Mapping Evapotranspiration and Drought at Local to Continental Scales Using Thermal Remote Sensing

Mapping Evapotranspiration and Drought at Local to Continental Scales Using Thermal Remote Sensing

M.C. Anderson, W.P. KustasUSDA-ARS, Hydrology and Remote SensingLaboratory

J.M NormanUniversity of Wisconsin - Madison

Remote Sensing of Surface MoistureRemote Sensing of Surface Moisture

Microwave emissivity

Surface dielectric properties

Surface moistureSurface layer: ~5cm

MICROWAVEMICROWAVE

Canopy temperature

Canopy transpiration

Root-zone moisture

(10-50 km resolution)(10-50 km resolution)

Soil temperature

Soil evaporation

Surface moisture

THERMALTHERMAL(60m – 10km resolution)(60m – 10km resolution)

Rootzone: ~1-2m

Satellite Thermal Imaging SystemsSatellite Thermal Imaging Systems

ASTERLandsat

Once every 8-16 days

90–120 mFine

VIIRSAVHRRATSR

MODISAVHRRATSR

2-4 times daily1 kmModerate

CrISGOESMSG

AIRSGOESMSG

15 min5-20 kmCoarse

FutureSources

CurrentSources

Temporal Resolution

Spatial Resolution

PixelScale

Table from S. Hook

… evapotranspiration… evapotranspirationAPPLICATIONSAPPLICATIONS

ALEXI – Atmosphere-Land Exchange Inverse ModelALEXI – Atmosphere-Land Exchange Inverse Model

(Anderson et al, 1997)(Anderson et al, 1997)

TRAD(θ) ~ fc(θ)Tc + [1-fc(θ)]Ts

(two-source approximation)Norman, Kustas et al. (1995)

Accommodates off-nadirthermal sensor view angles

Treats soil/plant-atmosphere coupling differences explicitly

Two-Source Energy Balance Model (TSEB)Two-Source Energy Balance Model (TSEB)

θθ

HSHS

HCHC

H = HC + HSH = HC + HS

RS

RA

RX

TATA

TCTC

TSTS

TAEROTAERO

fc(θ) = 1 – exp -0.5 Ω(θ) LAI

cos (θ)

ET = (RNET - G) - HET = (RNET - G) - HAvailable energy

Atmosphere-Land Exchange Inverse (ALEXI)Atmosphere-Land Exchange Inverse (ALEXI)

Regional scaleΔTRAD - GOESfc - MODIS

Landscape scaleTRAD - TM, ASTER, MODISfc - TM, ASTER, MODIS

Rsoil

TcTac

Hs

Ts

RaH = Hc + Hs

Rx

Hc

Ta

ABL

Ta

ALEXI DisALEXI5 km

30 m

TRAD (φ), fc

TRAD,i(φi), fc,i

i

Ra,i

Blending height

Rsoil

TcTac

Hs

Ts

RaH = Hc + Hs

Rx

Hc

TaTa

ABL

Ta Ta

ALEXI DisALEXI5 km

30 m

TRAD (φ), fc

TRAD,i(φi), fc,i

i

Ra,i

Blending height

Surface temp:Cover fraction:

Nor

man

and

Kus

tas

et a

l. (1

995)

Tw

o-S

ou

rce M

od

el

Anderson et al. (1997) Norman et al. (2003)

30

35

40

45

100

200

300

400

500

1 July 2002 – 10:30AM CST

Dis

ALE

XI(L

ands

at)

Dis

ALE

XI(U

SU

airc

raft)

Catchm

ent(60m

)Field scale

(30m)

CORNCORN

SOYSOY

SURFACE TEMPERATURE EVAPOTRANSPIRATION

ALE

XI(G

OE

S Im

ager

) Regional(5km

)A

LEXI

(GO

ES S

ound

er) C

ontinental(10km

)

Tem

pera

ture

(C)

Latent Heat (W

m-2)

Dis

ALE

XI(M

OD

IS)

Watershed

(1km)

