Land Suitability Assessment for Cotton Cultivation - A Case Study of Kumkurgan District, Uzbekistan

Abstract

Optimal land allocation for cotton cultivation requires more expert knowledge since we face increasing land degradation and climate change impact in Uzbekistan over the last years. In addition, because of not taking into account important agro-ecological conditions, crop requirements and crop rotation systems, land productivity is declining year by year. But, how can decision making on land allocation for cotton cultivation be supported? Here, a land suitability assessment models which allow to integrate and weight different spatial data (i.e., criteria) in a Geographic Information System (GIS) can be considered a promising tool. In this research, a land suitability assessment model was implemented on one of the arid areas of Uzbekistan (Kumkurgan district of Surkhandarya region) to identify and classify the suitability of lands for cotton cultivation. In terms of classifying land suitability for cotton cultivation, the Suitability Modeler functionality of ArcGIS Pro 2.7 was utilized. Overall, nine input variables related to important criteria of cotton cultivation such as soil texture, soil salinity, irrigation and collector-drainage networks, groundwater level/ mineralization, slope, current land use and temperature were used for a complex analysis of the area. After combining the data, the result was classified into five suitability classes based on the FAO land suitability classification method. The result delivered show that for Kumkurgan district the following: highly suitable (1.8% of the area; 36.2 km²), moderately suitable (11.5% of the area; 230.5 km²), marginally suitable (7.25% of the area; 145.0 km²), currently not suitable (24.3% of the area; 485.8 km²) and permanently not suitable (55.09% 55.09% of the area; 1101.1 km². Irrigated agricultural land parcel data with vector/shapefile which includes five years of cropping sequences were used for crop rotation system of 2:1 (cotton: wheat/other) to identify exact land parcels for a cotton seeding in the season of 2020-2021 and validating the results. As a result, about 78.5 km² of the agricultural area was assessed as very suitable for cotton for the 2020-2021 season. Validation revealed an accuracy of the model at 78% with the current state of cotton fields.