Identifying climate change vulnerabilities from the bottom up

Abstract:

Many climate change assessments apply a top-down approach. This starts with information from global climate models, then climate variables are downscaled and multiple models are linked together to make this global-scale information applicable to local systems. While these climate model-driven approaches have many benefits, models are imperfect representations of reality, and thus each step of the modeling cascade requires decisions on how best to span space and time, which can skew results. This research explores an alternative approach which uses the same modeling framework as past climate change studies (global climate models, hydrological models, and impact assessment tools), but reverses the direction of information flow: first, water resource managers identify a metric of concern, e.g., flow above X cfs at a certain gauge, then through model iterations, hydrologic factors that lead to the metric are diagnosed, and finally connections to climate drivers are quantified. In other words, instead of starting with the global climate information, which is often the least understood, the approach first investigates local water system sensitivities. From a hydrologic perspective, this capitalizes on using past weather events to better quantify conditions (e.g. extent, duration, and intensity of precipitation, snow pack, soil moisture) that cause extreme hydrologic events. This helps identify how the character of simulated future events, as they continue to evolve, differs from those of the past, and importantly, forces the assessment to consider the particulars of the local system, the impact of interest, and the decision makers at the start of the study. This presentation will discuss the opportunities and challenges of this approach in the Pacific Northwest. Floods in the Skagit River in western Washington, selected after numerous meetings with water resource managers throughout the region, will be highlighted as a test case for this bottom-up approach.