This project aimed to provide the U.S. Department of Defense (DoD) with a comprehensive analysis of uncertainty in generating regional-scale projections that can be used by stakeholders and decision makers to plan for future environmental impacts at specific locations. The merits and limitations of commonly used downscaling models, from simple to complex, were compared and evaluated for their suitability at installation scales. Downscaled projections were generated at selected DoD installations using dynamic and statistical methods, focusing on probability distributions of environmental variables and associated uncertainties. Site and variable selection for downscaling was based on a thorough understanding of the current and projected roles of weather and environmental factors in the operation, maintenance, and planning of DoD facilities.

Through interviews with DoD installation stakeholders, researchers identified key environmental variables and vulnerabilities of interest in the initial phase of the project. High-resolution projections for North America were generated by combining advanced dynamical and statistical downscaling models with quality-controlled observations and the latest global simulations of the Earth's system. Expected changes in environmental variables were evaluated by analyzing these downscaled model products at relevant spatial scales for selected DoD installations.

Researchers surveyed DoD installation stakeholders to identify environmental concerns, particularly changes in precipitation extremes. Using the PRISM dataset for precipitation and temperature, they evaluated downscaled projections, showing that a 12-km resolution reduced model bias compared to coarser models, especially in topographically complex regions.

Findings revealed that model accuracy varied by region and boundary conditions. Bias correction and nudging didn't always improve results, especially in high-precipitation and arid regions. The model’s ability to capture historical uncertainties was essential for future projections.

For 2085–2095, temperature projections showed increases of 5–7°C in summer and >7°C in winter in northern Canada and Alaska. Summer precipitation was projected to rise by up to 60% in Canada, Alaska, and the southwestern U.S., while winter precipitation in the Southwest and Southern Great Plains was expected to decrease by 40–60%. Heavy precipitation days increased by 3–5 days in the eastern U.S., Alaska, and Canada, with ~1 additional day in the western U.S. Projections also showed over 60 additional days/year with temperatures >90°F in the Great Plains and eastern U.S., and fewer than 20 additional days in the Rockies, Cascade Range, Alaska, and Canada. The West Mountain region could lose over 60 frost days annually, with significant elevation dependence in temperature extremes.

This project provided the knowledge, datasets, and tools needed to assess vulnerabilities, including severe-event probabilities, changes in operating parameters (e.g., heating and cooling needs), and other challenges for DoD installations and range management. By evaluating the value and appropriateness of downscaled projections for DoD installations, researchers advanced their use for local planning. The generated downscaled projections, along with associated uncertainties, offer a crucial resource for DoD to adapt to changing conditions at specific locations.

The methodology, data sets, and model products are available in an accessible database for future use by DoD. An illustrative example and a GIS application for exploring the data at selected installations will also be provided. These resources will help DoD develop informed policies for addressing future environmental changes.