Argonne researchers develop a tool for quantifying extremes under changing climate
In the past few decades, extreme storms and intense precipitation events have more frequently occurred in the United States, resulting in disruptive impacts on highly populated urban areas and critical infrastructure and leading to significant damage to property, loss of human life, and deterioration of human health. The catastrophic flooding associated with Hurricane Harvey (August 2017), for example, was Houston's third flood in recent years exceeding the "100-year flood" mark. Such precipitation and hydrologic extremes can be characterized using intensity-duration-frequency (IDF) analysis, which provides a design basis for engineering flood-sensitive infrastructure. However, the current IDF analysis framework was developed around the assumption that climate conditions will remain stationary over the next 50 or 100 years.
EVS hydrologist Eugene Yan and collaborating Argonne researchers, together with colleagues at Washington State University, have developed an improved IDF analysis framework, recently implemented in southern Maine. The updated framework (1) identified non-stationary changes in the climate by incorporating both historical observations and future climate projections; (2) developed processes capable of performing data pre-processing, multiple statistical distribution model evaluation and selection, model parameter uncertainty quantification, and performance validation; and (3) quantified runoff IDFs using precipitation IDF estimates for the Casco Bay Watershed in Maine. The updated IDF analysis framework considers non-stationary distribution models and snowmelt effects that are not incorporated in conventional methods.
The quantification of extreme precipitation events is vitally important for designing critical infrastructure and for urban planning. The improved IDF analysis framework developed for this study provides a better ability to quantify intensity of 10-, 50-, 100-, and 500-year storms in the context of changing future climate, and parameterizes the increasing trends in these events over time. Results from the improved analysis framework show a marked increase in intensity (45% on average) and frequency of the extreme precipitation events when compared to those using only historical records that assuming stationarity. This implies potential under-preparedness of flood-sensitive infrastructure in this region, which could be at risk for flood damage if original designs are based on lower IDF estimates using conventional method.
The improved IDF analysis framework developed in this study will provide a design basis for flood-sensitive infrastructure (urban drainage systems, roads, and hydraulic structures) that will aid in flood risk management. The results will enable and support a more science-based decision-making process for urban planning in the context of future flooding potential considering non-stationary climate shift.
This study is part of the Regional Resiliency Assessment Program (RRAP), which addresses a range of hazards that could be regionally significant for Portland, Maine. The RRAP is led by the Department of Homeland Security. Argonne is a partner in RRAP for Portland, Maine. Other researchers in this study include Alissa Jared, Julia Pierce, Vinod Mahat, and Mark Picel of EVS; Duane Verner and Thomas Wall of the Global Security Sciences Division; and Yonas Demissie and Edom Moges at Washington State University.