Development of high spatial resolution weather data using daily meteorological observations over Indian region

Finer spatial resolution interpolated weather data is essential to enable utilization of satellite-based images in studies related to crop growth dynamics, etc. Satellite data are available daily at 1 x 1 km or at the most within 5 x 5 km grid. To make the weather data timely available at the same spatial scale, the procedure has been developed to generate the spatially interpolated weather data product over India. Daily weather data (minimum & maximum temperature and rainfall) available at point scale on India Meteorology Department web site have been used in this study. A semi-automated user interactive Graphical User Interface (GUI) has been developed which quality checks the temperature data sets by filling the missing data sets as well as providing a platform to correct erroneous data which have been identified using statistical methods taking spatial as well as temporal incompatibility into account. Daily spatially interpolated product is generated in image form using thin plate spline interpolation technique that uses the quality checked weather data as well as elevation information from CARTODEM data in order to account for effect of elevation on temperature. The validation was performed using "Jack-knife testing method" for three different seasons i.e., monsoon, summer and winter. The mean absolute errors for decadal averaged products were found to vary within 1.2-1.5 degrees C for maximum temperature, 1.1-1.7 degrees C for minimum temperature, 1.0-7.0 mm for rainfall considering all seasons with higher error observed in monsoon for maximum temperature and rainfall and in winter for minimum temperature. It was found that errors were close to 1 degrees C for stations with elevation less than 550 m whereas in central portion of India, mean absolute errors were found to be less than 1 degrees C.