NOAA AVHRR land surface albedo algorithm development

TitleNOAA AVHRR land surface albedo algorithm development
Publication TypeJournal Article
Year of Publication1997
AuthorsToll, DL, Shirley, D, Kimes, DS
JournalInternational Journal of Remote Sensing
Pagination3761 -3796
Accession NumberKNZ00611

The primary objective of this research is to develop a surface albedo model for the National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR). The primary test site is the Konza prairie, Kansas (U.S.A.), used by the International Satellite Land Surface Climatology Project (ISLSCP) in the First ISLSCP Field Experiment (FIFE). In this research, high spectral resolution field spectrometer data was analyzed to simulate AVHRR wavebands and to derive surface albedos. Development of a surface albedo algorithm was completed by analysing a combination of satellite, field spectrometer, and ancillary data. Estimated albedos from the field spectrometer data were compared to reference albedos derived using pyranometer data. Variations from surface anisotropy of reflected solar radiation were found to be the most significant albedo-related error. Additional error or sensitivity came from estimation of a shortwave mid-IR reflectance (1.3-4.0 mum) using the AVHRR red and near-IR bands. Errors caused by the use or AVHRR spectral reflectance to estimate both a total visible (0.4-0.7 pm) and near-IR (0.7-1.3 mum) reflectance were small. The solar spectral integration, using the derived ultraviolet, visible, near-IR and SW mid-IR reflectivities, was not sensitive to many clear-sky changes in atmospheric properties and illumination conditions