Estimation of surface air temperature over central and eastern Eurasia from MODIS land surface temperature

Surface air temperature (T-a) is a critical variable in the energy and water cycle of the Earth-atmosphere system and is a key input element for hydrology and land surface models. This is a preliminary study to evaluate estimation of T-a from satellite remotely sensed land surface temperature (T-s) by using MODIS-Terra data over two Eurasia regions: northern China and fUSSR. High correlations are observed in both regions between station-measured T-a and MODIS T-s. The relationships between the maximum T-a and daytime T-s depend significantly on land cover types, but the minimum T-a and nighttime T-s have little dependence on the land cover types. The largest difference between maximum T-a and daytime T-s appears over the barren and sparsely vegetated area during the summer time. Using a linear regression method, the daily maximum T-a were estimated from 1 km resolution MODIS T-s under clear-sky conditions with coefficients calculated based on land cover types, while the minimum T-a were estimated without considering land cover types. The uncertainty, mean absolute error (MAE), of the estimated maximum T-a varies from 2.4 degrees C over closed shrublands to 3.2 degrees C over grasslands, and the MAE of the estimated minimum T-a is about 3.0 degrees C.