GERB TOA radiances and fluxes validation over the Valencia Ancho GERB TOA radiances and fluxes validation over the Valencia Ancho r Station r Station during the IV GERB Ground Validation Campaign during the IV GERB Ground Validation Campaign A. Velázquez Blázquez 1 , A. Cano 1 , N. Clerbaux 2 , S. Dewitte 2 , C. Doménech 1 , V. Estellés 3 , A.G. Ferreira 1 , L. Gonzalez 2 , J. Jorge Sanchez 4 , C. Narbón 1 , D. Pino 4,6 , A. Rius 6 , G. L. Smith 5,7 , Z. P. Szewczyk 5,8 , R. Tarruella 4 , J. Torrobella 6 , E. Lopez-Baeza 1 (1) Climatology from Satellites Group, Dept. of Physics of the Earth and Thermodynamics, University of Valencia, Spain, (2) Royal Meteorological Institute of Belgium (RMIB), Belgium, (3) Solar Radiation Group, Dept. of Physics of the Earth and Thermodynamics, University of Valencia, Spain , (4) Polytechnic University of Catalonia, Spain, (5) NASA Langley Research Center (LaRC),VA, USA, (6) Institute of Space Studies of Catalonia, Spain, (7) National Institute for Aerospace, Hampton, VA, USA, (8)Science Applications International Corporation, Hampton, VA, USA 4th ESA EO Summer School on Earth System Monitoring and Modelling. ESRIN, Frascatti, 4 – 14 August 2008 ABSTRACT ABSTRACT The Geostationary Earth Radiation Budget (GERB) instrument on bo The Geostationary Earth Radiation Budget (GERB) instrument on bo ard Meteosat ard Meteosat- 8 and 8 and - 9 provides accurate measurements of shortwave and longwave broad 9 provides accurate measurements of shortwave and longwave broad band radiances and fluxes at the Top band radiances and fluxes at the Top Of the Atmosphere (TOA) with very high temporal resolution of 15 Of the Atmosphere (TOA) with very high temporal resolution of 15 minutes. minutes. The aim of this study is to validate radiances and fluxes at the The aim of this study is to validate radiances and fluxes at the TOA measured by GERB during the IV GERB Ground Validation Campa TOA measured by GERB during the IV GERB Ground Validation Campa ign at the Valencia Anchor Station (VAS) area (31 ign at the Valencia Anchor Station (VAS) area (31 st st July July – 6 th th August August 2006). In the study, GERB enhanced spatial resolution data (GERB 2006). In the study, GERB enhanced spatial resolution data (GERB High Resolution) is used, where the resolution of the computed High Resolution) is used, where the resolution of the computed fluxes is improved through the combination of well fluxes is improved through the combination of well- calibrated GERB calibrated GERB broadband data with SEVIRI narrow broadband data with SEVIRI narrow- band high band high- sampling sampling- rate data. Clouds and the Earth rate data. Clouds and the Earth’ s Radiant Energy System (CERES) Terra FM2 data is also used from s Radiant Energy System (CERES) Terra FM2 data is also used from dedicated PAPS (Programmable Azimuth dedicated PAPS (Programmable Azimuth Plane Scanning) observations over the study area during Februar Plane Scanning) observations over the study area during Februar y 2004 campaign. y 2004 campaign. The validation capabilities of the Valencia Anchor Station have The validation capabilities of the Valencia Anchor Station have previously been assessed by successfully reproducing CERES TOA r previously been assessed by successfully reproducing CERES TOA r adiances and fluxes with the occasion of different ground valida adiances and fluxes with the occasion of different ground valida tion tion campaigns. The methodology consists in performing radiative tran campaigns. The methodology consists in performing radiative tran sfer simulations of CERES and GERB TOA radiances and fluxes from sfer simulations of CERES and GERB TOA radiances and fluxes from independent ground measurements of surface and atmospheric independent ground measurements of surface and atmospheric parameters (such as derived precipitable water parameters (such as derived precipitable water vapour vapour content from CIMEL sunphotometer and GPS (Global Positioning Sy content from CIMEL sunphotometer and GPS (Global Positioning Sy stem) instruments, radiosoundings from the Spanish stations of M stem) instruments, radiosoundings from the Spanish stations of Madrid and adrid