Soil erosion: assessing riverbank erosion dynamics in Jorhat district of Assam, India using revised MMF model and GIS...

The assessment of soil and nutrient erosion along the scattered Brahmaputra River in the Jorhat district of Assam has been attempted utilizing field and laboratory assessments, satellite remote sensing techniques, and the Revised Morgan-Morgan-Finney (RMMF) soil erosion model. River portions were mapped based on visually interpreted and geo-coded false colour composites (FCC) of Landsat-9 OLI data and Survey of India toposheets (1:50,000). Soil samples were collected from 62 different sites within a region of 86.78 km² on the banks of the Jorhat River. The main findings revealed that approximately 55.24% of the study region was inundated at least once during the flood period from May to July 2023. The study area showed total sand content between 26% and 84%, silt 5%-48%, clay 6%-36% and organic matter content 0.20%-2.08%. Erodibility indices like Clay Ratio (mean 4.91), Modified Clay Ratio (mean 4.67), Dispersion ratio (mean 0.19), Erosion Ratio (mean 0.11) and Erosion Index (mean 0.20) indicated that most soils in the study area were susceptible to erosion. Using the Revised Morgan-Morgan-Finney model, annual soil loss was estimated, with runoff transport capacity ranging from 5.16 to 164.68 t ha⁻¹ (mean: 37.27 t ha⁻¹), and total annual detachment ranging from 11.56 to 129.14 t ha⁻¹ (mean: 77.59 t ha⁻¹) and transport capacity was considered as estimated soil loss as it was lower than total detachment. Pearson's correlation matrix shows that aggregate stability had negative association with sand fraction and positive associations (p <  = 0.05) with clay and organic matter. Also, hydraulic conductivity (HC), water holding capacity (WHC) and available water (AW) demonstrated negative (p <  = 0.05) associations with soil loss. The total fertile soil loss was estimated as 322,668 tons annually, with annual losses of organic carbon, total nitrogen, total phosphorus, and total potassium estimated at 1,444.32 tons, 23.52 tons, 227.15 tons, and 4,710.64 tons, respectively. Change detection (from 1986 to 2022) was estimated based on various land use patterns, where riverine sand, built-up land, cropland showed percent increase as 61.16%, 52.75% and 14.38% respectively. But, negative percent change of 73.37%, 12.72% and 32.65% were showed against Land use/land cover (LULC) classes of barren land, river/waterbodies and grassland respectively. Hence, expanding cropland and bare sand on coarse-textured, low organic matter soil drove the extreme soil and nutrient erosion in the study area. This study underscores the potentiality of combining the Revised MMF model with geospatial analysis to effectively identify and quantify soil erosion in the study area.