On SEBI-SEBS validation in France, Italy, Spain, USA and China Massimo Menenti Li Jia 2 and ZongBo Su 2 - Laboratoire des Sciences de l Image, de l Informatique et de la Télédétection (LSIIT), Strasbourg, France -Istituto per i Sistemi Agricoli e Forestali del Mediterraneo (ISAFoM), National Research Council (CNR), Italy 2 ALTERRA Green World Research, Wageningen University and Research Centre, 6700 AA Wageningen, The Netherlands; September 17th 2003 - ICID RS ET - Montpellier 1
land evaporation: processes and observations evaporation land potential actual heat balance surface water balance soil water balance air Rn+ G+ H+ LE=0 P + I + Q + δm + δw + E = 0 δv δt + div q September 17th 2003 - ICID RS ET - Montpellier 2 v = E P
Actual evaporation: a history of methods heat balance land surface linear relationship [1] SVAT [4] local maximum ET [6] linear relationship B, n variable [2] SVAT + PBL [5] improved observations [7] NDVI vs. Trad [3] Menenti, 2001 September 17th 2003 - ICID RS ET - Montpellier 3
RS flux-algorithms albedo LE 0 SEBAL (Bastiaanssen, 1995) H 0 temperature pbl-temp SEBI (Menenti & Choudhury, 1993; Su et al, 2000) obs Dual - view angle measurements of surface temperature (Menenti et al, 2001) September 17th 2003 - ICID RS ET - Montpellier 4
SEBI: theory and definitions Menenti and Choudhury (1993) SEBI = (T 0 - T re (T 0 - T ( re ) a d a ) (T 0 - T a ) - ( re ) w ) d (T 0 - T a ) - ( re ) w w w Λ r H λe = λ E wet λe =1- SEBI Λ= R G n ( )( R G) = 1 Λ n Reference height = PBL top or blending height T a = potential air temperature θ max H ρ a max c u p * = 1 ln κ h z 0h C h L max θ min H ρ c u a min p * = 1 ln κ h z 0h C h L min H max = Q* G H min : s( Q G) + ρc / r ( e e) September 17th 2003 - ICID RS ET - Montpellier 5 λe wet = * a p s+γ a sat
Case studies France, DAISEX 1999 at Colmar and Hartheim (Jia et al., 2001) Italy, Pian di Rosia 1997 (Roerink et al., 2000). Spain (Jia et al., 2003) USA SGP 97 (Su et al., 2002) China (Su et al., 2003) September 17th 2003 - ICID RS ET - Montpellier 6
DAISEXL Overview of data collected - DAIS (Digital Airborne Imaging Spectrometer) - HYMAP (Hyperspectral Mapper) - ROSIS (Reflective Optics System Imaging Spectrometer) - POLDER (POLarized and Directional Earth's Reflectances) -WAAC (Wide Angle Airborne Camera) - LEANDRE LIDAR (aerosols lidar) - ARAT (in-situ atmospheric sensors) - ERS-1/ERS-2 Tandem SAR - Landsat TM / SPOT data (for the dates of campaigns) THREE YEARS OF CAMPAIGNS: Barrax: 1998, 1999 and 2000; Alsace: 1999 September 17th 2003 - ICID RS ET - Montpellier 7
SEBI from field measurements SEBI Normorlized (T0 - Ta) (K / (s m-1)) 0.6 Wet condition 0.4 Actual Dry condition 0.2 0-0.2-0.4 0.17 0.18 0.19 0.2 0.21 0.22 albedo Energy flux ( W m-2) 600 500 400 300 200 100 0-100 09:00 09:20 09:32 09:40 12:00 12:20 12:48 16:14 17:00 Hour(UTC) Rn G H_obs H_s ebi LE_obs LE_s ebi Flux densities 350 Colmar, sugar beet: SEBI estimates vs. field measurements RMSD = 17.4 Wm -2 LE estimation by SEBI ( W m-2 ) RMSD = 17.4 W m-2 300 250 200 200 250 300 350 LE obsevation ( W m-2 ) September 17th 2003 - ICID RS ET - Montpellier 8
Dry and Wet reference vs. Observations September 17th 2003 - ICID RS ET - Montpellier 9
From SEBI to LE - map Normalized (T0-Ta) ( K/(s m-1) ) 0.8 0.6 0.4 0.2 0-0.2 Colmar sugarbeet Wet boundary Actual Dry boundary normalized (T 0 T a ) versus surface albedo; airborne HYMAP, DAIS, Colmar -0.4 0 0.05 0.1 0.15 0.2 0.25 albedo LE Sugar beet Bare soil Distribution of LE at Colmar sub-site; cross indicates location of field measurements Fallow maiz Wm -2 September 17th 2003 - ICID RS ET - Montpellier 10
The Simplified Surface Energy Balance Index ( SEBI ) 60 250 Surface temperature ( o C) 50 40 30 20 T H T λe T 0 H max (r 0 ) λe max (r 0 ) mean temperature H [measured] (W/m2) 200 150 100 50 eddy correlation bowen ratio scintillometer 10 0 0.2 0.4 0.6 0.8 1 Surface reflectance (-) 0 0 50 100 150 200 250 H [S-SEBI] (W/m2) Menenti and Choudhury, 1993; Roerink et al., 1999 September 17th 2003 - ICID RS ET - Montpellier 11
