Remote sensing
& global databases
Claire Michailovsky
March 2nd 2020
Hydrological data collection & sharing
• How easy is it to collect traditional data?
• How easy is it to obtain traditional data?
• For what reasons?
A few examples …
Practical Political
remote areas classified data
data distribution/formats/storage ownership/data is power
cost of maintaining stations/collecting International/national conflicts
data Competeing users
lack of information on how the data is [Link]. is reservoir managed by
collected/validated (ex. a rain gauge energy or irrigation department?
with vegetation growing in/over it) Will they share data?
Example: Global Runoff Data Center
How to avoid these issues?
• Use of open access remote sensing data and global datasets
– Advantage for us
- data is available in a predictable manner
– Advantage for stakeholders
- accounts are reproducible, based on open source code and data
Transparency
Remote sensing
- a short introduction
What is remote sensing?
• “the science—and art—of
identifying, observing, and
measuring an object without
coming into direct contact with
it”
([Link]
[Link]
[Link]
What is remote sensing?
• “the science—and art—of
identifying, observing, and
measuring an object without
coming into direct contact with
it”
([Link]
• In the context of hydrology or more
generally earth sciences: [Link]
– The “object” measured is the
Earth
– Observation is typically done
from planes or satellites
[Link]
Characteristics
• Spatial resolution
Characteristics
• Coverage
• Temporal resolution or revisit time
– Depend on the satellite orbit:
Geostationary Satellites (GOES, Meteosat, …)
Good temporal resolution
Polar Orbiting Satellites (TRMM, Sentinel, …)
Good spatial coverage
How does it work?
• Passive • Active
– Sun is the source of energy – Satellite illuminates area to be
reflected or emitted by the measured with own energy
Earth source
Sun Satellite
2 main sources of EM radiation
1. Sun
Visible (VIS)
Near-infrared (NIR)
Shortwave-infrared (MIR)
“Shortwave radiation” (0.1 – 3 m)
2. Earth
Thermal-infrared: 8 – 14 m
Microwave radiation: 1- 20 cm
“Long wave radiation” (5 – 50 m)
How does it work?
• Sensors measure different parts of the electromagnetic spectrum
Example: Landsat bands
• Landsat (NASA/USGS) has data dating back 30 years
• 30m spatial resolution
• 16 day revisit time
– Data at: [Link] or [Link]
(need free registration to download)
• Data is collected for a range of spectral bands
BLUE
Bands?
GREEN
RED
What can you see with other bands?
NDVI
Example: NDVI
B G R NIR
Landsat 7 channels: 1 2 3 4
Landsat 8 channels: 2 3 4 5
Application: Multi-spectral land use mapping
Application: Multi-spectral multi-temporal land use mapping
1.0
Irrigated mixed
Irrigated and rainfed cropland cotton,wheat
rotation/orchards
0.9 Irrigated mixed
cotton,wheat
rotation/sugarcane
0.8 Irrigated rice, wheat
rotation
0.7 Irrigated mixed
rice,wheat
rotation/cotton
Irrigated wheat,fodder
0.6 rotation
NDVI Value
Irrigated rice,fodder
0.5 rotation
Irrigated mixed
0.4 rice,wheat
rotation/sugarcane
Rainfed crops
wheat/grams
0.3
Rainfed crops mixed
cotton,wheat
0.2 rotation/fodder
Rainfed crops general
0.1
Rainfed crops and
woods
0.0
11-May
21-May
1-May
11-Mar
21-Mar
1-Mar
1-Jan
1-Jun
11-Feb
21-Feb
1-Nov
1-Dec
1-Feb
11-Aug
21-Aug
11-Sep
21-Sep
1-Oct
11-Jan
21-Jan
11-Apr
21-Apr
11-Jun
21-Jun
1-Aug
1-Sep
1-Apr
11-Nov
21-Nov
11-Dec
21-Dec
11-Oct
21-Oct
1-Jul
11-Jul
21-Jul
Date
Thermal Infra-Red (TIR)
• Unlike the previous examples with
reflected radation, TIR measures emitted
radiation
Example: Iceland – geothermal plant
Land surface temperature from Landsat
Application: ET monitoring
Surface energy balance
Application: ET monitoring - SEBAL
Weather station data
NDVI, albedo, cloud
Competitors
cover
surface temperature,
Vegetation
Index
ET
Biomass (yield)
Rn G c p
es ea
ra
ET
r
Bio APARt t dt 1 s 24
ra
Bastiaanssen et al., 1998
Some limitations…
Image: ESA
Passive Microwave remote sensing
• Concept similar to thermal
• Lower energy
– Need to observe larger areas to
get useable signal = low spatial
resolution
• Applications:
– Soil Moisture
- SMOS
- SMAP
– Precipitation
Passive Microwave remote sensing
• Concept similar to thermal
• Lower energy
– Need to observe larger areas to
get useable signal = low spatial
resolution
• Applications:
– Soil Moisture
- SMOS
- SMAP
– Precipitation
Active Microwave remote sensing - RADAR
• Radio Detection and Ranging
• Emission of Microwave pulse and
analysis of the portion of the signal
returned to the sateillte
Different scattering mechanisms & wavelengths
for different applications
• Forest monitoring
Source: Bourgeau-Chavez, Laura L.A1 - Riordan, KevinA1 - Powell, Richard
B.A1 - Miller, NicoleA1 - Nowels, MitchY1 - 2009-10-01PY - 2009T1 -
Improving Wetland Characterization with Multi-Sensor, Multi-Temporal SAR
and Optical/Infrared Data FusionN2 - A2 -
Radar – Flood mapping
CCRS
UN Spider
Radar Altimetry
• Active microwave remote sensing
• Measurement of surface water
elevation
• Lake and river level monitoring
Exception to the rule: Gravity RS!
