Railroad Valley Playa for use invicarious calibration of large

footprint sensorsK. Thome, J. Czapla-Myers, S. Biggar

Remote Sensing Group Optical Sciences Center

University of Arizona

P Background!Reflectance-based vicarious calibration!Test sites

P Results from Terra, EO-1, and Landsat

P Issues with current methodology!Temporal sampling!Noise/errors!Railroad Valley test site

P Calibration without ground-personnel!LED radiometers and atmospheric monitoring!Results from early measurements

P Vicarious calibration test site modeling

P Intercomparison possibilities

P Conclusions and future work

Introduction

Reflectance-based ApproachCombine surface reflectance and atmospheric

transmittance data to predict at-sensor radiance

RadiativeTransfer Code

RSG Test Sites

Rely on dry lakesand gypsum saltflats in California,

Nevada, andNew Mexico

(USA)

RSG Test Sites

Railroad Valley Playa,Nevada

Ivanpah Playa, California

White Sands MissileRange not shown

Landsat-7 results

ETM+ work indicates that there has not been significantdegradation of the sensor so use average and standard

deviation of difference in at-sensor radiance

Band 1 Band 2 Band 3 Band 4 Band 5 Band 7-9

-6

-3

0

3

6

P Sample a 1-km by 1-km area

P Takes approximately 1 hour to collect data

Large-footprint approachApplication to large-footprint sensors requires a different

surface reflectance sampling approach

Terra MODIS resultsTerra MODIS is also well-behaved with no significant

degradation at Level 1B

412 nm469 555 645 858 905 1240 1640 2130

-10

-5

0

5

10

15

20Terra MODIS

Landsat 7 and Terra Results

0.4 0.5 0.6 0.7 0.8 0.9

Wavelength (micrometers)

-9-6-30369

ASTERMISRMODISETM+

1.2 1.4 1.6 1.8 2 2.2 2.4

Wavelength (micrometers)

-9-6-30369

ASTERMISRMODISETM+

Results with EO-1

412 nm 469 555 645 858 905

-15

-10

-5

0

5

10

15Terra MODIS ETM+

ASTER ALI

Hyperion MISR

P Features in the percent difference are repeatable

P Same features are seen in the lunar calibrationresults

Hyperspectral exampleResults below are average and standard deviation of

five VNIR Hyperion data sets

0.40.40.40.4 0.50.50.50.5 0.60.60.60.6 0.70.70.70.7 0.80.80.80.8 0.90.90.90.9 1111

Wavelength (micrometers)

-20-20-20-20

-10-10-10-10

0000

11110000

22220000

AverageAverageAverageAverage1-sigma1-sigma1-sigma1-sigma std.std.std.std. devdevdevdev

P Examination of ETM+ results does not show anobvious cause

P Scatter most likely “errors” in surface reflectance

P Outliers due to anomalous atmospheres

Noise - outlier data sets One major drawback of the reflectance-based approach

are outlier data sets

0 500 1000 1500 2000

Days since launch

1.4

1.45

1.5

1.55

1.6

PreflightAverageGround-reference

P All data from RRVPlaya

P Mostly coincidentdates for plotshown here

P Standarddeviations slightlydifferent for two

P Note the outlierfrom thereflectance-basedapproach

Noise, MODIS-ASTER exampleRecent work has used MODIS as a reference for an

intercomparison with ASTER

600 800 1000 1200 1400

Days Since Launch

0.7

0.75

0.8

0.85

Reflectance-basedMODIS intercomparison

Temporal sampling issuesRailroad Valley

Playa is in centralNevada about 13hours by car fromthe University of

Arizona 0 500 1000 1500Days since Jan. 1, 2000

-10

-5

0

5

10Terra

P Require personnel on the ground at sensor overpass!Expensive in personnel and travel!Reduces opportunities for calibration attempts

P Currently make approximately one trip per month!Cannot get all sensors on all trips!Weather prevents success in some cases!Fortunate to obtain 8-10 data sets per sensor per

year

P These 8-10 data sets may not be sufficient for trendanalysis

P Goal - increase the number of data sets per sensorwithout sacrificing accuracy

Improved temporal samplingPoor temporal sampling has always been an issue with

the reflectance-based method

Ground-based instrumentationUse the same methodology but replace instruments

with sensors that do not require personnel to be present

RadiativeTransfer Code RadiativeTransfer Code

P Sunphotometer provides atmospheric optical depths anduses sky radiance data to produce aerosol size and type

P Data are available via the Web from Goddard SpaceFlight Center’s Aeronet

P Meteorological station provides ancillary data includingrainfall

Atmospheric dataAtmospheric measurements rely on a meteorological

station and automated Cimel sunphotometer

P Test sites used for vicariouscalibration are typically bright

P Uncertainties in the surfacereflectance cause the samelevel of uncertainty in thevicarious results

P Could assume the surface isinvariant

P NOT a good assumption at Railroad Valley

Surface reflectanceMost critical measurement is the

surface reflectance

P Benefit of combining spectral selection and detector!Reduces cost! Improves spectral and radiometric stability over time!Others have shown this stability to be much better

than 1% over periods in excess of 10 years

P Have a range of wavelengths available!Focus is currently on the visible and near infrared!Detector wavelength shifts relative to the emitting

wavelength

LED radiometersMonitor surface reflectance via a set of robust,

inexpensive radiometers relying on light emitting diodes(LEDs) operating as detectors