GOES CoverageGOES Coverage

GOES – West (Pacific)

Copyright 2008 EUMETSATCopyright 2008 EUMETSAT

GOES - East

AP

RA

PR

MA

YM

AY

JUN

JUN

JUL

JUL

AU

GA

UG

SE

PS

EP

20022002 20032003 20042004

(Wm

-2)

EV

AP

OT

RA

NS

PIR

AT

ION

LowH

igh

(Anderson et al, 2007)

ALEXI validation sitesALEXI validation sites

SMEX02/05SMEX02/05

SGP97SGP97

BondvilleBondville

Fort PeckFort Peck

Walker BranchWalker Branch

AudubonAudubon Goodwin CreekGoodwin CreekGainesvilleGainesville

EvergladesEverglades

SevilletaSevilletaSMEX04SMEX04BushlandBushland

Black HillsBlack Hills

BARCBARC

Clear-sky fluxes using Landsat thermal (60m)Clear-sky fluxes using Landsat thermal (60m)

• rangeland• pasture• corn• soybean…

0

200

400

600

800

0 200 400 600 800

Tower flux (Wm-2)

Mod

eled

flux

(Wm

-2)

RNETHG

RMSD: 33 Wm-2 (10%)

(Anderson et al, 2007)

… drought & stress monitoring… drought & stress monitoring

ESI = 1 –ESI = 1 –AETAETPETPET

APPLICATIONSAPPLICATIONS

Evaporative Stress IndexEvaporative Stress Index

0

10

20

30

40

50

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Date

Cum

ulat

ive

area

l per

cent

age

Climatological Study: 2002-2005Climatological Study: 2002-2005

Extreme dryExtreme dryExtreme wetExtreme wet

2002 2003 2004Palmer Drought Index - NCDC

• 2002: extreme-severe drought conditions covering 40% of the US in July• 2003: some improvement (10-25% extreme-severe drought coverage) • 2004: extreme drought coverage falls < 5% due to increased late rainfall• 2005: wetter than average in western US, drought in east

2005

20022002 20032003 20042004 20052005A

PRAPR

MA

YM

AY

JUN

JUN

JUL

JUL

AU

GA

UG

SEP

SEP

DRY WETMONTHLY ESI ANOMALIESMONTHLY ESI ANOMALIES

APRAPR

MA

YM

AY

JUN

JUN

JUL

JUL

AU

GA

UG

SEP

SEP

ΔΔESIESI ΔΔZZ20022002

2002WetDry

EvaporativeStress IndexEvaporativeStress Index

2002

APRAPR

MA

YM

AY

JUN

JUN

JUL

JUL

AU

GA

UG

SEP

SEP

ΔΔESIESI ΔΔZZ20032003

2003WetDry

EvaporativeStress IndexEvaporativeStress Index

2003

APRAPR

MA

YM

AY

JUN

JUN

JUL

JUL

AU

GA

UG

SEP

SEP

ΔΔESIESI ΔΔZZ20042004

2004WetDry

EvaporativeStress IndexEvaporativeStress Index

2004

APRAPR

MA

YM

AY

JUN

JUN

JUL

JUL

AU

GA

UG

SEP

SEP

ΔΔESIESI ΔΔZZ20052005

2005WetDry

EvaporativeStress IndexEvaporativeStress Index

2005

20022002

20032003

20042004

ALEXIALEXI PrecipitationPrecipitation

ESI anomaly Precipitation anomaly (mm)Dry Wet

20052005

Multi-scale Drought MonitoringMulti-scale Drought Monitoring

GOES Evaporative Stress IndexGOES Evaporative Stress Index

Palmer Z IndexPalmer Z IndexAnomalyAnomaly

JUNE 2002JUNE 2002

Landsat2/13/02

Landsat2/13/02

Multi-scale Ecosystem Health MonitoringMulti-scale Ecosystem Health Monitoring

MODIS (1km) L7 (60m)

(hourly) (daily) (monthly)

FLORIDA EVERGLADESFLORIDA EVERGLADESGOES (10km)

