Abstract

This study aimed to evaluate the efficiency of smart agriculture technologies, particularly unmanned aerial vehicles (UAVs), in surveying and classifying desert soils in Karbala Governorate, Iraq. The study area is located in the western desert of Karbala between longitudes 43°51′29″–43°59′08″ E and latitudes 32°34′13″–32°35′41″ N, covering approximately 10,000 ha. High-resolution aerial images were acquired using a UAV at an altitude of 200 m and compared with Sentinel-2 satellite imagery acquired on 21 April 2024 during the same period. Geometric correction was applied to both datasets, and the images were used as base maps for soil survey and classification at the soil series level. Supervised and unsupervised classification techniques were applied to the UAV and satellite images. UAV imagery achieved a spatial resolution of 0.04 × 0.04 m, whereas Sentinel-2 imagery provided a resolution of 10 × 10 m, demonstrating the superior spatial accuracy of the UAV data. The results revealed that compared with conventional satellite-based approaches, UAV-based soil surveys offer higher accuracy, faster implementation, and lower costs. A total of twelve distinct soil series were identified and mapped, all belonging to the Aridisols soil order. This study confirms the effectiveness of UAV technology as a reliable tool for detailed desert soil surveys and classification in arid environments.