for forested areas 1.9

P0PADJ Precipitation adjustment factor for each band 0.0

Groundwater percolation. (Maximum capacity of

P0PERC sub-surface storage. Excess runoff goes to 0 - 50 15
interflow.) Units of mm.

Fraction of precipitation intercepted in forested

P0PINT 0.12
areas

P0PINX Maximum value of interception, mm 10

P0RREP AES adjustment factor for rainfall data -1.0 to 1.0 0.0

P0SREP AES adjustment factor for snow fall data -1.0 to 1.0 0.0
 Appendix B - 7

Value added to temperature before determining the

P0TASR 0.0
form of precipitation, oC

Lapse rate of maximum temperature range for

P0TEDL 6.0
elevations below 2000 m, oC / 1000 m

Lapse rate of maximum temperature range for

P0TEDU 0.0
elevations above 2000 m, oC / 1000 m

P0UGTK Upper groundwater runoff time constant (days) 10 - 50 30

Maximum wind speed at elevations below 1000 m,

P0VBMX 8.0
km/h

R0GLAC Glacier indicator: 1 = glacier present, 0 = no glacier 0

Indicates whether snow covered area contributes to

R0SNET evapotranspiration: 0 = snow covered areas 0
contribute , non-zero = no contribution

S0CLDF Snowpack cold content storage in forested areas 0.0

S0CLDO Snowpack cold content storage in open areas 0.0

S0CLDGL Glacial cold content storage 0.0

Snowpack below which snow fails to cover the

S0FATS 100 - 400 200
entire band surface for forested areas, mm

Snowpack below which snow fails to cover the

S0PATS 100 - 400 200
entire band surface for open areas, mm

Soil moisture storage deficit. Calculated: enter

S0SOIL initial value only. Units mm. To view the 0.0
calculated value see the FORECAST.WAT file.

Snowpack in forested areas (water equivalent), mm.

S0SWEF Calculated: enter initial value only. To view 0.0
calculated value see the FORECAST.WAT file.

Snowpack in open areas (water equivalent), mm.

S0SWEO Calculated: enter initial value only. To view 0.0
calculated value see the FORECAST.WAT file.
 Appendix B - 8

Snowpack holding storage deficit in forested areas.

S0WEDF 0.0
Calculated: enter initial values only.

Snowpack holding storage deficit in open areas:

S0WEDO 0.0
Calculated: enter initial values only.

TICAES Time increment of AES data: 1 = hourly, 24 = daily 24

TIWSC Time increment of WSC data: 1 = hourly, 24 = daily 24

Temperature below which snow capping may occur

TSNCAP 0-2 2
upper station, oC

TSTEP Time step in hours: 1 = hourly, 24 = daily 24

V0EMF1 Monthly adjustment factor for snowfall 0.0

V0EMF2 Monthly adjustment factor for rainfall 0.0

Monthly adjustment factor for the

V0EMOF 1.0 - 2.0 1.0
evapotranspiration rate, A0EDDF

Flash flood threshold, mm (If runoff greater than

V0FLAS 20 - 40 30
V0FLAS, FLASH SHARE is initiated), mm

V0FLAX Maximum flash runoff, mm 1800

Monthly solar radiation adjustment factor for north

V0NOTH 0.0 - 1.0 1.0
facing bands

Monthly solar radiation adjustment factor for south

V0SOTH 0.0 - 1.0 1.0
facing bands

WSCNAME$ Name of .WSC file. Type VOID for no data.

 Downloading Hydat flow data and converting to
UBC Watershed Model Dos Format.

Insert the Hydat CD and click on Hydat.exe.

Click “display”and request “summary data” . Then search for required station using the
“station lookup”

Click on the required station in the displayed list, and it will be displayed in the lower
window. Click on it and then click the tick symbol, which will cause several more
buttons to appear. (or click the OK button.)

Click on “display” button at bottom of screen and Monthly Summary data will be
displayed. (You can print or save this summary, if you wish)

Click on “export” at top of page, in header task bar and click on “comma delimited”
because there is no other reasonable choice (you can change to “card format” on the next
screen.).

