Revealing Bare-Soil Texture using GLCM and Multispectral Indicators: Applications to Sustainable Agricultural Land-Use Planning in Enrekang Nurfadila JS 1*, Syaeful Rahmat 2, Rismaneswati 3
1. Agricultural Sciences Study Program, Post-Graduate School, Hasanuddin University, Makassar, Indonesia
2. Remote Sensing and Geographic Information Systems Study Program, Faculty of Vocational Studies, Hasanuddin University, Makassar, Indonesia
3. Soil Science Study Program, Faculty of Agriculture, Hasanuddin University, Makassar, Indonesia
Abstract
Remote sensing provides a scalable pathway to characterize soil texture in data-sparse tropical landscapes, but the signal is indirect and highly conditioned by surface state. This study develops a PlanetScope based workflow to reveal bare-soil texture patterns in Enrekang for sustainable agricultural land-use planning. We construct dry season, low vegetation composites to isolate bare soil, apply atmospheric and topographic corrections, and compute Gray Level Co-occurrence Matrix (GLCM) metrics on red-edge and near-infrared channels within moving windows. Texture features are fused with multispectral soil indicators and terrain covariates to enhance separability under variable illumination and micro relief. Field and laboratory observations guide model development and spatially aware validation. Qualitatively, the resulting maps delineate coherent zones aligned with landform and management gradients, distinguishing five dominant texture classes without relying on quantitative thresholds: sandy clay loam, silt loam, clay, silty clay loam , and clay loam. Red-edge and NIR texture metrics consistently emerge as the most informative cues under bare soil, dry conditions, while terrain variables stabilize transitions across slope positions. Principal limitations residual vegetation, surface moisture, iron oxide coatings, and tillage imprints are mitigated through strict masking and multi-date compositing. The workflow is repeatable, low cost, and transferable, offering decision support for site selection, crop and tillage choices, irrigation prioritization, erosion control, and targeted ground sampling across smallholder agroecosystems.
Keywords: GLCM, PlanetScope, soil texture, bare soil, Enrekang
