Invited Paper

Sensor Performance and Atmospheric Effects using

NvThermIP/NV-IPM and PcModWin/MODTRAN® models – a
historical perspective

Authors:

Dylan Payne and John Schroeder

Ontar Corporation
9 Village Way
North Andover, MA

www.ontar.com
[email protected], [email protected]

978-689-9622

ABSTRACT

This paper will focus on the effect of atmospheric conditions on EO sensor performance using computer models.
We have shown the importance of accurately modeling atmospheric effects for predicting the performance of an EO
sensor. A simple example has demonstrated how real conditions for several US sites will significantly change the
Probability of Detection, and hence Recognition and Identification.

The current state-of-the-art model for computing atmospheric transmission and radiance is, MODTRAN® 5,
developed by the US Air Force Research Laboratory and Spectral Science, Inc. Research by the US Air Force,
Navy and Army resulted in the public release of LOWTRAN 2 in the early 1970’s. Subsequent releases of
LOWTRAN and MODTRAN® have continued until the present.

The corresponding state-of-the art in sensor models, is the NVESD’s NVThermIP and soon to be released, NV-IPM.
These models have also undergone an evolutionary process starting with the FLIR and ACQUIRE models to the
present. NVThermIP, versions 2002 through 2009. NVTherm-IP has the capability to interface with 3rd party
MODTRAN® software, along with a Beers law and table inputs. The replacement, NV-IPM, uses a different
approach. It continues to use Beers law and table, however, its primary interface is via a library of pre-calculations.
The capability to interface to a complete set of MODTRAN® inputs no longer exists.

It is recommended that the full capabilities of MODTRAN® be used when modeling atmospheric transmission and
radiance, and that this capability be added to the new NVESD NV-IPM.

Infrared Imaging Systems: Design, Analysis, Modeling, and Testing XXIV, edited by
Gerald C. Holst, Keith A. Krapels, Proc. of SPIE Vol. 8706, 87060G · © 2013 SPIE
CCC code: 0277-786X/13/$18 · doi: 10.1117/12.2016101

Proc. of SPIE Vol. 8706 87060G-1

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 1. INTRODUCTION

Atmospheric effects can have a significant impact on the target detection, recognition and identification capabilities
of electro-optical sensors. Adverse atmospheric and weather conditions can render the most sophisticated hardware
system virtually blind.

This paper will demonstrate the importance of accurately modeling atmospheric effects and their impact on the
performance of electro-optical sensors. Calculations using two state-of-the art models, NvThermIP/NV-IPM and
MODTRAN®/PcModWin, are given, which show how to accurately model EO sensor and atmospheric conditions
and the interaction between them.

Ontar and NVESD entered into a Technology Transfer Agreement to interface MODTRAN® and PcModWin with
NVTherm-IP in the late 1990’s. This resulted in the Atmosphere input screen options in NVThermIP and the Ontar
developed interfaces with PcModWin called PcNvMod and NvMod.

The work presented here used NvThermIP 2009, the most recent version released by the NVESD. NVESD is in the
process of releasing a significant update called NV-IPM. This is discussed in more detail below in Section 4. A beta
version of NV-IPM is available; however it does not allow the user to fully utilize the capabilities of MODTRAN®
that are needed to make accurate atmospheric attenuation calculations. The conclusion given in this paper applies to
NVThermIP 2009 and earlier versions of NVThermIP, as well as NV-IPM.

2. DESCRIPTION OF NVTHERMIP/NV-IPM AND PCMODWIN/MODTRAN® AND

PCNVMOD

“NVThermIP is a PC based computer program which models parallel scan, serial scan, and staring thermal imagers
that operate in the mid and far infrared spectral bands. The model can only be used for thermal imagers, which sense
Figure 1
emitted infrared radiation. NVThermIP predicts the
Contrast Threshold Function (CTF) of the system, and uses
the CTF to predict the target acquisition range performance
Test field likely to be achieved using the sensor. The model also
performance Predict field predicts a Laboratory Measurement of the Minimum
erformance Resolvable Temperature Difference (MRT) of the system.”

