Investigating the effects of Changing Land Use/Land Cover on Curve Number, Land Surface Temperature, and Vegetative Cover in the Ong Sub-Basin of the Mahanadi River, India

Abstract

Over the past few decades, the Curve Number (SCS-CN or CN) approach has been employed to estimate direct surface runoff. In the times of urban sprawl, rapidly growing socioeconomic anthropogenic activities, and environmental changes all have a rapid growth that leads to spatial and temporal variability in the land use/land cover (LULC) complex which in a way affects the direct surface runoff, LST, and Vegetative cover. The study utilized the pixel based (PB), object directed (OD) machine learning algorithm i.e., Random Forest (RF) classifier for the LULC change study of the Google Earth Engine (GEE) platform for the verification of the SCS-CN, LST, and vegetative cover variability over 4 decades from 1980 to 2020 of the Ong River sub-basin (area = 4650 sq. km) of the Mahanadi River Basin of India. Sentinel-2 and Landsat satellite products were processed and utilized to conduct the LULC change analysis. The Kappa Coefficients of LULC maps for each decade from 1980 to 2020 equaled 0.86, 0.90, 0.891, and 0.895 with overall accuracy percentages of 97.89, 96.16, 96.79, and 96.44, respectively. The study determined the associated effects of each LULC class (i.e., built-up areas, Barren land, Water Bodies, Cropland, and Forest) on CN variability, LST, and vegetation index i.e., Normalized difference vegetation index (NDVI). The CN values varied from 64 to 78 in the 4 decades, suggesting the effects of decrease in forest cover and an increase in the built up. The LULC change analysis revealed a considerable decrease in the forest area from 1864.8 sq. km to 1098.34 sq. km and a sizeable increment in built-up area from 123.3 sq. km to 458.9 sq. km.  Furthermore, the study investigated and correlated each LULC class with the LST and NDVI. LST and NDVI for the forest and built-up areas were correlated with R2 equal to 0.91 and 0.72, respectively. Overall, the results suggest considering dominating LULC class change led to a rise in the average LST of the watershed from 28.06 °C to 28.68 °C. The increase in the average LST and increasing Curve Number value is a consequence of the scanty forest area combined with the reduction in the greenness of vegetative cover due to an increase in the built-up area indicating the alteration in the land-use patterns, land management practices, and streamflow due to anthropogenic activities. This type of study is helpful for watershed planning and management.