The Potential Impacts of Climate and Land Use Changes on Water Yield in the Croatan National Forest, USA

Coastal forests are highly sensitive to both climate change and land use change, which can strongly affect hydrological processes and long-term water yield. This study quantifies the individual and combined impacts of climate change and land use/land cover (LULC) change on water yield in the Croatan National Forest (CNF), a coastal ecosystem in North Carolina, USA, from 2003 to 2070. To produce high-resolution climate projections, we extended the MIDAS (Machine Learning-Based Integration of Downscaled Projections for Accurate Simulation) approach by applying a full statistical downscaling of temperature and precipitation from CMIP6&ndash:SSP5-8.5 scenarios using the Random Forest algorithm. Future LULC scenarios were generated using machine learning and Markov Chain-based modeling to predict spatial changes up to 2070. The downscaled climate and LULC data were integrated into the WaSSI hydrological model to simulate their potential effects on water yield under the following four scenarios: baseline, LULC change only, climate change only, and combined change. The results showed that climate change alone could reduce annual water yield by about 11%, while LULC change alone could increase it by roughly 3% due to lower evapotranspiration from forest-to-urban conversion. Under the combined scenario, water yield decreased by about 6%, indicating that climate change dominated, but LULC change could locally alter or influence its effects. Overall, the findings highlight that climate change could be the primary driver of reduced water yield in coastal forests, while LULC change mainly affects its spatial variability. This integrated framework improves the accuracy of regional hydrological projections and provides useful insights for climate adaptation and sustainable water resource management in coastal forest ecosystems.