The popular Soil Conservation Service-Curve Number (SCS-CN) method is widely used for direct surface runoff estimation from a given amount of rainfall in a watershed. The present urban sprawl, socioeconomic anthropogenic activities, and environmental changes all have affected the cosmic extent of land use-land cover (LULC) complex and climate, both spatially and temporally, which directly affect the parameter curve number (CN) and, in turn, the direct surface runoff. Therefore, the study propels the disparity of representative CNs of SCS-CN methodology, which is usually derived from NEH-4 tables based on land use and soil type (CNLU−ST) and from the observed rainfall(P)-runoff(Q) events (CNP−Q). The annual series of CNP−Q and CNLU−ST (from 1980 to 2020) showed the existence of trends and the inconsistency between CNP−Q and CNLU−ST for the Ong River basin (India). The land use and land cover (LULC) alteration analysis utilized the supervised machine learning algorithm and indicated two major LULC classes as the contributing factors for increasing CNs. Furthermore, the study attributes the implications of shifting LULC dynamics (~ 70%) and climate variations (~ 30%) to the watershed. Employing Aridity Index (AI), as a parameter in solving the disparity for representative CNs for annual/decadal values revealed strong evidence with a fit of high R2 range (0.72, 0.99) of LULC and aridity influencing CNs.