“NVThermIP is the latest iteration of the NVESD (Night

Vision and Electronic Sensors Directorate) thermal
models. NVESD was previously called the Night Vision
Laboratory (NVL). The first NVESD thermal model
Laboratory MRT "Night Vision Laboratory Static Performance Model for
"Quick Check" Thermal Viewing Systems," was published by Ratches and
others in 1975. Later versions of the thermal model
included Forward Looking Infrared 90 (FLIR90) and
FLIR92. The original 1975 performance model included both a prediction of sensor Minimum Resolvable
Temperature Difference (MRTD or just MRT) and also prediction of target acquisition performance using the
Johnson criteria. FLIR90 and FLIR92 predict only the sensor MRT. The MRT is then used as an input to the
Acquire computer model which uses the Johnson criteria to predict target acquisition. NVTherm 2002 returned to
the original format by including range prediction and MRT prediction in the same computer program. NVThermIP
includes range prediction based on the imagers calculated CTF (Contrast Threshold Function) and the Targeting
Task Performance (TTP), Metric and a MRT prediction based on Laboratory Procedures. NVThermIP replaces the
2002 model because the use of the Johnson Metric in performance predictions leads to pessimistic results when
modern thermal imagers are modeled. NVThermIP also makes a better attempt at trying to accurately predict MRT
measurements taken during laboratory testing of the system.” (1)

Proc. of SPIE Vol. 8706 87060G-2

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 MODTRAN® is an atmospheric irradiative transfer model, co-developed by the US Air Force Research Laboratory,
and more recently by AFRL and Spectral Sciences Inc. The atmosphere is modeled as stratified horizontal layers
and molecular and particulate constituent profiles may be defined either using built-in models or by user-specified
vertical profiles. MODTRAN®5 performs accurate and speedy calculations from the UV through the visible,
infrared and microwave spectrum (0 to 50,000 cm-1). Calculations in the visible and ultraviolet spectral regions (>
22600 cm-1) are performed at lower spectral resolution (20 cm-1), while those in the IR and longer wavelengths are
done at 0.1 cm-1 resolution. MODTRAN®5 is distributed as a large (> 35000 lines) FORTRAN program oriented
around command line operation, with inputs taken from numerous formatted input files. Operating MODTRAN®5
directly requires extensive knowledge of the file formats and input parameters for the code. (2)
http://www.ontar.com/Software/ProductDetails.aspx?item=PcModWin

PcModWin©/MODTRAN® 5 is a commercial Windows version of the United States Air Force Research
Laboratory's MODTRAN® model. MODTRAN® is the recognized standard for computing atmospheric
transmission and radiance at medium to high spectral resolution (0.1 cm-1). PcModWin wraps around MODTRAN®
for ease of use. (3)

Figure 2 NVThermlP
File Edit Inputs) Run Save/View Results Help PcNvMod: allows the user to run a full
Ia®Il Type Of Imager version of PcModWin - MODTRAN®,
System Parameters

Atmosphere
Input F
Optics
from within NVThermIP with complete
-Atmospheric Transmission Type o Detector
Ch
access to all the MODTRAN® options (4)
Electronics
r Beer's Law Night Display & Human Vision r:
.........................

F ;ModTran;
.
Sensor Alli
Therm
Range Custom MTF
rn The atmospheric conditions are specified
Month
r Table Help
Atmosphere 0:
by selecting the Atmosphere input screen
aaaaa Target
-Table Beer's Law
Additional Options
shown to the left (Figure 2).
sense
Number of Values 0 Transmiss File

Range(Km) I Transmission
ModTran The user has three choices: Beer’s Law,
Aerosol Model
INo Aerosol Attenuation
MODTRAN® and Table. A Technology
Transfer Agreement between NVESD and
Model Environment
IUS Standard 1976 Ontar specified this screen and implement
Edit Values Rain /Cloud Aerosol Extensions the MODTRAN® options and interface to
INo Clouds or Rain
PcModWin. Selecting the MODTRAN®
Index Structure Parameter (Turbulence) -Smoke
2 -2t3
option allows the user to specify three
1
CN ( meters ) 11.00E-14 (T Yes Alpha
I

MODTRAN® parameters: Aerosol

r No Concentration Length 1
Model, Model Environment and the
Rain/Cloud Aerosol Extensions.
411rI
OK I CANCEL
I +I
When NVThermIP is run, there are two
Press F1 for Help. 2:02PM 3/14/2013
paths the software can take depending on
what additional software is installed.

