A high-resolution, event-based modeling framework for understanding the impact of climate change on hydrometeorological extreme events

Dynamical downscaling through regional climate modeling is often used in order to better address the needs of society, aiming to supply projections or predictions to stakeholders and decision-makers at a higher, more useful resolution than that provided by current global climate models. The details of downscaling (e.g., model resolution, domain selection, model physics) become of critical importance when addressing extreme weather phenomena at a local scale, as such events are often determined by and/or sensitive to small-scale processes. Particularly for flood events in complex terrain, it is necessary to resolve the details of both the terrain itself as well as driving fine-scale atmospheric processes to form a realistic picture of future flood risk.

This study explores high-resolution (1-km) simulations of warm-season extreme precipitation events in the western US, with the objective of better understanding the factors and physical processes responsible for changes in observed vs. future possible extreme events. The events examined are selected from observed extreme events in the Colorado Rocky Mountain and Front Range regions, such as the Fort Collins, CO flood of July 1997, as well as select events from the longer historical record used for water resources management planning. These events are simulated in high resolution using the Weather Research and Forecasting (WRF) model, and then a “climate perturbation” approach is applied in which the initial conditions of a single extreme precipitation event are modified by adding a “climate change signal” (derived from a spectrum of climate model projections) to the thermodynamic fields of the original event.

The objective of this work is to address questions related to storm maximization methods and probable maximum precipitation (PMP) limits -- two concepts commonly used for water resources management. For example, is a storm of PMP magnitude physically able to develop in a specific region? In regions where paleoflood data do not suggest that floods of such magnitude have actually occurred historically, can a model be “forced” to produce such a storm following reasonable maximization of key fields? Do such changes reflect realistic possibilities of future climate change? Potential applications for decision-making in the realm of water resources management will be discussed.