New Mexico Statewide Drought Vulnerability Analysis Under Future Climate Change Scenarios Using a Physically Based Coupled Model

New Mexico, one of the fastest-warming states in the United States, faces increasing challenges in managing water resources because of declining snowpack and intensified groundwater extraction. The goal of this project was to assess drought vulnerability for the entire state using historical and forecasted climate data from 1940 to 2100. Specifically, the objectives were to: (1) determine an appropriate spatial resolution for statewide hydrologic modeling, (2) develop a computationally feasible coupled surface–subsurface hydrologic model, (3) implement distributed computing techniques for large-scale optimization, (4) forecast drought conditions to 2100 with associated uncertainty using CMIP6 climate projections, and (5) calculate a statewide Drought Vulnerability Index (DVI) to inform water resources management.

To achieve these objectives, we developed a high-resolution, fully automated hydrologic modeling framework VIC-MF6 that couples the Variable Infiltration Capacity (VIC) surface water model with the MODFLOW 6 (MF6) groundwater model. The VIC-MF6 framework operates at a spatial resolution of 1/32° (~3 km) and integrates observed and projected climate data from 1940 to 2100. Automated modules handle parameter downscaling, data conversion, and model synchronization via the MF6 Application Programming Interface (API) based on the eXtended Model Interface (XMI), allowing monthly exchange of baseflow and groundwater discharge. The framework was applied over the domain of the Rio Grande Transboundary Integrated Hydrologic Model (RGTIHM) and demonstrated scalability, computational efficiency, and physical consistency.

The statewide DVI reproduces observed drought episodes from 2010–2023, including 2011–2013, 2018, and 2020–2021, with strong agreement with U.S. Drought Monitor data. Historical statewide monthly mean DVI shows high interannual variability, with 12-month running means peaking around 2011–2013, when areas in high/extreme DVI (≥ 0.5) briefly exceeded 50%, and 2020–2021. After 2024, the smoothed DVI remains lower and more stable, with typical peaks below ~0.4 and high/extreme coverage under 20%. Spatially, historical mean DVI (1991–2024) is low to moderate and relatively homogeneous, while future mean DVI (2025–2100) increases in northern and west-central mountains (up to +0.3) and decreases in central and eastern plains (down to −0.3), indicating a redistribution of vulnerability rather than a uniform increase. The frequency of high-vulnerability months mirrors this pattern, increasing in headwater regions while remaining low elsewhere. Month-to-month DVI class transition probabilities show strong persistence (91.5% low, 90.7% moderate, 87.0% high, 72.6% extreme), supporting the use of backward windows and demonstrating gradual relaxation of extreme events. Distribution analysis reveals a lower future median DVI (~0.26 vs. ~0.32 historically) with increased spatial spread, reinforcing greater heterogeneity across grid cells.

These results indicate that, although the statewide average drought vulnerability may decline under projected climate conditions, high-elevation basins face increasing and more persistent risk. The findings provide quantitative, spatially explicit insight into New Mexico’s drought vulnerability and offers a robust foundation for region-specific water resources management strategies and long-term planning.