Duo satellite-based remotely sensed land surface temperature prediction by various methods of machine learning

The entire world is witnessing challenges in climate change and control catastrophes. Addressing and investigating sustainable climate and urban thermal harmony can be done by getting deeper into pattern and anomalies identification in land surface temperature studies. In this paper, we have discussed duo satellite-based remotely sensed land surface temperature prediction using various methods of machine learning in the western Indian smart city, Ajmer from data range 2003-2021. Quad LST datasets (two for daytime and two for nighttime) from MODIS sensor on board Aqua and Terra satellites were incubated in this study. In the field of remote sensing, neural networks and the essence of artificial intelligence are now being smelled in a diverse assortment of applications, and regular new application proposals are being made. In the present study, machine learning techniques are used to predict the temperature of the same day and future day. To evaluate the accuracy of the predictions made by each method, the R-2 score and RMSE were calculated. R-2 score is a statistical measure that indicates the proportion of the variance in the dependent variable that is predictable from the independent variable(s). A score of 1 indicates a perfect fit, while a score of 0 indicates that the model does not explain any of the variability in the data. RMSE (Root Mean Square Error) is a measure of the difference between the predicted values and the actual values. By calculating these metrics, the researchers and urban dwellers are able to determine the accuracy of the different Machine Learning methods. These findings and future developments can be used in land surface temperature prediction in different contexts such as quality and quantity of available different satellite data, duration, and study area in climate studies, earth and atmospheric sciences.