Parcel-Level Land Cover and Soil Fertility Mapping using Multi-Sensor Satellite Imagery, Cadastral Overlays and Soil Health Card Records in a Semi-Arid Taluka of Saurashtra, Gujarat, India

DOI: 10.22161/ijeab.115.24

Abstract:
Reliable, spatially explicit information on land cover and soil fertility is a prerequisite for micro-level planning of land and water resources in rainfed, groundwater-dependent agricultural regions. This study demonstrates a workflow that links multi-sensor Indian Remote Sensing satellite data, digitised cadastral records and Soil Health Card (SHC) data for Amreli Taluka (83,112 ha), Saurashtra, Gujarat. CARTOSAT-1 panchromatic, IRS-P6 LISS-III and LISS-IV scenes were georeferenced and overlaid with village boundaries and survey-number polygons for seven randomly selected villages. A LISS-III scene of 31 January 2009 was classified with a maximum-likelihood supervised classifier using training signatures derived from a false colour composite, and the result was evaluated with a confusion matrix and the kappa coefficient. Nutrient status (N, P, K, organic carbon) and soil reaction (pH, electrical conductivity) from the SHCs were joined to cadastral polygons and mapped at survey-number level for Sanosara village. Dense, medium and poor agriculture occupied 28.27%, 22.24% and 14.23% of the taluka, respectively, which together with fallow land (16.81%) accounted for 81.55% of the area; habitation covered 12.01%, waste or eroded land 4.73% and water bodies 1.71%. The summary statistics reported with the matrix are an average accuracy of 83.08%, an overall accuracy of 79.32% and a kappa coefficient of 0.754 (standard deviation 0.00165; 95% confidence limit ±0.00324). SHC records showed predominantly medium nitrogen, low phosphorus and high potash, with normal pH (6.5-8.2) and electrical conductivity below 1 dS m-1. The cadastral-level approach converts farm-scale soil data into mappable, plot-specific recommendations, but accuracy reporting, class design and sampling density need strengthening before the maps are used operationally. The workflow is transferable to other talukas where SHC and cadastral data exist.

Keywords:
IRS-P6 LISS-III, Maximum likelihood classification, Micro-level planning, Soil fertility mapping

Article Info:
Received: 23 Sep 2026; Received in revised form: 19 Oct 2026; Accepted: 25 Oct 2026; Available online: 31 Oct 2026

Cite this Article: