资源环境科技发展态势分析平台

Precise characterization of surface air temperature at different spatial scales remains a challenge as it often requires a dense network of temperature sensors. The most common approach for estimating surface air temperature at a desired location in hydrological and terrestrial models is to interpolate surface air temperature observations obtained from a sparse set-up of temperature sensors using the global standard uniform lapse rate. This interpolation technique, though easy to use, can lead to unreliable results, as the global standard uniform lapse rate may not be representative of seasonal and local variations in lapse rate. In this study, linear and nonlinear regression relationships are derived to estimate mean minimum, mean and mean maximum near surface air temperatures (at monthly and annual time scales) across India. The normal daily air temperature data from 439 stations and their corresponding latitude, longitude and elevation are used in the analysis. The computed temperature gradients exhibit seasonal variation. The effects of elevation and longitude are more pronounced in pre-monsoon (March-May) and post-monsoon (October-November) months, whereas the effects of latitude are more in winter months (December-February). Results indicate that the global standard uniform lapse rate (-0.65 degrees C 100m(-1)) is only applicable for monthly mean maximum temperature for April and May. The highest interpolation reliability is obtained for monthly mean minimum temperature. The regression equations though simple, hold promise for describing surface air temperature across India.