Mei et al. (2026) Phase-dependent responses of surface water dynamics to precipitation in the South China Karst based on remote sensing
Identification
- Journal: Journal of Hydrology Regional Studies
- Year: 2026
- Date: 2026-11-16
- Authors: Qianying Mei, Rui Li, Zhongfa Zhou, Wang Kaicun, Qiaoyan Zhang, Chaoyang Xue, Shitian Tang, Jie An, Wanquan Zhang, Zhenhong Yi
- DOI: 10.1016/j.ejrh.2026.104227
Research Groups
- School of Karst Science, Guizhou Normal University, Guiyang, Guizhou 550001, PR China: Qianying Mei, Rui Li, Zhongfa Zhou, Qiaoyan Zhang, Chaoyang Xue, Shitian Tang, Jie An, Wanquan Zhang, Zhenhong Yi
- State Engineering Technology Institute for Karst Desertification Control, Guiyang, Guizhou 550001, PR China: Qianying Mei, Rui Li, Zhongfa Zhou, Qiaoyan Zhang, Chaoyang Xue, Shitian Tang, Jie An, Wanquan Zhang, Zhenhong Yi
- College of Urban and Environmental Sciences, Peking University, Beijing 100871, PR China: Kaicun Wang
Short Summary
This study investigated the phase-dependent responses of surface water dynamics to precipitation variability in the South China Karst (SCK) region from 1990 to 2020. It revealed a pronounced phase-dependent pattern with initial expansion (1990–2007) followed by contraction (2008–2020), and a paradoxical shift in the precipitation–surface water relationship from positive to negative, driven by complex interactions of climate variability, human activities, and karst hydrogeological processes.
Objective
- To characterize the spatiotemporal patterns and long-term trends of surface water dynamics in the SCK during 1990–2020.
- To quantify the phase-specific effects and interactions of precipitation, temperature, Normalized Difference Vegetation Index (NDVI), and socioeconomic factors on surface water dynamics.
- To determine whether a phase-dependent precipitation–surface water response paradox exists under the combined influences of climate change and human activities.
Study Configuration
- Spatial Scale: The Southern China Karst (SCK) region, covering approximately 1.938 × 10⁶ km², encompassing eight provinces (Yunnan, Guizhou, Guangxi, Sichuan, Chongqing, Guangdong, Hubei, and Hunan). A representative watershed, the Nanpan–Beipan River Basin (approximately 83,449 km²), was also analyzed.
- Temporal Scale: 1990–2020 (31 years).
Methodology and Data
- Models used:
- Hybrid Water Index (combining Modified Normalized Difference Water Index (MNDWI), Normalized Difference Vegetation Index (NDVI), and Enhanced Vegetation Index (EVI)) for surface water extraction:
Water = (mNDWI > EVI or mNDWI > NDVI) and EVI < 0.1. - Google Earth Engine (GEE) platform for cloud-based processing.
- Mann-Kendall (MK) trend test for identifying change points and trends.
- Ordinary Least Squares (OLS) regression for trend validation.
- Pearson correlation analysis for quantifying linear relationships between variables.
- Stepwise Multiple Regression (SMR) for quantifying impacts of driving factors and identifying interaction effects.
- Fisher’s z-transformation for evaluating differences between phase-specific correlation coefficients.
- Confusion matrix and accuracy indicators (Overall Accuracy (OA), Producer's Accuracy (PA), User's Accuracy (UA), F1 score, Kappa coefficient) for classification accuracy assessment.
- Hybrid Water Index (combining Modified Normalized Difference Water Index (MNDWI), Normalized Difference Vegetation Index (NDVI), and Enhanced Vegetation Index (EVI)) for surface water extraction:
- Data sources:
- Remote Sensing Data:
- Landsat 5 Thematic Mapper (TM), Landsat 7 Enhanced Thematic Mapper Plus (ETM+), and Landsat 8 Operational Land Imager (OLI) Surface Reflectance (SR) imagery (1990–2020).
- Sentinel-2 MultiSpectral Instrument (MSI) imagery (2020) for validation.
- Shuttle Radar Topography Mission (SRTM) Digital Elevation Model (DEM) for terrain correction and shadow removal.
- MODIS Terra MOD13C1 and MOD13C2 monthly products for NDVI (0.05° resolution).
- Defense Meteorological Satellite Program/Operational Linescan System (DMSP/OLS) and Suomi National Polar-orbiting Partnership/Visible Infrared Imaging Radiometer Suite (NPP/VIIRS) data for night-time light index.