1. Ontar’s PcNvMod is NOT installed: NVThermIP will use the spectral interval and geometry specified
elsewhere on the other inputs screens. Together with the three parameters above and default to running
MODTRAN® 4 v1r1. It is the user’s responsibility to obtain and copy Mod4v1r1.exe to the files directory.
Please Note: The Air Force Research Laboratory no longer supports MODTRAN® 4 and highly
recommends the use of MODTRAN® 5.
2. Ontar’s PcNvMod IS installed: The screen shown below (Figure 3) will be displayed.

Proc. of SPIE Vol. 8706 87060G-3

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 Figure 3

Ontar P<NvMod 5.1 The user is given two options, the Express and Advanced
Choose Mode modes. The Express mode is similar to the option above in
Please choose mode:
that the user cannot specify any additional inputs,
e Express
utomatically run NVMod with the values setup in NVThermIP. This requires no further input on your part. HOWEVER the calculations will be done using
o Advanced MODTRAN® 5 (NOT MODTRAN® 4 as above).
Launch Ontar's PcoydWin to edit the MODTRAN input values. This allows you to modify the parameters
using a rich graphical user interface.

The advanced stetting is the mode of choice. This will bring

Run Keep Window Open: 'Always J
up the screen shown in Figure 4 below and the user has
- Status access to all the capabilities of PcModWin/MODTRAN® 5.

Reading Config File C:IProgram Files (x86)INVESDIModels 20091NVThermIP-

20091Fi IeslNvMod.cfg

About... I Manual... I I Close

Figure 4
PcNvMod report for SPIE 2013

Observer Height: 0.00000 km Bandpass (cm " -1): 833.33 to 1307.2 cm"-1
Length: 10.00000 km Bandpass (um): 12.000 to 7.6500 um
Slant Type:
Atmosphere: Meteorological Data (Horizontal Path)
Extinction: Rural -VIS = 23km

Horizontal Path St. Louis SPIE 2013

Trans. Total
Trans. H2O (Band Model)
Trans. Water Continuum

0.0 2.0 4.0 e.0 8.0 10.0

Down Range Distance Km

0.0
7.60 8.00 8.60 9.00 9.50 10.50 11.60 12.60

Wavelength Microns

The importance of using the full MODTRAN® capabilities is the emphasis of the remainder of this paper.

3. IMPORTANCE OF COMPLETE ATMOSPHERIC MODELING

The significance of atmospheric attenuation can be shown by using the Beer’s Law option in NVTherm-IP – please
refer to Figure 3 above. In general Beer’s Law is not an accurate way to compute transmission, and should only be
used when the user does not have better data. However it will suffice in this case to illustrate the importance of
modeling atmospheric effects.

Proc. of SPIE Vol. 8706 87060G-4

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 The NVTherm-IP distribution is provided with 13 test cases. They are setup for different sensor types, spectral
intervals; scenarios etc. and serve as an excellent tutorial to sensor modeling. NVTherm calculates sensor properties
including the Probabilities of Detection [P(det)], Recognition and Identification ([P(rec)] [P(id)) as a function of
range to the target.

This paper will only report P(det)s to save space. Similar results and conclusions are valid for the Probabilities of
Recognition and Identification.
Table 1

Sensor 2 Beer's Law Atmospheric Transmission

Table 1 to the left, and Figure 5 below, give the
8 -12 µm Spectran Region Sensor 2 [P(det)] using Beer’s Law for an average
P(det) T (ave) = T (ave) = T (ave) = T (ave) = transmission over 10 Kms of 100%, 50%, 25% and
Range 100% 50% 25% 10% 10%. This was done by setting the Beer’s Law
C 1 1 1 1 Transmission per Kilometer to 1.0, 9331, 0.8705,
1 1 1 1 _1 0.795 respectively. As is shown the P(det) at 10 Km
1.000 1.000 varies from 85% at an average transmission of
2 1.000 1.000
100% to 32% at a transmission of 10%. That is, the
3 0.999 0.998 0.996 0.991
sensor performs well on a clear day and poorly
4 0.994 0.987 0.976 0.952 under adverse conditions. Another way to look at it
5 0.982 0.963 0.934 0.874 is the sensor performs well at 10 Km with okay
6 0.963 0.926 0.870 0.766 condition (e.g. 50% transmission), but only out to 6-
7 0.938 0.878 0.792 0.644 7 Kms on adverse condition days.
8 0.909 0.824 0.706 0.523
9 0.878 0.765 0.620 0.412
10 0.845 0.706 0.537 0.315