Check the information on the display. Change the report type to “Daily”, then change
Format to “card format”(you have to click the up-arrow to find Card Format!!) and make
sure it is set to download “flow”. Enter the years required. The available years are
displayed near bottom of screen.

Click “Export” button on right of screen. (And not the header task).

Specify the file name and the folder to save in.

Click “Export” button on right of screen. (And not the header task).

A message will tell you that it has been successfully exported.

When you have exported all the files you require, close Hydat.

Re-formatting Hydat data to UBC format:

Find the saved file and rename it to “Name”.wsc, where “name” must not exceed 8
characters.

Open the file in Notebook or WordPad. (It is clearer in WordPad and easier to edit.)

Add a new line to the top of the file as follows:

3 Name of Station
The 3 must be in the first column. The name can contain any information you find
useful! But do not exceed the one line.

Save the file into your Data Folder, (e.g. Basin Name). This folder must also contain the
conversion files, CON_WSC.EXE for the Flow Files, and CON_AES.EXE for the
Climate files
.
Double click on the program “CON_WSC.EXE”

The program will display the available *.WSC files. Select the file you have just saved
and press “ENTER”.

The program may convert the file, or it may tell you that there is missing data. This
missing data information will be saved in a text file to assist you to edit it.

All missing data must be replaced, either with zeros or with data from another year. The
date information must be corrected to run in sequence.
(For example, if winter flows are missing for several months for a station in which the
winter flows are very small, then it is reasonable to replace these months from another
year. But be careful to change the year value in each line to the correct year. Also be
careful about LEAP YEARS, when February must have 29 days of data. Perhaps select
from another leap year, if possible.

Make sure you have not left a blank line at the top of the file, and make sure your inserted
top line is immediately above the data, with no blank line.

Now when you call CON_WSC.EXE again and click on the required file, it should tell
you that conversion is successful.

Conversion to Windows format can now be done using the Convert Standard UBC
DOS to Windows Structure option in Tools menu of UBC watershed model in
windows.
 Steps to run UBC Snow Water Equivalent (SWE) Updating Program

Select a previously calibrated Watershed Wat file, and run this file from the previous
October 1. Review this calculated flow against the available recorded flow. Use this
comparison to decide the date for the first update, based on any systematic diversion of
the recorded and calculated flows.

 The start of the update period can now be specified by selecting the Update
procedure and specifying the start date. Similarly, the end date should also be
entered.

o The updating procedure will automatically generate and name the

Forecast.wat file. The naming convention for the Forecast WAT file uses
the first four letters of the watershed file name, and then adds a four digit
date for the start of the Forecast.wat period. For example,
Camp0501.WAT. This forecast WAT file has been generated on May 1 of
the specified year for the Campbell watershed.

o When a calibrated WAT file is selected, a year sub-folder will be created

in the folder where the calibrated WAT file is located. The forecast WAT
file, together with input data files for a particular year, will be placed in
this year folder. Any additional forecast.wat files run for later dates in this
same year will be placed in this same folder.

Figure 1
 o If the observed flow file is specified in the WAT file, it will be displayed
in the box as the default. Otherwise, the observed flow data can be
selected by clicking the “Browse” button. ( If flow data only exists for
part of the period, then the missing data should be filled in with zeros in
the flow file.)

o The snow pillow data file (*.SNC) is selected by clicking the “Browse”
button. The *.SNC should be created first by running SnowHist.exe
which, in British Columbia, can be downloaded from the River Forecast
and Snow Surveys Center. Otherwise, the *.SNC file may have to be
generated by the user in the same format as the example *.SNC files
provided.

After clicking the “OK” button in Figure 1, another window will display the band
elevation, snow covered area, in percentage, and snow water equivalent (SWE).
The navigation buttons, Figure 2, allow the user to navigate to any dates within
the specified run period.

Figure 2

A graphical view, Figure 3, of band SWE for a particular date is also shown on
screen. The graph shows the daily values of SWE as the user moves along the run
time frames by clicking the “Previous” or “Next” buttons on the table shown in
Figure 2. The snow pillow data will also be shown on the same graph, indicated
by a triangular symbol, if the snow pillow data is available for a particular date.