• Gravity Recovery and Climate
Experiment (GRACE)
Exception to the rule: Gravity RS!
• Gravity Recovery and Climate
Experiment (GRACE)
Advantages/disadvantages
Remote sensing vs. In-situ data
• In-situ data
– High temporal resolution
– Low spatial coverage
– Data access
– Data gaps
• Satellite remote sensing
– Near-global coverage
– Spatially distributed data
– Many open access databases
- all NASA and ESA missions provide free data access regardless of
use intended and country of origin
– Complexity of interpretation
Database examples
& value added products
Major databases used for hydrology
• Precipitation
• Evapotranspiration
• Meteorological
• Land use/Land cover
• Soil moisture
• River flow
• Water Levels
• Groundwater storage
• Surface water extent
• Vegetation
• Elevation
* Data Discovery Portals
Precipitation: TRMM
• Tropical Rainfall Measuring Mission (TRMM)
• NASA & JAXA mission
• Spatial resolution: 0.25° ~27 km
• Temporal resolution: 3-hourly, daily, monthly
• From 1998 to April 2015
• Instruments
– Precipitation radar
– Microwave imager
– Visible and Infrared Scanner (VIRS)
– Lightning Image Sensor (LIS)
Source: [Link]
Precipitation: GPM
• Global Precipitation Measurement
(GPM)
• NASA & JAXA mission
• Spatial resolution: 0.1° ~11 km
• Temporal resolution: 30min
• Initial date: March 2015
• Instruments
– Dual-Frequency Precipitation radar
(DPR) x2
– Microwave imager
Source: [Link]
Precipitation: CHIRPS
• Climate Hazards Group InfraRed
Precipitation with Station data (CHIRPS)
• Funded by USGS and USAID
• Spatial Extent: -180, 180, -50, 50°
• Temporal resolution: daily, monthly
• Initial date: 981
Source: [Link] and [Link]
Evapotranspiration: MOD16
• Moderate Resolution Imaging
Spectroradiometer (MODIS)
• Two satellites / 36 bands
• Spatial Resolution: 1km
• Temporal resolution:
8-day, monthly
• Initial date: 2000
Source: [Link] and [Link]
Evapotranspiration: GLEAM
• Global Land Evaporation Amsterdam
Model (GLEAM)
• VU - ESA
• Global, 0.25° ~27km
• Temporal resolution: daily, monthly
• From 2003 to 2012
Source: [Link]
Meteorological: ECMWF ERA-interim
• European Center for Medium- Variables
range Weather Forecasts • Temperature
(ECMWF) – Surface
• Data Assimilation System – Soil
– Model – Air
– Observations • Albedo
• Spatial resolution: 80km • Surface Pressure
• Temporal resolution: 3-hourly • Cloud cover
• From 1900 to 2010 • LAI
• Etc.