Figure 5
Next, a series of calculations were
Sensor 2: 8 -12 µm done for 5 different locations in the
United States to demonstrate the
effects of using actual atmospheric
5 Km Range conditions. The results are
summarized in Table2 and Figure 6.
The table is divided into three parts.
The top third gives the average
temperature (oC), relative humidity
(%), and transmission (%) for Tampa,
Saint Louis, San Diego, Boston and
0.2 Seattle for winter conditions, i.e.
February. The transmissions are for
o bands from 3.4 to 5 µm and 7.65 to 12
20 40 60 80 10 µm, and were calculated using
T. (ave) PcModWin/MODTRAN®5. The
average transmissions are taken from

Proc. of SPIE Vol. 8706 87060G-5

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 Table 2

Average Temperature, Relative Humidity and Transmission the bottom of the tape6/Modout1
PcModWin /MODTRAN® 51 1
files. They were done using the
Meteorological Data (Horizontal Path
Only) for the Model Atmosphere and
February Ave Ave 't (ave) °. o T (are) using the New Model Atmosphere
Temp °C RH % 3.4 - 5pm 7.65 -12µuí and User Supplied Profile inputs.
Tampa These options are only available
17.06 73 31 17
when the user has installed the
St. Louis 1.89 72 39 48 PcNvMod software. The middle third
San Diego 14.94 66 33 25 gives the same information for
Boston -0.28 62 41 54 summer conditions, i.e. August.
Seattle 6.28 75 37 39
At the bottom of the table is the
average transmission using the
default atmosphere in MODTRAN®,
e.g. US Standard, Midlatitude etc.
August Ave Ave 't (ave) °.. ó T (ave) %y The default atmospheres are
Temp °C RH % 3.4 - Sµm 7.65 - 12µm statistical averages over many years
and latitude – longitude positions.
Tampa 28.17 78 23 1.9
Some are good, some not so. For
St. Louis 25.67 70 27 5.4 example, Saint Louis in February is
San Diego 22.50 74 28 7.8 somewhere between a US Standard
Boston 22.39 71 28 9.1 and Midlatitude Winter. However,
Seattle 18.67 68 31 16.7 Tampa in February is more like a
Subarctic Summer. The point is that
the default atmospheres are
Default Atmospheres Ave T (ave) °,-oT (ave) % acceptable for bounding the problem,
T /RH 3.4 - 5µm 7.65 - 12µm but are not good for a specific
US Standard 16/46 36 38 scenario.
Midlatitude Summer 21/76 28 8.9
Midlatitude Winter -1 In 40 51
Suba rtic Summer 14/75 32 22
Subartic Winter -16/80 45 66
Tropical 27/76 25 32

Proc. of SPIE Vol. 8706 87060G-6

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 i (ave) 7.65 -12 µm
Figure 6

US MS MW SS SW T

70

o
Tampa St. Louis San Diego Boston Seattle

i (ave) 3.4 -5µm

US MS MW SS SW T
50
10 Km Range
45

40
35

30

25
Summer
20
Tampa St. Louis San Diego Boston Seattle

Proc. of SPIE Vol. 8706 87060G-7

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 Table 3, and Figure 7, below summarizes the NVThermIP calculations for Sensor 13, a 7.65 to 12 µm scanned
sampled sensor. The table is divided into two main parts. The left hand side gives the range dependent P(det)s
values at the top (highlighted in yellow) and average in-band transmission values (highlighted in green) at the
bottom using MODTRAN® 4 and MODTRAN® 5 with a US Standard atmosphere. As can be seen, the difference
between the MODTRAN®4 and MODTRAN®5 calculations are insignificant. Please see below for a discussion of
this result.
ic Table
scenario.
3

