• Pallab Kumar Das

The spatiotemporal trade-off associated with satellite-derived datasets necessitates the downscaling of Land Surface Temperature (LST) to understand temperature dynamics and related phenomena at a finer level. This study employs a Support Vector Regression (SVR) model to enhance the spatial resolution of Moderate Resolution Imaging Spectroradiometer (MODIS) LST from a coarse resolution of 1 kilometer to a finer scale of 90 meters over an approximately 2,500 km2 area in Haryana, India. The model was trained on datasets from three distinct dates representing three different months (March, April, and May) to refine the MODIS LST and assess its predictive capability. The results demonstrated the model’s strong downscaling efficacy with high Coefficient of Determination (R2) and low Root Mean Square Error (RMSE) values, indicating the feasibility of learning cross-scale relationships

Spatiotemporal trade-off, downscaling, land surface temperature, support vector regression, MODIS, Landsat, spatial resolution

"Spatial Refinement of MODIS Land Surface Temperature from 1,000 m to 90 m Using Machine Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.4, Issue 11, page no.160-166, November-2017, Available :http://www.jetir.org/papers/JETIR1711254.pdf

"Spatial Refinement of MODIS Land Surface Temperature from 1,000 m to 90 m Using Machine Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.4, Issue 11, page no. pp160-166, November-2017, Available at : http://www.jetir.org/papers/JETIR1711254.pdf