Incorporating Climate Change into the Reclamation Safety of Dams Flood Risk Assessments

Climate change threatens to alter hydrology in basins across the United States, potentially resulting in increases in flood risk. As the owner of hundreds of high hazard dams in the western United States, the Bureau of Reclamation (Reclamation) considers flood risk in making decisions for its dams. Reclamation’s Safety of Dams process relies on probabilistic flood loadings information (among other types of loadings) to describe and assess risk of potential dam failure modes. However, up until now, Reclamation has not considered quantitative impacts of climate change on probabilistic flood loadings in the Safety of Dams process, leaving managers with an unknown vulnerability to future changes in climate.

The question of how to estimate probabilistic flood loadings and flood risk under transient climate conditions is not new. Across the public and private sectors, many have undertaken to study specific basins or propose methodologies. While many of these studies focus on using climate model output (e.g., statistically or dynamically downscaled fields) to estimate projected changes in streamflow, there are often challenges with these modeling frameworks. To circumvent concerns, other studies have focused on drawing conclusions about future climate conditions using atmospheric processes better represented by Earth system models. Decision scaling is an alternative approach to understanding climate impacts that falls into a different modeling framework. Specifically, decision scaling is a general approach to identify the atmospheric process that drives risk and then identify a subset of climate projections that adequately represent that atmospheric process (e.g., the risk driver) in order to determine likelihood of risk increasing and what decisions should be made. Because decision scaling begins first by looking at what drives the risk at a dam, for some basins this initial analysis may suggest further modeling with climate projections is not needed (e.g., risk is not sensitive to large perturbations in the climate).

Very recently, Reclamation has begun developing a decision framework and methodology for incorporating climate change projections into the Safety of Dams process for probabilistic flood loading estimates and decision making. The proposed flood risk methodology involves applying decision scaling to flood risk assessments at Reclamation dams. For example, risk at many Reclamation dams is sensitive to changes in hydrologic extremes, such as increasing flood volume, larger flood peaks, or changing seasonality of flows. These hydrologic extremes may be driven by large-scale atmospheric processes (e.g., the location of persistent high or low pressure regions, etc.) affected by climate change. The intent of the proposed flood risk methodology is to establish a process of quantitatively analyzing potential changes to flood risk present in future climate projections that can be applied to dams across the western US. In this presentation, we will share both the flood risk methodology and decision framework we developed. Furthermore, we hope to share some examples of how these can be applied across Reclamation’s dam portfolio.