Assessing the surface water area dynamics of Chahannur National Wetland Park with Landsat and Sentinel-2 multi-temporal imageries

In semi-arid inland basins, the surface water area of lacustrine wetlands is a key indicator for monitoring watershed hydrological processes and ecosystem health. Understanding its spatiotemporal dynamics is essential for regional lake conservation and ecological civilization initiatives. However, the inherent spatiotemporal limitations of single-sensor remote sensing systems hinder accurate monitoring of rapidly changing seasonal lakes. To address this, this study developed an innovative multi-sensor fusion framework leveraging the Google Earth Engine (GEE) cloud computing platform. By integrating Landsat-7 ETM+, Landsat-8/9 OLI, and Sentinel-2A/B MSI surface reflectance imageries from 2017 to 2022, we conducted high spatiotemporal resolution monitoring of seasonal water bodies in the Chahannur National Wetland Park. This park encompasses a climate-sensitive terminal lake located in Inner Mongolia’s semi-arid agro-pastoral ecotone. Various water indices and threshold methods were evaluated for water extraction. Among all the tested spectral indices and threshold methods, the rule-based approach integrating the Land Surface Water Index (LSWI) and a Vegetation Index (VI) yielded the highest water extraction accuracy, with an average overall accuracy of 95.57% and a Kappa coefficient of 0.90. By integrating dense time-series imagery from Landsat-7/8/9 and Sentinel-2 satellites (2017—2022) with an average temporal resolution of 9 d, we achieved near-decadal monitoring of wetland water surfaces. Our observations reveal that from 2017 to 2022, only a small, isolated lake in the northern part of Chahannur held water permanently, while all other water bodies were seasonal. The inundation frequency was 40% or higher in the eastern lake area, but below 20% in the western part. The annual maximum water surface area of Chahannur increased initially and then decreased, peaking at 34.50 km² in 2020 and dropping to a minimum of 8.65 km² in 2022. Seasonally, the average area from January to June was merely 1.24 km². It increased rapidly during the rainy season (from July to September) before gradually declining, and shrank back to pre-rainy-season levels after October. The lake area responded rapidly to precipitation during high-water periods but failed to maintain its extent, after reaching the maximum, it contracted within 10-30 d depending on subsequent rainfall. This study provides critical support for accurately understanding the seasonal dynamics of Chahannur Lake and guiding ecological restoration in the Chahannur National Wetland Park.