Sensor 13
NVThermIP RANGE PERFORMANCE PREDICTIONS 7.65 - 12 pm

MODTRAN 4 MODTRAN 5 MODTRAN 5

US Std US Std Local Conditions

P(det) August
Range P(det) P(det) Tampa St. Louis San Diego Boston Seattle
0 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1
2 0.999976 0.999975 0.998718 0.999644 0.999779 0.999818 0.999922
3 0.998703 0.998659 0.963891 0.98728 0.99138 0.992719 0.99643
4 0.991125 0.990905 0.831478 0.931213 0.951045 0.957823 0.977668
5 0.972334 0.971708 0.603775 0.819845 0.866955 0.883404 0.934372
6 0.941094 0.940032 0.328605 0.670899 0.748268 0.77651 0.867694
7 0.899314 0.897754 0 0.509718 0.61379 0.653113 0.784652
8 0.850089 0.84807 0 0.354791 0.479759 0.527599 0.693583
9 0.796003 0.793608 0 0.213995 0.356318 0.409708 0.601469
10 0.740365 0.737696 0 0.079624 0.245911 0.304219 0.513875

Atmospheric Transmission P(det) February

Tampa St. Louis San Diego Boston Seattle
Range T (ave) % T (ave) % 1 1 1 1 1
1 0.879 0.8745 1 1 1 1 1
2 0.808 0.8027 0.999926 0.999985 0.999958 0.999988 0.999979
3 0.7514 0.745 0.99657 0.999137 0.997893 0.999287 0.998827
4 0.702 0.6955 0.978407 0.993841 0.986096 0.994835 0.991869
5 0.6581 0.6516 0.936329 0.980262 0.957792 0.983243 0.974313
6 0.6185 0.6121 0.871391 0.957235 0.912035 0.963339 0.944953
7 0.5823 0.576 0.790063 0.925699 0.852557 0.935965 0.905436
8 0.5491 0.543 0.700175 0.887594 0.784654 0.902657 0.858316
9 0.5184 0.5125 0.609918 0.845208 0.71251 0.865353 0.806777
10 0.4899 04843 0.523116 0.800399 0.640553 0.82563 0.752865

Proc. of SPIE Vol. 8706 87060G-8

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 Figure 7

On the right hand side of the table are the

10 Km Range August and February P(det)s for the
conditions of the 5 cities Tampa, Saint
Louis, San Diego, Boston and Seattle.
These calculations were all done using
MODTRAN®5 using the capabilities of
PcNvMod to model the User Supplied
Profiles.

There is both a significant difference

between the August and February
conditions and between these results and the
ones using e.g. a US Standard Atmosphere.
MOD S For example, the P(det) in Tampa during
US Std Tampa St. Louis San Diego Boston Seattle February is reasonable at 8 Km range, while
it drops to 4-5 Kms during August.

The reason for these differences can be explained by looking at the transmission calculation in Figure 4. The total
transmission, the red curve in the figure, has a structure determined by the molecular water absorption and a level
determined by the water continuum absorption. That is, the absorption in the 7 – 12 µm is highly dependent on the
H2O concentrations, i.e. the relative humidity.

It is important to realize that this is a relatively simple example which illustrates the importance of accurately
modeling the atmospheric conditions. We have not discussed the importance of other atmospheric constituents, e.g.
aerosols, which MODTRAN®/PcModWin/PcNvMod allow the user to model with great accuracy.

Section 3.1. MODTRAN®4 versus MODTRAN®5 – Transmission Mode & Radiance Mode

There were several significant enhancements in going from MODTRAN® 4 to MODTRAN® 5. For example,
MODTRAN® 5 has the capability to model the absorption at 0.1 to 0.2 cm-1 spectral resolution. Consequently
MODTRAN® 5 can be used for IR laser applications where line brooding in the lower atmosphere is on the order of
.2 cm-1. However this enhancement has no impact on NVThermIP simulations where we are dealing with
broadband sensor.

It is not surprising that for the example given above there is not a significant difference between the MODTRAN® 4
and 5 transmission calculations. The two most noteworthy changes between versions 4 and 5 effecting transmission
is:

1. MODTRAN® 5 used a newer version of the HITRAN database to compute the band models used for the
molecular absorption. There were no major changes in the water spectral parameters which will effect
NVThermIP predictions for the spectral region of the above example.
2. MODTRAN® 5 has the capability to model a boundary layer aerosol whose extinction coefficient obeys
the Angstrom law or to modify the extinction of a model aerosol with an Angstrom law perturbation. We
did not make an attempt to model aerosols in this example.