LED radiometersCurrent results are based on a simple design with a

four-channel approach

P The spectral bands are green, red, and NIR!Bands are similar to those of several earth-

imaging sensors!Bands are wider

than those typically used in imagers

P Spectral response varies somewhat from LED to LED but the wider bands help mitigate this

LED radiometers - Spectral responseOf the four channels, three survived assembly and early

deployment to Railroad Valley

0.0

0.2

0.4

0.6

0.8

1.0

400 500 600 700 800 900 1000

Wavelength (nm)

Green Red NIR

LED radiometer - reflectance retrieval

Output of LED radiometerdepends on the incidentsun angle, atmosphericconditions, and response

Correcting for these effectsallows the reflectance to befound

0.0

0.3

0.6

0.9

78.5 79.5 80.5

Day of Year (UTC)

00.10.20.30.40.50.60.7

78.5 79.5 80.5

Day of Year (UTC)

LED radiometer - resultsLED results are used to determine a hyperspectral

surface reflectance for the vicarious calibration

0.35

0.40

0.45

0.50

500 600 700 800 900

Wavelength (nm)

LED Radiometer 1

LED Radiometer 2

Fitted hyperspectralreflectance

P Also shown are the results from the full ground-based data (open circles)

P Results are very good in visible and poorer atlonger wavelengths

LED radiometer - resultsGraph below shows the reflectance-based results for

the three sensors

0.4 0.8 1.2 1.6 2 2.4Wavelength (micrometers)

-20-15-10-505

10

ETM+ LED-basedETM+ reflectance-basedASTER LED-based

ASTER reflectance-basedMODIS LED-based

P This model will provide at-sensor radiance for a givensun-sensor geometry!Hyperspectral at 10-nm intervals from 350-2500 nm!30-m spatial resolution

P Combination of ground-based LED and satelliteimagery!Rainfall data give information regarding sharp

changes in reflectance!ETM+ data (or similar system) give spatial information!MODIS can give directional reflectance data along

with the LED data!Cimel provides atmospheric data

Model-based playaOne goal of this work is to develop a model of the

Railroad Valley Playa

P BRDF effects

P Spectral variabilityacross the playa

P Spatial variabilityover time

P Anomalousbehavior of playa

P Fortunately, all ofthese are also ofinterest to thereflectance-basedmeasurements

Model-based approach

Numerous issues must be addressed for this to work

P Clearly not invariant

P Note large changein middle row

P Last image takenone month aftersnow melt

Model-basedapproach

ASTER Band 3data from RRV

2002-06-15 2003-02-09

2003-06-20 2003-07-04 2003-07-06 2003-07-22

2003-08-23 2003-11-11 2004-03-18

2001-01-05

P Note that this shows that repeatability/precisionand accuracy are not identical

P Many of the cases shown have not overlappingtime periods

Reflectance-based intercomparisonsResampling of Landsat ETM+ results show that as fewas five data sets can provide a repeatable estimate of

sensor calibration

All 1 2 3 4 5 6 7 8 9 10

-6

-3

0

3

6

Intercomparison - MODISReflectance-based method allows for direct comparison

of results from two sensors without concurrent views

412 nm469 555 645 858 905 1240 1640 2130

-10

-5

0

5

10

15

20Terra MODIS Aqua MODIS

Intercomparison - MODISStill need some work to understand behavior of ground

data results relative to other vicarious methods

0 5 10 15 20

-2

-1

0

1

2

3

4

5

6Reflectance-based

Lunar-based

P Better understandvicarious approach (effectof atmospheric models)

P Self-consistency within theUofA laboratory andconsistency with fieldmeasurements!Same panels are used in

field and for radiancecalibration

!Multiple calibrationapproaches forlaboratory radiometer

Future IntercomparisonsIntercomparisons between laboratory radiometers

calibrated to radiance and model-predicted radiancesare currently being done

P Vicarious methods such as reflectance-based methodare now more repeatable

P Vicarious do not require coincident collections (evenallows gaps in the data record)!Does require consistent application of single method!Best when there is consistent sensor collection

methodologies (view angles, protocols)

P Results shown here showed some small biasesbetween several sensors!Biases could be real!Shows need for multiple intercomparison methods! In the case of large biases a decision must be made

regarding the “right” answer

Conclusions - IntercomparisonsVicarious methods can be used for sensor

intercomparisons

P ETM+ results between the two approaches agreedto better than 3% in the VNIR!ASTER did not have as good agreement -

possibly due to spatial atmospheric effects!SWIR results poorer due to assumptions used to

obtain the hyperspectral reflectance

P Two single point LED values gave good results forMODIS! In reality, this was somewhat fortuitous!Area of Railroad Valley was very uniform on this

date in the region of the LEDs!Future work will deploy more radiometers to

assess the spatial uniformity

Conclusions - LED resultsLED and Cimel results gave similar accuracies as the

full up ground-based measurements

P Precision of vicarious methods is improving!Repeatability used here as a surrogate for precision!Becoming more difficult to determine error sources

and how to correct

P Links/traceability to laboratory standards are needed!Solar-based calibration approaches!Laboratory-quality field radiometers

P Temporal sampling issues!What is the optimal sampling frequency?! “Clumping” of vicarious results may be preferred

P Vicarious methods should be considered whenplanning preflight characterizations!Size of source!Spectral nature

General Conclusions and Comments