A MACHINE LEARNING APPROACH FOR SPATIAL SUITABILITY MODELLING FOR COTTON CULTIVATION IN NORTH-EAST BALUCHISTAN-PAKISTAN

Abstract

Over the past few years, cotton cultivation in Pakistan has witnessed a decline, with farmers increasingly turning to sugarcane and maize. This shift has had detrimental effects on the local textile industry. To revive cotton farming in the region, this study sought to identify vacant, naturally available land suitable for cotton cultivation. The research focused on the suitability of cotton cultivation in Northeast Baluchistan, a challenging terrain characterized by extreme slopes and mountainous areas. The data were processed in Google Earth Engine (GEE) and for different parameters i.e. Land use and Landcover (LULC), Normalized Difference Vegetation Index (NDVI), and Normalized Difference Moisture Index (NDMI), Rainfall, Temperature, Geology, Slope, DEM, Soil pH and Soil texture. Different satellite data were used to measure the parameters in the study area. Then all the parameters were assigned different weights and the Weighted Overlay Method was applied from the Spatial Analyst tool in Arc Map and suitability modelling was done. The study revealed that Barkhan in Northeast Baluchistan exhibited the highest suitability for cotton cultivation, covering 0.71% of the area. On the other hand, approximately 47.86% of the region displayed suitability despite restrictions on cultivation, including well-furnished drainages and irrigation and important infrastructures. The remaining land was not suitable for cotton cultivation in the northeast of Baluchistan. Overall, this publication offers information on land suitability for cotton farming in Northeast Baluchistan and creates a path for the discovery of feasible cotton farming.