and Murcia, aerosol optical thickness also from CIMEL, broadband alb Murcia, aerosol optical thickness also from CIMEL, broadband alb edo and temperature over shrubs, bare soil and vineyards in the edo and temperature over shrubs, bare soil and vineyards in the study area) in conjunction with other satellite products such as study area) in conjunction with other satellite products such as TOMS TOMS (Total Ozone Mapping Spectrometer) ozone, CERES/SARB emissivity (Total Ozone Mapping Spectrometer) ozone, CERES/SARB emissivity and MODIS BRDF (Bidirectional Reflectance Distribution Function) and MODIS BRDF (Bidirectional Reflectance Distribution Function) . The latter allows us to analyze the contribution of each land . The latter allows us to analyze the contribution of each land use use to the anisotropy of the shortwave radiation field and constitut to the anisotropy of the shortwave radiation field and constitut es a good improvement of the methodology that had been tested fo es a good improvement of the methodology that had been tested fo r the case of CERES in previous campaigns. The comparison betwee r the case of CERES in previous campaigns. The comparison betwee n n simulations and CERES calibrated and validated data provides a g simulations and CERES calibrated and validated data provides a g ood indicator of the reliability of the methodology to be applie ood indicator of the reliability of the methodology to be applie d as a validation tool for GERB. d as a validation tool for GERB. THE VAS SITE THE VAS SITE LANDSAT 5 LANDSAT 5 5TH JULY 2003 5TH JULY 2003 (RGB) (RGB) ACKNOWLEDGMENTS ACKNOWLEDGMENTS We would like to thank the Spanish Institute for Meteorology for their collaboration on the radiosounding ascents and the Institute for Space Studies of Catalonia for their collaborations with the GPS measurements. The authors wish to thank the CERES Science Team for their support and provision of the data. To To reproduce reproduce radiances radiances and and fluxes fluxes at at the the TOA, TOA, it it is is needed needed a a good good characterization characterization of of the the surface surface and and the the atmosphere atmosphere. Atmospheric Atmospheric profiles profiles • Water Water Vapour Vapour: Radiosounding ascents interpolated to 1m resolution and Radiosounding ascents interpolated to 1m resolution and completed with MLW ( completed with MLW ( Mid Latitude Winter Mid Latitude Winter ) profiles. Then, select 94 ) profiles. Then, select 94- 95 levels 95 levels and scale the profile to the integrated water vapor retrieval ob and scale the profile to the integrated water vapor retrieval obtained with the tained with the GPS ( GPS ( Global Positioning System Global Positioning System) • Ozone: Ozone: MLW profile scaled to TOMS ( MLW profile scaled to TOMS ( Total Ozone Mapping Spectrometer Total Ozone Mapping Spectrometer ) ) values values • Aerosols: Aerosols: STREAMER MLW standard profile, assuming background STREAMER MLW standard profile, assuming background tropospheric aerosols and background stratospheric aerosols and tropospheric aerosols and background stratospheric aerosols and aerosol aerosol optical thickness from CIMEL photometer measurements optical thickness from CIMEL photometer measurements Se Se lection of surface parameters lection of surface parameters Methodology Methodology • Surface Emissivity Surface Emissivity Surface Emissivity Surface Emissivity Surface Emissivity Surface Emissivity Surface Emissivity Surface Emissivity from CERES/SARB ( from CERES/SARB ( from CERES/SARB ( from CERES/SARB ( from CERES/SARB ( from CERES/SARB ( from CERES/SARB ( from CERES/SARB ( Surface and Atmospheric Radiation Budget Surface and Atmospheric Radiation Budget Surface and Atmospheric Radiation Budget Surface and Atmospheric Radiation Budget Surface and Atmospheric Radiation Budget Surface and Atmospheric Radiation Budget Surface and Atmospheric Radiation Budget Surface and Atmospheric Radiation Budget ), Wilber et al, 1999 ), Wilber et al, 1999 ), Wilber et al, 1999 ), Wilber et al, 1999 ), Wilber et al, 1999 ), Wilber et al, 1999 ), Wilber et al, 1999 ), Wilber et al, 1999 • Surface Temperature Surface Temperature Surface Temperature