(Wm-2)EvapotranspirationEvapotranspiration

Multi-scale Ecosystem Health MonitoringMulti-scale Ecosystem Health Monitoring

MODIS (1km) L7 (60m)

(hourly) (daily) (monthly)

FLORIDA EVERGLADESFLORIDA EVERGLADESGOES (10km)

Evaporative stressEvaporative stress (low)(high)

TIR descopedfrom LDCM

TIR descopedfrom LDCM

Impending data gap in high-resolution TIRImpending data gap in high-resolution TIR

ASTERLandsat

Once every 8-16 days

90–120 mFine

VIIRSAVHRRATSR

MODISAVHRRATSR

2-4 times daily1 kmModerate

CrISGOESMSG

AIRSGOESMSG

15 min5-20 kmCoarse

FutureSources

CurrentSources

Temporal Resolution

Spatial Resolution

PixelScale

Table from S. Hook

HyspIRI-TIR?HyspIRI-TIR?

… soil moisture mapping… soil moisture mappingAPPLICATIONSAPPLICATIONS

AW

F (% 0-100cm

)

May 12, 2003

May 29, 2003

July 5, 2003

August 1, 2003

May 7, 2004

June 2, 2004

August 3, 2004

June 19, 2002

ALEXIALEXIMESONETMESONET ALEXIALEXIMESONETMESONET

(Hain et al., 2008)

0

100

80

60

20

40

Volumetric Soil Moisture RetrievalsVolumetric Soil Moisture Retrievals

0

0.1

0.2

0.3

0.4

0.5

0 0.1 0.2 0.3 0.4 0.5

Observed AWF

ALE

XI A

WF

<11-22-33-4>4

LAI

r2 = 0.53RMSD = 0.06Bias = -0.013

ALEXI

Observed VSM

ALE

XI

VS

M

(0-100 cm)

Improving weather forecast accuracyImproving weather forecast accuracy

1800 UTC 1 Aug 2004

1800 UTC 2 Aug 2004

1800 UTC 1 Aug 2004

1800 UTC 2 Aug 2004

2-m temperature bias 2-m dewpoint bias

Con

trol

ALE

XI

SM

Assimilation into hydrologic modelsAssimilation into hydrologic models

Assimilating thermal-basedET/PET

Assimilating surface temperature directly

No Assimilation Assimilation

Crow et al., RSE 2007

THERMAL REMOTE SENSING DATAHAVE GREAT UTILITY:

… ET mapping… drought monitoring… soil moisture mapping

THERMAL REMOTE SENSING DATAHAVE GREAT UTILITY:

… ET mapping… drought monitoring… soil moisture mapping

CONCLUSIONSCONCLUSIONS

NEED TO MAINTAIN MULTI-SCALE THERMAL DATA SOURCES

NEED TO MAINTAIN MULTI-SCALE THERMAL DATA SOURCES

Martha.Anderson@ars.usda.govMartha.Anderson@ars.usda.gov

ALEXI_EUROPE and _AFRICA (coming soon)ALEXI_EUROPE and _AFRICA (coming soon)

METEOSAT COVERAGE

0o Operational

Copyright 2008 EUMETSATCopyright 2008 EUMETSAT

Indian Ocean

METEOSAT ApplicationMETEOSAT Application

0.6

0.7

0.8

0.9

1

0

100

200

300

400

500

600

EvapotranspirationEvapotranspiration

ESIESI (low)

(high)

LEBRIJA, SPAINLEBRIJA, SPAIN

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.2 0.4 0.6 0.8 1.0

Available water fraction

Act

ual/p

oten

tial E

T

Tra

nspi

rati

on

Canopy

Available water fraction (root zone)

Act

ual/p

oten

tial E

T

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.2 0.4 0.6 0.8 1.0

Available water fraction

Act

ual/p

oten

tial E

T

Soi

lev

apor

atio

n

Soil

Act

ual/p

oten

tial E

T

Available water fraction (soil surface)

0-5cm0-5cm

1-2m1-2m