Source: [Link]
Meteorological: GLDAS
• Global Land Data
Assimilation System
(GLDAS)
• Models: VIC and Noah
• Spatial resolution:
1° ~110km
• Temporal resolution:
3-hourly, monthly
• Initial date:1948
Source: [Link] and [Link]
Land Use / Land Cover: GlobCover
• ESA – land cover
• Spatial resolution 300m
• Two periods
– December 2004 – June 2006
– January – December 2009
• Meris mission
– Land surface
reflectance
– Classification
Source: [Link]
Land Use / Land Cover: Global Cropland Extent
• Global Agriculture Monitoring Program (GLAM)
• MODIS
• Spatial resolution: 250m data
• Period: 2000 – 2008
• Pixel classification
• Cropland probability per-pixel
• Cropland/non-cropland
indicator map
Source: [Link]
Land Use / Land Cover: IWMI
• International Water
Management Institute (IWMI)
• Agricultural Areas
– Irrigated
– Rainfed
• Spatial extent: Africa and Asia
• Period: 2000
• Spatial resolution: 250m
• Based on MODIS NDVI data
Source: [Link]
1.0
Irrigated mixed
Irrigated and rainfed cropland cotton,wheat
rotation/orchards
0.9 Irrigated mixed
cotton,wheat
rotation/sugarcane
0.8 Irrigated rice, wheat
rotation
0.7 Irrigated mixed
rice,wheat
rotation/cotton
Irrigated wheat,fodder
0.6 rotation
NDVI Value
Irrigated rice,fodder
0.5 rotation
Irrigated mixed
0.4 rice,wheat
rotation/sugarcane
Rainfed crops
wheat/grams
0.3
Rainfed crops mixed
cotton,wheat
0.2 rotation/fodder
Rainfed crops general
0.1
Rainfed crops and
woods
0.0
11-May
21-May
1-May
11-Mar
21-Mar
1-Mar
1-Jan
1-Jun
11-Feb
21-Feb
1-Nov
1-Dec
1-Feb
11-Aug
21-Aug
11-Sep
21-Sep
1-Oct
11-Jan
21-Jan
11-Apr
21-Apr
11-Jun
21-Jun
1-Aug
1-Sep
1-Apr
11-Nov
21-Nov
11-Dec
21-Dec
11-Oct
21-Oct
1-Jul
11-Jul
21-Jul
Date
Land Use / Land Cover: FAO global irrigated areas
• Map of irrigated areas
• Spatial resolution 5’ ~9km
• % or Ha per pixel
– Equipped
– Groundwater
– Surface water
Source: [Link]
Soil Moisture: JPL - SMAP
• Jet Propulsion Lab (JPL)
• Soil Moisture Active Passive
(SMAP)
• Launched 2015
• Spatial resolution: 9km
• Soil Moisture
– Surface
– Root zone
Source: [Link]
Soil Moisture: ASCAT
• Advanced SCATterometer (ASCAT)
• EUMETSAT
• Spatial resolution: 12.5km
• Temporal resolution: daily
• Initial date: 2007
• SWI (Soil Water Index)
at different depths
Source: [Link] and
[Link]
Water Levels: Radar Altimetry
• Multiple satellites & databases:
• Database for Hydrological Time Series of
Inland Waters (DAHITI)
• Global Reservoir and Lake Monitor
• Hydroweb
• …
Many databases combine the data from
multiple satellites (Topex, Jason 1-3, ERS2,
Envisat, Saral-Altika, Cryosat)
• Spatial extent: will depend on satellite.
Need to check if your target of interest
is on groundtrack.
• Lake & river level variations
Source:
Groundwater Storage: GRACE
• Gravity Recovery and
Climate Experiment
(GRACE) - NASA-JPL
• Measures changes in gravity
• Time resolution: ~ monthly
• Spatial resolution: ~3° [Link]
(~300km)
• ! Some solutions provide 1 °
resolution but with high
correlation to neighboring
cells
• Initial date: April 2002
• Equivalent water depth
• Groundwater
• Root zone
• Surface
Source:
Water Surface Extent/Change
• Global Surface Water
Explorer (JRC)
• Surface Water maps based
on 30-years of Landsat
imagery.
• Occurrence, change,
seasonality, recurrence,
transitions and maximum
extent available
• Landsat Resolution :
• 30 m
• 16 day repeat
• Other Landsat uses in WA+:
• NDVI for vegetation
growth/parameters
[Link]
Vegetation: MOD15
• Moderate Resolution Imaging
Spectroradiometer (MODIS)
• Two satellites / 36 bands
• Spatial Resolution: 1km
• Temporal resolution:
8-day, monthly
• Initial date: 2000
Variables:
• NDVI
• LAI
• Fpar
Source: [Link]
Vegetation: Vito – Proba-V
• Flemish Institute of
Technology (VITO)
• Spatial resolution: 100m
• Spatial extent: -180, 180,
-56, 75°
• Temporal resolution: daily
satellite passes
• Initial date: 16-10-2013
Source: [Link]
Elevation: SRTM
• Shuttle Radar Topography
Mission (SRTM)
• NASA – JPL
• Spatial resolution: 90m – 30m
• Spatial extent: -180, 180,
-60, 60°
Digital Elevation Model
(DEM)
Source: [Link]
Elevation: HydroSHEDS
• Hydrological data and maps based on Shuttle Elevation Derivatives at
multiple Scales (HydroSHEDS)
• Based on SRTM
• Spatial resolution: 90m
• Conditioned DEMs
• Flow direction and
Flow accumulation
grids.
Source: [Link]
Some Data Discovery Portals
• EUMETSAT
[Link]
• NASA
[Link]
• GEO (Group on Earth Observations)
[Link]
• Hydroshare
[Link]