Surface Temperature Surface Temperature Surface Temperature Surface Temperature Surface Temperature from from from from from from from from Valencia Anchor Station Valencia Anchor Station Valencia Anchor Station Valencia Anchor Station Valencia Anchor Station Valencia Anchor Station Valencia Anchor Station Valencia Anchor Station and mobile station (land use classification weighted) and mobile station (land use classification weighted) and mobile station (land use classification weighted) and mobile station (land use classification weighted) and mobile station (land use classification weighted) and mobile station (land use classification weighted) and mobile station (land use classification weighted) and mobile station (land use classification weighted) BRDF = k 0 + k 1 * f ross-thick + k 2 * f li-sparse GERB OBSERVATIONS GERB OBSERVATIONS CERES PAPS observations over VAS CERES PAPS observations over VAS Sample of CERES FM2 TOA Shortwave Radiances over the VAS. 1st August 2006, 10z. Sample of CERES FM2 TOA Shortwave Radiances over the VAS. 1st August 2006, 11z. Sample of CERES FM3 TOA Shortwave Radiances over the VAS. 12th February 2004, 13z. • BRDF: BRDF: calculated from the isometric, volumetric and geometric calculated from the isometric, volumetric and geometric Kernels of the MOD43B1 product for the Ross Kernels of the MOD43B1 product for the Ross- Thick Thick- Li Li- Sparse Sparse- Reciprocal model. Reciprocal model. Parameters from MODIS/Terra 1km resolution, SIN Grid, 1 Parameters from MODIS/Terra 1km resolution, SIN Grid, 1 Parameters from MODIS/Terra 1km resolution, SIN Grid, 1 Parameters from MODIS/Terra 1km resolution, SIN Grid, 1-16 day 16 day 16 day 16 day RMSE_SW = 4.8 W m -2 sr -1 RMSE_LW = 1.3 W m -2 sr -1 RMSE_SW = 4.1 W m -2 sr -1 RMSE_LW = 2.1 W m -2 sr -1 RMSE_SW = 14.9 W m -2 RMSE_LW = 2.7 W m -2 RMSE_SW = 8.6 W m -2 RMSE_LW = 7.8 W m -2 CERES simulations CERES simulations RMSE_SW = 4.3 W m -2 sr -1 RMSE_LW = 2.3 W m -2 sr -1 RMSE_SW = 17.5 W m -2 RMSE_LW = 7.1 W m -2 GERB simulations GERB simulations CONCLUSIONS CONCLUSIONS The methodology here shown is able to reproduce CERES and GERB T The methodology here shown is able to reproduce CERES and GERB T OA OA unfiltered radiances and fluxes under clear sky conditions. unfiltered radiances and fluxes under clear sky conditions. Simulated radiances reproduce the anisotropy of the radiance fi Simulated radiances reproduce the anisotropy of the radiance fi eld eld showing low showing low RMSEs RMSEs for clear sky conditions. for clear sky conditions. The inclusion of a higher resolution BRDF in the methodology ha The inclusion of a higher resolution BRDF in the methodology ha s s significantly improved the comparison between simulated and meas significantly improved the comparison between simulated and meas ured ured fluxes and will allow further studies over wider areas. fluxes and will allow further studies over wider areas. CERES dedicated PAPS observations over the VAS are of great val CERES dedicated PAPS observations over the VAS are of great val ue to ue to develop the methodology to validate low spatial resolution remot develop the methodology to validate low spatial resolution remot e sensing e sensing data and products. In this way, the methodology has been extende data and products. In this way, the methodology has been extende d and d and applied to GERB products. applied to GERB products. Classified Classified LANDSAT LANDSAT image image (5th (5th July July 2003): 11 2003): 11 categories categories for for the the Valencia Valencia Anchor Station Anchor Station area area (50 x 50 km (50 x 50 km 2 ) 1: Water, 2:Pine trees, 3: Low-density Pine trees,4: Shrubs, 5: Irrigated, 6: Vineyard, 7: Low-density vineyard, 8: Very low density, 9: Dry crops, 10: Bare soil, 11: Degraded Classified Classified LANDSAT LANDSAT image image (5th (5th July July 2003): 11 2003): 11 categories categories for for the the Valencia Valencia Anchor Station Anchor Station area area (50 x 50 km (50 x 50 km 2 ) 1: Water, 2:Pine trees, 3: Low-density Pine trees,4: Shrubs, 5: Irrigated, 6: Vineyard, 7: Low-density vineyard, 8: Very low density, 9: Dry crops, 10: Bare soil, 11: Degraded REFERENCES