SEBS - The Surface Energy Balance System Input SEBS Core Modules Output Meteorological Data Boundary Layer Variables Remote Sensing Data Boundary Layer Similarity Theory Roughness for Heat Transfer Turbulence Heat Fluxes Evaporative Fraction NIR VIS TIR Surface Energy Balance Index Actual Evaporation September 17th 2003 - ICID RS ET - Montpellier 12
Spain: validation with scintillometers Location of scintillometer measurements sites Location and characteristics of scintillometer experimental sites in Spain. Site Tomelloso Lleida Badajoz Transmitter Receiver Distance between Surface Location transmitter and receiver (m) Characteristics Location 39 07.357 N 2 55.314 W 41 32.644 N 0 51.644 E 38 55.697 N 6 36.590 W Height (m) 4.56 39 07.653 N 2 55.951 W 39 41 34.962 N 0 52.444 E 68 38 56.298 N 6 40.141 W Height (m) 4.15 1070 Dry vineyard 45 4440 Small scale irrigation area with fruit trees, alfalfa 56 5250 Large scale irrigation area with wheat, corn, alfalfa, lettuce, olives, beans, tomatoes. September 17th 2003 - ICID RS ET - Montpellier 13
Surface temperature (K) Surface albedo ( ) NDVI ( ) Fractional vegetation cover ( ) PBL depth (m) PBL pressure (Pa) PBL potential temperature (K) PBL speci.c humidity (%) PBL wind speed (m/s) Channel SEBS Data Requirements ATSR ATSR ATSR ATSR NWP-RACMO RACMO RACMO RACMO RACMO Central wavelength 50% band wid (mm) (mm) 1 * 12.0 11.60-12.50 2 * 11.0 10.52-11.33 3 3.7 3.47-3.90 4 1.6 1.575-1.642 5 * 0.87 0.853-0.875 6 * 0.65 0.647-0.669 7 0.55 0.543-0.565 ATSR channels September 17th 2003 - ICID RS ET - Montpellier 14
Sensible heat flux: SEBI vs. scintillometers H estimated by SEBS (W m -2 ) 350 300 250 200 150 100 Tomelloso Lleida Badajoz SEBI (SEBS version) ATSR-2: surface temperature, albedo and NDVI vertical error bars = standard deviation over pixels along path horizontal error bars = error LAS measurements 100 150 200 250 300 350 H observed by LAS (W m -2 ) September 17th 2003 - ICID RS ET - Montpellier 15
SEBI: Forward and nadir T 0 USA: Southern Great Plain 1997 Hydrology Experiment (SGP 97) (Nadir View) (Forward View) Colour from red to yellow, green, blue indicates increasing evaporative fraction forward view = higher fraction of foliage, lower surface brightness temperature higher estimate of evaporative fraction September 17th 2003 - ICID RS ET - Montpellier 16
China: Monitoring Droughts July 1, 2000 July 8, 2000 September 17th 2003 - ICID RS ET - Montpellier 17
Comparison with Soil Moisture Measurements China: Relative evaporation vs relative soil moisture April 2000 April 2000 April 2000 Relative Soil Moisture (at10 cm depth) 80 70 60 50 40 30 20 10 0 0 10 20 30 40 50 Relative Evaporation (%) SOIL MOISTURE (at10 cm depth) Predicted Average Relative Soil Moisture Avergae Relative Soil Moisture up to 20 cm Depth (%) 80 70 60 50 40 30 20 10 0 0 10 20 30 40 50 Relative Evaporation (%) Average Relative Soil Moisture up to Depth of 20 cm Predicted Average Relative Soil Moisture Average Relative Soil Moisture up to depth of 50cm (%) 80 70 60 50 40 30 20 10 0 0 20 40 60 Relative Evaporation (%) 6 Average Relative Soil Moisture up to Depth of 50 cm Predicted Average Relative Soil Moisture Time Series of Drought Severity Index September 17th 2003 - ICID RS ET - Montpellier 18
Development related changes in land cover and Colour composite of multitemporal images of the Soil Adjusted Vegetation Index (SAVI) calculated with AVHRR Ch1 and Ch2 reflectance at 1 km spatial resolution, Aral Sea Basin 1992: SAVI (April)= red; SAVI (june)= green; SAVI (august)= blue; area is 1568 km x 1232 km; greenish areas = irrigated lands water balance September 17th 2003 - ICID RS ET - Montpellier 19
Concluding Remarks Method performs well under different conditions Use of NWP models to determine PBL variables makes large area applications feasible Error budgets give significantly higher error estimates SEBS will be further developed by integrating generic algorithms to process TOA radiometric data Case studies should be expanded to time series September 17th 2003 - ICID RS ET - Montpellier 20