Surface Reflectance Retrieval using the AATSR-ADV Dual View Algorithm.
Sogacheva, Larisa1; Kolmonen, Pekka1; Virtanen, Timo H.1; Rodriguez, Edith1; Sundström, Anu-Maija2; de Leeuw, Gerrit1
1Finnish Meteorological Institute, FINLAND; 2University of Helsinki, FINLAND

Global satellite observations of land properties are important to Earth system science and global climate research. Land surface albedo, defined as the ratio of the incoming and reflected solar radiation at the surface, is one of the key geophysical variables controlling the surface radiation budget. Snow-free albedo is especially important for land surface models that compute the exchange of energy, water, carbon for various land use categories. In this study, a new algorithm was developed for the retrieval of surface reflectance, using the aerosol optical depth retrieved from the Advanced Along-Track Scanning Radiometer (AATSR).

AATSR on board the European Space Agency (ESA) satellite ENVISAT covers, together with ATSR-2, the 17-years period of observations (1995-2012). (A)ATSR measures top-of-the-atmosphere (TOA) radiances at 7 wavelengths in the spectral range from the visible to the thermal infrared. (A)ATSR has two views, one at nadir and the other one at 55o forward view. (A)ATSR has a conical scan covering a swath of 500 km. The (A)ATSR resolution is 1 km at nadir.

Designed primarily to measure Sea Surface Temperature (SST), the radiometer is also successfully used for the Aerosol Optical depth (AOD) retrieval. Several AOD retrieval algorithms have been developed to achieve this. The AATSR Dual View (ADV) algorithm used at Finnish Meteorological Institute in Helsinki was originally developed by Veefkind et al. (1988). The current version of AATSR-ADV algorithm is described in Kolmonen et al. (2012). Using the dual view, the surface contribution is eliminated by using the K-ratio (ratio between forward and nadir TOA reflectance at 1.6µm).
Since the surface reflectance is eliminated, the AOD is retrieved as an independent parameter and can be used as a contribution to the atmospheric correction needed to compute the surface reflectance from the measured TOA radiance. The determination of surface relectance is executed after AOD retrieval. The atmospheric reflectance due to aerosols (and gases) is subtracted from the measured TOA reflectance taking into account also the multiple scattering between surface and atmosphere. The end result is the instantaneous surface reflectance. Although the surface reflectance is a derived result (i.e. it is computed after aerosol retrieval) it is semi-independent retrieval product since the assumptions about surface during the aerosol retrieval are not connected to the value of the surface reflectance itself.

The computed surface reflectance was validated against the spectral surface bidirectional reflectance and albedo data available from the AERONET-based Surface Reflectance Validation Network (ASRVN). The ASVRN data set is based on Moderate Resolution Imaging Spectroradiometer (MODIS) TERRA and AQUA data (Wang et al., 2011). The preliminary results show good (R=0.9) correlation between AATSR and ASRVN derived surface reflectances.

The method can be applied to the Sentinel-3 Sea and Land Surface Temperature Radiometer (SLSTR), which is a dual-view Earth observing instrument that builds on the heritage of the Along Track Scanning Radiometer (ATSR) series.

Acknowledgement: This work contributed to the ESA Climate Change Initiative project Aerosol-CCI

Kolmonen, P., Sundstroem, A.-M., Sogacheva, L., Rodriguez, E., Virtanen, T. and de Leeuw, G. (2012). The uncertainty characterization of AOD for the AATSR ADV/ASV retrieval algorithm -Towards the assimilation of the satellite retrieved aerosol properties. Submitted to Atmos. Meas. Tech.
Veefkind, J.P., de Leeuw, G., A new algorithm to determine the spectral aerosol optical depth from satellite radiometer measurements, J. Aerosol Sci., 29, 1237-1248, 1998.
Wang, Y, J Czapla-Meyers, A Lyapustin, K Thome and E. G. Dutton (2011), AERONET-based surface reflectance validation network (ASRVN) data evaluation: Case study for railroad valley calibration site, Review of Scientific Instruments, 115(10), p. 2710.