Remote sensing albedo variability over an arctic Siberian tundra site

Laboratory of Geo-Information Science and Remote Sensing
  Education
  Research
  Publications
  Models
  News & Calendar
  News
  Calendar
  Archive
  News
  Calendar
  2011
  2010
  2009
  2008
  2007
  2006
  2005
  2012
  Staff
  Equipment
  Contact details
  Workshops

12 Dec 2007 09:00 - 12 Dec 2007 09:30
Unit: Laboratory of Geo-Information Science and Remote Sensing
Location: GAIA 1
Organisation: Wageningen University

 By Martin Claverie  

Abstract:
" Land surface albedo is one of the key drivers of the Earth’s surface energy budget. Over high latitude, the Surface albedo feedback is an important component of the climate models. This study leads to understand the seasonal shape of the satellite inferred albedo over an arctic Siberian tundra. MODIS (Moderate Resolution Imaging Spectroradiometer) is a spectral and angular sensor, which provides various atmospheric and surface products. Albedo is derived from MODIS reflectance data, based on Bidirectional Reflectance Distribution Function (BRDF) model. It is composed by MCD43A1 (BRDF parameters) and MCD43A3 (DHR and BHRiso ). The analysis is carried out through three components: analysis of the data quality, analysis of the seasonal shape of the MODI albedo products, and comparison of MODIS data with ground measurements.
MODIS albedo data are provided with a so-called quality flag data (MCD43A2). It informs users about the algorithm used to derive BRDF parameters. The full inversion algorithm (QF 0) is based only on surface reflectance data, whereas the magnitude inversion algorithm (QF 1) includes a priori knowledge and assumptions in the form of the BRDF function. Throughout the year, more than 50% of the data are derived from the magnitude inversion algorithm. The analysis of those data leads to a deviation from full inversion ones, inferior to 0.05 point of albedo. This good accuracy allows to use data from both algorithms for the rest of the study.
Seasonal shape of albedo is mainly driven by snowmelt. Vegetation has influence during the summer over visible and NIR bands. DHR, BHRiso, and isotropic parameter are very similar. Their differences depend mainly on the solar zenith angle. Volumetric and geometric component informs about the vegetation development. Two objects (lakes and shrubland) have been selected on this tundra area. During the spring, the lakes have a slower snowmelt compared to the vegetation patterns. In the summer, the shrubland is characterized by an increase of the albedo mainly in the NIR band.
Snow height data is compared with MODIS albedo data. Over the visible and NIR bands, DHR, BHRiso and isotropic parameter have a high correlation with the snow height (around 0.9). Volumetric and geometric parameters are not correlated with this data over those bands. Concerning the MIR bands, the snow height is anti-correlated (around -0.6 for 1.2 band and -0.8 for 2.1 band) with the five components.
The validation is done with 2003 in-situe albedo measurements. Because of a lack of data, only six validation points are used. The accuracy of MODIS albedo data is 0.05 for all data, and 0.02 when the analysis is restricted to full inversion data. "
Print this activity