Proc. of SPIE Vol. 8706 87060G-9

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 This is not to say that there are no cases where both aerosols and changes in HITRAN will not impact sensor
performance.

Finally, MODTRAN® 4 should not be used for accurate radiance modeling. MODTRAN® 5 is a necessity for
radiance computations.

Table 4, and Figure 8, below is a similar set of calculations for Sensor 7, a 3.4 – 5.0 µm staring sensor. The
difference between using a standard atmosphere and the local conditions are not as pronounced as they were for
Sensor 13. This is because water, the
Figure 8
constituent we are changing, is not as
Sensor 7: 3.4 -5 µm significant a contributor to the molecular
0.4
absorption.

0.2
MOD4 MODS
US Std US Std Tampa St. Louis San Diego Boston Seattle

Proc. of SPIE Vol. 8706 87060G-10

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 Table 4

'Sensor 7 1

NVThermIP RANGE PERFORMANCE PREDICTIONS 3.4- 5 µm

MODTRAN 4 MODTRAN S MODTRAN S

US Std US Std Local Conditions

P(det) August
Range P(det) P(det) Tampa St. Louis San Diego Boston Seattle
0 1 1 1 1 1 1 1
1 0.999468 0.999478 0.999309 0.999386 0.999419 0.999429 0.99948
2 0.969134 0.969726 0.963042 0.965973 0.967255 0.967684 0.969787
3 0.883047 0.885041 0.864713 0.873365 0.877166 0.878486 0.884976
4 0.774181 0.777529 0.744627 0.758318 0.764335 0.766494 0.777076
5 0.667912 0.672168 0.630709 0.647661 0.655106 0.657841 0.671254
6 0.573902 0.578634 0.532187 0.551006 0.559194 0.562255 0.577291
7 0.49399 0.498891 0.450358 0.469593 0.478213 0.48143 0.497183
8 0.427109 0.431961 0.383161 0.402295 0.41068 0.413931 0.429985
9 0.371181 0.376033 0.328094 0.346699 0.354853 0.358054 0.37385
10 0.324619 0.329083 0.282816 0.300663 0.308488 0.311592 0.326774

Atmospheric Transmission P(det) February

Tampa St. Louis San Diego Boston Seattle
Range T (are) °..'ó T (ave) °'ó 1 1 1 1 1
1 0.6593 0.6635 0.999486 0.999588 0.999518 0.999603 0.999565
2 0.5831 0.5909 0.970001 0.97457 0.971417 0.975293 0.973488
3 0.5311 0.5406 0.885621 0.90046 0.890142 0.902921 0.896838
4 0.491 0.5012 0.7781 0.803301 0.785663 0.807652 0.797001
5 0.4581 0.4687 0.672515 0.705488 0.682279 0.711385 0.697074
6 0.4303 0.4409 0.578662 0.616615 0.589782 0.623612 0.606757
7 0.4061 0.4167 0.498594 0.539296 0.510414 0.547002 0.528555
8 0.3849 0.3953 0.431385 0.473235 0.443445 0.481349 0.462043
9 0.3659 0.3761 0.375202 0.417129 0.387199 0.425433 0.405798
10 0.3489 0.3588 0.328055 0.369506 0.339957 0.37785 0.358231

4. SUMMARY & RECOMMENDATIONS

This paper has shown the importance of accurately modeling atmospheric effects to predicting the performance of
an EO sensor. A simple example has demonstrated how real conditions for several US sites will significantly

Proc. of SPIE Vol. 8706 87060G-11

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 change the Probability of Detection, and hence Recognition and Identification, using the NVESD NVThermIP
simulation.

We have not exploited the full capabilities of the MODTRAN® atmospheric transmission and radiance model. Nor
was it our intent to do so. It is beyond the scope of a paper to illustrate the extent and sophistication of
MODTRAN®. NVThermIP is a similarly sophisticated model and its complete capabilities are best utilized when
the atmosphere is correctly quantified.

NV-IPM: The NVESD is in the process of releasing a significant upgrade to their modeling capabilities called NV-
IPM. "The Night Vision Integrated Performance Model (NV-IPM) is designed to provide a single, comprehensive
interface which contains all of the functionality of the previous generation of Night Vision imager performance
models, along with many requested features and additions. The interface is designed for ease of use in conducting
trade studies, changing inputs, and incorporation of new sensor modalities. The underlying structure of the model is
designed to be as completely generic as possible in order to provide as powerful and complete an interface as
possible". (5) The Java application has been released in a Beta version, most recently last spring.

