Use of AVHRR-derived surface temperatures in evaluating a land–atmosphere model

Abstract

The vertical transports of heat and moisture at the Earth's surface provide a critical linkage between meteorological, hydrological and ecological processes. Unfortunately, estimates of energy and water fluxes by land surface–atmosphere models are difficult to validate. This is due to the fine spatial and temporal resolution of the models, as well as a lack of widespread observations of these fluxes and other components of the surface energy budget. However, modelled fluxes of energy and water are often based largely on accurate determination of the modelled soil surface temperature. Land surface temperatures (LSTs) derived from the Advanced Very High Resolution Radiometer (AVHRR) onboard polar-orbiting satellites can be determined over large spatial extents a few times per day. They provide a means by which modelled surface temperatures, and thus modelled fluxes, can be evaluated. In this paper, two versions of the Simulator for Hydrology and Energy Exchange at the Land Surface (SHEELS) model are evaluated using AVHRR-derived LSTs. Model vegetation cover is held constant in the first version, whereas Normalized Difference Vegetation Index (NDVI) data are used to temporally modify vegetation cover in the second version. Inclusion of the NDVI data is found to decrease the bias between modelled and satellite-derived LSTs by 0.5 K. However, even with this improvement in vegetation parametrization, bias remains large (3.7 K).