Geomatic techniques for precise Dayas detection in arid zones: a case study in Northwestern Wilaya of Naama, Algeria

Arid zones are characterized by limited rainfall and scarce surface water, making them highly susceptible to droughts and water scarcity. However, temporary water bodies known as “Dayas” play a crucial role in providing surface water in such regions. These Dayas often serve as the sole water source, supporting local communities and biodiversity. This study aims to identify the most effective index for extracting Dayas by evaluating seven water indices on the Google Earth Engine (GEE) platform. Using a 30 m spatial resolution Landsat-5 Thematic Mapper (TM) imagery from October 2008, indices including the Modified Normalized Difference Water Index (MNDWI), the Augmented Normalized Difference Water Index (ANDWI), the Automated Water Extraction Index: for images with shadow (AWEIsh) and without shadow (AWEInsh), the Water Index (WI₂₀₁₅), the New Water Index (NWI), and the Enhanced Water Index (EWI) were computed and compared using the Otsu thresholding method. MNDWI and EWI exhibited superior performance, achieving overall accuracies of 94% and 92%, respectively, with a kappa coefficient exceeding 0.8, signifying accurate Dayas identification. This study focuses on the Wadi Ateuch watershed; however, the identified indices have the broader potential for precise Dayas detection and monitoring in diverse arid regions.