Global surface temperature for climate studies using NOAA-AVHRR data

Surface temperature is a good indicator of the energy balance at the Earth´s surface and is a key parameter in the physics of land-surface processes on a regional as well as global scales. It is a result of surface-atmosphere interactions and energy fluxes between the atmosphere and the ground. Therefore it is required for a wide variety of climate, hydrologic and biogeochemical studies. A long term data set of satellite-derived land-surface temperature, such as that from NOAA-AVHRR can be used to answer questions about possible climate change. Numerous split window equations have been developed. Most of these algorithms have an accuracy of +/- 2 K, which is far from the 1 deg. K error required for land surface studies. Based on Becker and Li´s [1995] evaluation, the algorithms developed by Ulivieri et al. [1994] and by Sobrino et al. [1997] were selected for the land split window in this study, because of their simplicity and accuracy. The sea split window equation proposed by Njoku [1985], and intensively used to assess sea surface temperature was also selected. Using the MODTRAN algorithm, a data set representative of different AVHRR acquisition geometry and atmospheric properties was generated for each NOAA satellite, NOAA 7 to NOAA 14 using the filter function specific to each satellite. Three sets of coefficients were generated for the split window equations, one for the sea and dense vegetation, and two for land surfaces. The Pathfinder II data set was used to generate global surface temperature maps and BOREAS data were used to evaluate and validate our surface temperature assessments