2
 Figure 3

The overall graphical view of band SWE, mean watershed SWE and the available
snow pillow data for the specified run period are displayed for easy reference in
Figure 4, to assist snowpack updating. The triangular symbols on the graph are
the snow pillow data for a particular snow pillow station. A list of available snow
pillow stations is also displayed for the user’s selection. When the user double-
clicks a selected station, the snow pillow data will be shown on the graph with the
corresponding snow pillow station name.

Figure 4

 The column table, Fiure 2, for band SWE is editable. If the user wants to update
the SWE for a particular band on a particular date, the user just clicks the SWE
field corresponding to the band. After making changes to the SWE, the user has to
click the “Save Update” button to save all the changes on a particular date to a
Forecast WAT file. The Forecast WAT will be named according to the naming
convention discussed previously.

3
 For example, if the band SWE changes are made on May 1, 2004. Then the
Forecast WAT file is named as CAMP0501.WAT and placed in the year folder of
2004.

 If the user does not like the updating after reviewing the hydrograph against the
observed flow data, the user can reset all the band SWE changes to the original
values for a particular date by clicking the “Reset SWE to Original” button.

 When the user clicks the “Forecast” button shown in Figure 2, the user is
prompted to select a forecast WAT file. The default forecast end date is set to 14
days ahead of selected forecast WAT file. The user can extend the forecast period,
as required. The format of the forecast end date is yyyymmdd.

Figure 5

The program will plot the hydrograph with the observed flow, the forecast flow
without any SWE updating, and the forecast flow with SWE updating. The plot
will extend from the specified run start date to the specified forecast end date.

If the user clicks on any point of the hydrograph as shown in Figure 6, the value
of a selected date for a particular graph will be shown at the bottom of the
hydrograph.

4
 Figure 6

 The updating procedure can be a repetitive process. The user may update the band
SWE several times, and then review the resulting hydrograph after each update.
The user can continue the process until a satisfactory result is achieved.

 For reference and to record the whole process, the program saves all the SWE
updated information to a history list file which has an extension of LST, and is
stored in the same folder as the forecast WAT file.

5
----- Original Message -----
From: edmond yu
To: Michael Quick
Sent: Wednesday, August 11, 2010 9:57 PM
Subject: Last email
I couldn't send executable.
Please rename setup.eee to setup.exe.
I have already sent 10 cab files in the previous emails.
Please let me know if you have any problems to get the files.
Thanks.
Edmond
Xu,and Naeem,

I am sending a series of emails with two CAB files in each, and these extra setup files.
Let me know if you receive them safely.
There will be one more email with some documentation, manual etc.
Put all the CAB files and setup files into one new folder and then click on "setup exe".
It should install. You may get some messages about newer system files, and just say "Yes" to these.
Windows 7 may report some errors, but just ignore and it should install correctly. An extra email covers
this.
There are 10 emails in total.
Let me know if you have any problems.
Michael Quick
4598 West 1st Avenue
Vancouver, BC, V6R 1H8
604-224-8895
----- Original Message -----
From: edmond yu
To: Michael Quick
Sent: Thursday, August 12, 2010 9:00 AM
Subject: UBCWM XP package
Dear Dr. Quick,
Have you got any problems to download the whole package?
The whole XP package should include 10 cab files. The" setup.exe "must be renamed from setup.eee,
and 1 setup.lstmust be included.
I am working to output the soil deficit.
When will you be back from your trip? I would like to discuss when you are back.
Thanks.
Edmond
his relates to Windows 7.
----- Original Message -----
From: edmond yu
To: Michael Quick
Sent: Wednesday, August 11, 2010 8:18 PM
Subject: Install UBC WindowsXP package on Windows7
Dear Dr.Quick,
When I installed the UBC Windows XP package on Windows7 operating system, I got several error
messages as follows:
- Entry Point Not found, RegEnumKeyExI could not be located in the dynamic link library MSDART.DLL
- An error occured while registering the file msadox.dll.
I do not understand exactly what these messages meant. I just ignored all these message, and finished
the installation.
It seemed to work. I could run the MIca watershed and plot the flows in graphic.
Edmond