REFERENCES Gonzalez L., A. Hermans, S. Dewitte, A. Ipe, G. Sadowski, N. Clerbaux. Resolution Enhancement of GERB data, RMIB Technical Note: MSG-RMIB-GE-TN-0003 Key, J. and A. J. Schweiger (1998): Tools for atmospheric radiative transfer: Streamer and FluxNet, Computers & Geosciences, 24(5), 443-451. Loeb, N.G, S. Kato and B.A.Wielicki. Defining Top-of-Atmosphere Flux Reference Level for Earth Radiation Budget Studies. Journal of Climate, vol: 15, nº22, pp3301-3309. Velazquez, A. et al. (2006) Use of CERES PAPS observations over the Valencia Anchor Station to validate low spatial resolution remote sensing data and products. Proceedings of the 12th Conference on Atmospheric Radiation Wanner, W, X.Li, and A.H. Strahler (1995): On the derivation of kernels for kernel-driven models of bidirectional reflectance, J. Geophys. Res., 100, pp. 21077-21090 Wilber, A.C., D.P. Kratz, and S.K. Gupta (1999): Surface Emissivity Maps for Use in Satellite Retrievals of Longwave Radiation. NASA/TP-1999-209362 RESULTS RESULTS Collocation Collocation algorithm algorithm: CERES PAPS footprints vary in CERES PAPS footprints vary in location location and and size size from one to from one to another. another. To know the BRDF at CERES scale it will be necessary to To know the BRDF at CERES scale it will be necessary to convolve the Imager Surface Properties with the convolve the Imager Surface Properties with the CERES CERES Footprint Point Spread Function Footprint Point Spread Function δ δ δ and and and and β β β are the are the are the are the coordinates of a coordinates of a coordinates of a coordinates of a point in the FOV point in the FOV point in the FOV point in the FOV δ β δ β δ δ β δ β δ β δ d d P d d x P x FOV FOV ⋅ ⋅ ⋅ ⋅ ⋅ ⋅ = ∫ ∫ cos ) , ( cos ) , ( ) , ( The The collocation collocation algorithm must be able to handle with the algorithm must be able to handle with the shape of the footprints in PAPS mode and take into shape of the footprints in PAPS mode and take into account the scanning direction because of the asymmetry account the scanning direction because of the asymmetry of the CERES PSF. of the CERES PSF. The results are the The results are the weighted means weighted means in which the in which the weighting corresponds to the weighting corresponds to the PSF PSF By applying the BRDF model to the weighted means of the paramete By applying the BRDF model to the weighted means of the paramete rs we obtain a rs we obtain a BRDF BRDF for every CERES footprint for every CERES footprint and for the 7 MODIS narrow and for the 7 MODIS narrow- bands present in the MOD43 product, and for the 3 broad bands present in the MOD43 product, and for the 3 broad- bands bands First, we determine the limits of the First, we determine the limits of the CERES footprint and then we CERES footprint and then we locate locate the pixels from MODIS that are within the pixels from MODIS that are within the CERES footprint the CERES footprint 20 40 60 80 100 30 210 60 240 90 270 120 300 150 330 180 0 SW radiances, 12-Feb-2004, 13:05:04-13:11:53 UTC VZA (°) RAA ( ° ° ° ° ) 30 35 40 45 50 55 60 65 forward scattering direction backward scattering direction CERES TOA SW radiances and geometry. Radial axis CERES TOA SW radiances and geometry. Radial axis corresponds to VZA and corresponds to VZA and azimuthal azimuthal direction corresponds direction corresponds to RAA to RAA 20 40 60 80 100 30 210 60 240 90 270 120 300 150 330 180 0 SW radiances, 10-Feb-2004, 13:17:04 - 13:23:53 VZA (°) RAA ( ° ° ° ° ) 30 35 40 45 50 55 60 65 70 75 forward scattering direction backward scattering direction CERES TOA SW radiances and geometry. Radial axis CERES TOA SW radiances and geometry. Radial axis corresponds to VZA and corresponds to VZA and azimuthal azimuthal direction corresponds direction corresponds to RAA to RAA Radiative transfer simulations have been run every 15 min, from 7:00 until 17:00 UTC, for the 1st of August 2006 over the 4 selected GERB footprints that cover the study area. RMSE in SW radiances are of the same order as the obtained for CERES comparisons. However, GERB shows higher RMSE than CERES for the derived fluxes comparison, needing this problem further investigations.