Duo Satellite Based Surface Temperature Comparative Study of Jaipur City Using Soft Computing

In this paper, the remotely sensed land surface temperature distribution from two satellites i.e., Aqua and Terra of Jaipur city were used in back propagation neural network system to model and predict the thermal activity. As an input dependent variable TDLST and TNLST was fed to BPNN and showcased the regression of independent variable ADLST and ANLST by keeping constant initial weight, bias and test and training sets. In day-day LST model R
²
values for training and testing are 0.9827 and 0.9012 and in night-night LST model R
²
values for training and testing are 0.9812 and 0.9270 respectively. Model prediction results showed model ability to predict reliable results when compared with the collected data.