- Climatic Data:
- TerraClimate high-resolution global climate water balance dataset (1/24° (~4 km) spatial resolution, 1990–2020) for annual precipitation, temperature, potential evapotranspiration (PET), extreme precipitation, extreme drought, extreme low temperature, and extreme high temperature.
- Socio-economic and Land-use Data:
- Global Human Settlement Population (GHS-POP) dataset (100 m spatial resolution) for population density.
- China Land Cover Dataset (CLCD) (30 m spatial resolution, 1990–2020) for cultivated land area.
- Provincial statistical yearbooks (2003–2020) for agricultural, industrial, and domestic water use, and irrigated area.
- GeoDAR dataset for global dams and reservoirs.
- Validation Data:
- Official provincial Water Resources Bulletins (2000–2020) for annual surface water resources.
- Remote Sensing Data:
Main Results
- Surface Water Dynamics: The SCK surface water area exhibited a distinct two-phase pattern: an expansion phase from 1990 to 2007, with an average annual growth of approximately 471 km², reaching a peak around 2010 (approximately 33,333 km²). This was followed by a contraction phase from 2008 to 2020, with a decline to approximately 25,728 km² by 2020, representing a net loss of about 7606 km² from the 2010 peak.
- Phase-Dependent Precipitation–Surface Water Relationship:
- During 1990–2007, surface water area showed a significant positive correlation with precipitation (r = +0.277, p < 0.05).
- During 2008–2020, this relationship shifted to a significant negative correlation (r = -0.298).
- Precipitation variability increased considerably between the two phases, with the coefficient of variation rising from 6.30% to 8.68%.
- Multi-factor Regulation and Interaction Effects:
- Phase 1 (1990–2007): Precipitation positively interacted with NDVI (β= 0.184), GDP (β= 0.182), and reservoir abundance (β= 0.239). Negative interactions were observed with population density (β= −0.152), irrigation area (β= −0.12), temperature (β = −0.116), extreme heat days (β= −0.072), and extreme rainfall days (β= −0.064).
- Phase 2 (2008–2020): Negative interactions strengthened for population density (β= −0.252), temperature (β= −0.172), and irrigation area (β= −0.210). Positive interactions with NDVI (β= 0.230), GDP (β= 0.140), and reservoir abundance (β= 0.310) persisted but were insufficient to offset the combined negative influences.
- Representative Watershed Analysis: The Nanpan–Beipan River Basin reproduced the same phase-dependent transition and precipitation–surface water paradox (r = +0.47 in Phase 1; r = -0.32 in Phase 2), further supporting the robustness of the regional pattern.
- Accuracy Assessment: The surface water extraction achieved an Overall Accuracy (OA) of 87.44% and a Kappa coefficient of 0.7489. Remotely sensed summer surface water area was significantly and positively correlated with officially reported annual surface water resources across all eight provinces (Pearson's r ranging from 0.59 to 0.88, p < 0.05).
Contributions
- Provides the first large-scale, long-term (1990–2020) analysis of surface water dynamics and their phase-dependent responses to precipitation in the complex South China Karst region using multi-source remote sensing data.
- Identifies a novel "precipitation–surface water response paradox" where the relationship shifts from a significant positive correlation to a significant negative correlation over time, challenging conventional hydrological assumptions.
- Quantifies the interactive effects of climatic variability (e.g., temperature, extreme precipitation), human activities (e.g., population density, irrigation, reservoirs), and karst hydrogeological processes in regulating surface water evolution.
- Offers critical insights for understanding hydrological responses in karst environments and provides a scientific basis for water-resource monitoring, climate-change adaptation, and sustainable water management in ecologically vulnerable regions.
Funding
- National Natural Science Foundation of China (32572153; 32360421)
- Guizhou Province Department of Science and Technology (QianKeXie KJLYRC- [2026] 035)
Citation
@article{Mei2026Phasedependent,
author = {Mei, Qianying and Li, Rui and Zhou, Zhongfa and Kaicun, Wang and Zhang, Qiaoyan and Xue, Chaoyang and Tang, Shitian and An, Jie and Zhang, Wanquan and Yi, Zhenhong},
title = {Phase-dependent responses of surface water dynamics to precipitation in the South China Karst based on remote sensing},
journal = {Journal of Hydrology Regional Studies},
year = {2026},
doi = {10.1016/j.ejrh.2026.104227},
url = {https://doi.org/10.1016/j.ejrh.2026.104227}
}
Original Source: https://doi.org/10.1016/j.ejrh.2026.104227