Background and Objective: Wetlands are highly sensitive, dynamic, and vital ecosystems that play a key role in biodiversity conservation, climate regulation, water quality improvement, and carbon sequestration. These ecosystems also provide habitats for a wide range of plant and animal species and contribute to flood mitigation and soil stabilization. However, they are increasingly affected by climate change, recurrent droughts, and human activities such as excessive water extraction and unsustainable development, leading to significant water loss and degradation. Therefore, identifying and analyzing long-term trends in wetland dynamics, particularly in systems with strong seasonal variability, is essential for sustainable water resource management and planning.

Materials and Methods: This study investigated changes in the water surface area of the Miankaleh Wetland in Iran from 2013 to 2025 using Landsat 8, Landsat 9, and Sentinel-2 satellite data. Initially, all available images were preprocessed through atmospheric correction and cloud filtering (less than 5% cloud cover). Water bodies were then extracted using the Modified Normalized Difference Water Index (MNDWI) combined with Otsu automatic thresholding. Post-processing steps were applied to reduce noise caused by clouds, shadows, and classification errors. The resulting datasets were integrated into a unified time series, and accuracy was evaluated using Relative RMSE. Finally, Seasonal Sen’s slope and the Modified Seasonal Mann–Kendall test were applied to assess long-term trends and seasonal effects.

Materials and Methods: This study investigated changes in the water surface area of the Miankaleh Wetland in Iran from 2013 to 2025 using Landsat 8, Landsat 9, and Sentinel-2 satellite data. Initially, all available images were preprocessed through atmospheric correction and cloud filtering (less than 5% cloud cover). Water bodies were then extracted using the Modified Normalized Difference Water Index (MNDWI) combined with Otsu automatic thresholding. Post-processing steps were applied to reduce noise caused by clouds, shadows, and classification errors. The resulting datasets were integrated into a unified time series, and accuracy was evaluated using Relative RMSE. Finally, Seasonal Sen’s slope and the Modified Seasonal Mann–Kendall test were applied to assess long-term trends and seasonal effects.

Materials and Methods: This study investigated changes in the water surface area of the Miankaleh Wetland in Iran from 2013 to 2025 using Landsat 8, Landsat 9, and Sentinel-2 satellite data. Initially, all available images were preprocessed through atmospheric correction and cloud filtering (less than 5% cloud cover). Water bodies were then extracted using the Modified Normalized Difference Water Index (MNDWI) combined with Otsu automatic thresholding. Post-processing steps were applied to reduce noise caused by clouds, shadows, and classification errors. The resulting datasets were integrated into a unified time series, and accuracy was evaluated using Relative RMSE. Finally, Seasonal Sen’s slope and the Modified Seasonal Mann–Kendall test were applied to assess long-term trends and seasonal effects.

Results: The results indicated a significant decreasing trend in the wetland’s water surface area (Sen’s slope ≈ −0.045 km²/day, τ = −0.714, p < 0.001), representing a strong and statistically significant decline. This corresponds to an average annual loss of approximately 16 km². Validation results confirmed high accuracy of the extracted water areas, with Relative RMSE below 10% for all sensors. Although all datasets showed a consistent declining trend, Sentinel-2 provided more detailed seasonal and short-term variations. The most pronounced decrease occurred in the later years, especially after 2022. Seasonal analysis revealed that the decline was not uniform throughout the year; the strongest reductions were generally observed in winter, while spring and summer exhibited more gradual but persistent decreases. Overall, despite short-term fluctuations and occasional temporary increases, the long-term trend remained strongly negative and concerning.

Discussion and Conclusion: Seasonal analysis enabled the separation of periodic effects and autocorrelation, demonstrating that seasonal approaches provide more reliable detection of true long-term trends compared to non-seasonal methods. It also highlighted distinct seasonal responses of the wetland, emphasizing the importance of accounting for seasonal variability in trend analysis. Ignoring these components may lead to misinterpretation of results. Overall, seasonal methods are essential for accurate monitoring of dynamic wetlands, and the findings of this study provide a scientific basis for improving the management, conservation, and restoration of vulnerable wetland ecosystems.