NVESD is conducting a NV-IPM Workshop during this meeting on Wednesday evening from 5 to 6:30 pm.

NV-IPM offers similar atmospheric options to NVThermIP, e.g. Beer’s Law, Lookup Table, Turbulence and a
MODTRAN® Database. The difference is that the user can no longer run MODTRAN®, but relies on a lookup
table approach based on the scenario. The approach is also only available for ground-to-ground calculations. This
approach was taken, we understand, for speed considerations. It is much faster to use precalculated values from a
lookup table than to do the calculations. Unfortunately the precalculations distributed with the Beta version only
“cover” the limited MODTRAN® capabilities of the basic NVThermIP model. It is virtually impossible to
construct a lookup approach that includes all the capabilities encompassed MODTRAN®. Also the calculations in
the most recent Beta version were done using a version of MODTRAN® 4. NVESD has indicated that they may
redo the precalculations using MODTRAN®5.

Finally, NV-IPM may be using Path Radiance calculations as a component of detection, recognition, identifications
calculations in the future. It is impossible, not just virtually, to construct a lookup approach for radiance
calculations.

Recommendations for NVThermIP (2009 and previous) users:

1. At a minimum, update from MODTRAN®4 to MODTRAN®5. This can be done by “tricking” the
software. First you need to obtain a PC MODTRAN®5 executable. Copy this to the NVThermIP files
directory. Delete the file mod4v1r1.exe, if it is present in the directory. Rename your MODTRAN®5
executable mod4v1r1.exe. This will work for the input files generated by NVThermIP, however the user
should check the MODTRAN® output file, tape7 to insure that the program correctly executed.
2. Install software that will let you use the full MODTRAN®5 capabilities with your NVThermIP
calculations.

Recommendations for NV-IPM users:

1. Contact NVESD and recommend that they put the hooks in NV-IPM that will allow the user full access to
all the MODTRAN®5 capabilities that exist in e.g. NVThermIP. We realize that this will “slow down” the
overall calculations time. However, the hare frequently wins the race.

Proc. of SPIE Vol. 8706 87060G-12

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms

 Mr. Michael Soel of FLIR System, Inc., Billerica, MA, has suggested two additional recommendations regarding
NVThermIP and NV-IPM.

1. NvTherm/MODTRAN® range sampling: There is a limitation of the # of range points and how the two
models interact with each other. There is a need to increase the sampling increment when using
MODTRAN® with NvTherm.
2. It is important for the uses to better understand the Table input capability on the NVTherm Atmospheric
Transmission screen. The user must combine the thermal background contrast Planck function weighting
and the spectral D* weighting with the atmospheric transmission and any other system transmission
functions, e.g. a filter function as spectral inputs to the table.

Acknowledgements

The authors wish to acknowledge the suggestions and recommendations made by Mr. Michael Soel of FLIR
Systems Inc., Billerica, MA (http://www.flir.com/US) and Dr. Gerald Holst of JCD Publishing Co., Winter Park, FL
(http://www.jcdpublishing.com)

This work was funded by the Ontar Corporation.

References:

1. Night Vision Thermal and Image Processing Performance Model, User’s Manual, August 2009,
DOCUMENT : Rev 10, U.S Army RDECOM CERDEC, Night Vision and Electronic Sensors Directorate,
Modeling & Simulation Division, Fort Belvoir, VA

2. Acharya, P.K., Anderson, G.P., Berk, A., Shettle, E.P. “MODTRAN®5.2.0.0 USER’S MANUAL.” July
2008. Burlington, MA. Spectral Sciences, Inc.

3. Ontar Corporation. “PcModWin® Manual Version 5.0 v2r0.” August 2009. North Andover, MA.

4. Ontar Corporation. “NvMod and PcNvMod Manual”. July 2006. North Andover, MA.

5. Teaney, Brian. “Night Vision Integrated Performance Model, NV-IPM, Version 0.3 (Beta)”.

Proc. of SPIE Vol. 8706 87060G-13

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/07/2013 Terms of Use: http://spiedl.org/terms