GIS010 – Prediction of High Spatio-Temporal Resolution Land Surface Temperature using ANN and MICROWAVE VEGETATION INDEX


Land Surface Temperature (LST) with high spatio-worldly goals is sought after for hydrology, environmental change, nature, urban atmosphere and ecological investigations, and so forth. Moderate Resolution Imaging Spectroradiometer (MODIS) is a standout amongst the most usually utilized sensors attributable to its high spatial and fleeting accessibility over the globe, yet is unequipped for giving LST information under shady conditions, bringing about holes in the information. Interestingly, microwave estimations have an ability to enter under mists.

The current investigation proposes a technique by investigating this property to foresee high spatio-worldly goals LST under shady conditions amid daytime and evening time without utilizing in-situ LST estimations. To accomplish this, Artificial Neural Networks (ANNs) based models are utilized for various land cover classes, using Microwave Polarization Difference Index (MPDI) at better goals with subordinate information. MPDI was determined utilizing resampled (from 0.25° to 1 km) brilliance temperatures (Tb) at 36.5 GHz channel of double polarization from Advance Microwave Scanning Radiometer (AMSR)- Earth Observing System and AMSR2 sensors.

The proposed procedure is tried over Cauvery bowl in India and the execution of the model is quantitatively assessed through execution estimates, for example, relationship coefficient (r), Nash Sutcliffe Efficiency (NSE) and Root Mean Square Error (RMSE). Results uncovered that amid daytime, AMSR-E(AMSR2) determined LST under clear sky conditions compares well with MODIS LST bringing about estimations of r extending from 0.76(0.78) to 0.90(0.96), RMSE from 1.76(1.86) K to 4.34(4.00) K and NSE from 0.58(0.61) to 0.81(0.90) for various land cover classes. Amid evening, r esteems ran from 0.76(0.56) to 0.87(0.90), RMSE from 1.71(1.70) K to 2.43(2.12) K and NSE from 0.43(0.28) to 0.80(0.81) for various land cover classes. RMSE esteems found between anticipated LST and MODIS LST amid daytime under clear sky conditions were inside satisfactory breaking points. Under overcast conditions, aftereffects of microwave determined LST were assessed with air temperature (Ta) and demonstrate that the methodology performed well with RMSE esteems lesser than the outcomes acquired under clear sky conditions for land cover classes for both day and evenings.

BASE PAPER:  Prediction of High Spatio-Temporal Resolution Land Surface Temperature using ANN and MICROWAVE VEGETATION INDEX

Leave a Reply

Your email address will not be published. Required fields are marked *