Comparing Water Indices for Landsat Data for Automated Surface Water Body Extraction under Complex Ground Background: A Case Study in Jilin Province

Derived from Landsat imagery and capable of enhancing the contrast between surface water bodies background, indices are widely used in body extraction. Whether one index its optimal threshold can maintain best all year round is a question. At present, however, little research has considered effect time or conducted experiments with data different months. To identify for extraction, two regions Jilin Province were selected case study comprehensive comparative analysis considering acquisition was conducted. Ten calculated based on 30 m spatial resolution TM/OLI acquired 1999 2001 2019 2021 May to October. The included Modified Normalized Difference Water Index (NDWI3 MNDWI), Automated Extraction (AWEI) images without shadow, Multi-Band (MBWI), New (NWI), Ratio (WRI), Sentinel-2 (SWI) originally imagery, Comprehensive (NCIWI), Surfaces (IWS), Enhanced (EWI). OTSU algorism adopted adaptively determine segmentation each compared terms inter-class separability, sensitivity, robustness, extraction accuracy. result showed that NWI EWI performed months years, enhancement could suppress background information, especially water-related land use types cloud pollution. Their thresholds throughout period more stable than others, ranges −0.342 −0.038 −0.539 −0.223, respectively. Based thresholds, they achieved overall accuracies 0.952 0.981 0.964 0.981, commission errors 0 28.2% 7.7%, omission 15% 8%, kappa coefficient above 0.8 indicating good results. results demonstrated effectiveness combined algorithm better monitoring during periods offering reliable Even though this only focuses lakes within regions, have potential accurately over other regions.