Process-based Evaluation of Temperature and Precipitation Projections and Downscaling Methods over the CONUS: Charting a Path for End-Users from the CMIP6 Ensemble to Multivariate Facility-Level Risks

Daniel Feldman | Lawrence Berkeley National Laboratory

RC19-1391

Objective

The objective of this research is to develop a process-based evaluation of the large number of climate model simulations from the Coupled Model Intercomparison Project – Phase 6 (CMIP6) in order to develop robust estimates of changes, and associated confidence intervals, in temperature and precipitation distributions at the Department of Defense (DoD) facility level across the Conterminous United States (CONUS). As part of this, the project team seeks to understand when and where statistical and dynamical downscaling methods diverge to ultimately help end-users navigate the landscape of climate model simulations and downscaling approaches.

Back to Top

Technical Approach

The research will develop an approach to look at the question of how climate model projections of temperature and precipitation can be used to develop robust risk estimates for the changing threats to DoD facilities projected under climate-changed future conditions. The distinguishing feature here is that the project team will develop process-based metrics to establish climate model skill regarding CONUS temperature and precipitation projections at the regional level for CMIP6 for the business-as-usual scenario Representative Concentration Pathway 8.5/ Shared Socioeconomic Pathways 5  (RCP8.5/SSP5) and a scenario with mitigated climate change effects Representative Concentration Pathway 4.5/ Shared Socioeconomic Pathways 1 (RCP4.5/SSP1). Due to the importance of downscaling, the project team will concurrently evaluate, through observations, whether the assumption of stationarity inherent in statistical downscaling techniques is valid over decadal time-scales. The project team will do this by applying Localized Constructed Analogs (LOCA) statistical downscaling, which was trained with data from 1950-2005, on data since 2005 to determine the error incurred from applying LOCA where it has not been trained. Next, the project team will use existing dynamical downscaling experiments to determine the temporal and spatial scales and settings at which statistical and dynamical downscaling diverge in the 21st Century. Additionally, the project team will update LOCA to LOCA2 and apply it to most CMIP6 models that report results for the RCP4.5/SSP1 and RCP8.5/SSP5 scenarios. Finally, the project team will use process-based skill metrics and updated downscaling methods to develop ensemble temperature and precipitation projections and associated confidence intervals.

Back to Top

Benefits

A key deliverable from this work will be new projections of future climate-changed conditions of temperature and precipitation at finer spatial scales than is most often available for climate vulnerability assessments. These can be widely utilized by the scientific and DoD user communities. Additionally, one specific use-case for this deliverable is that it will help characterize risk in specific watersheds around potentially vulnerable DoD installations. Additionally, the LOCA2-derived future climatology from this project may help update the findings in the FY18 “National Defense Appropriations Act (NDAA; Public Law 115-91) Section 335 Evaluation of Climate Change Effects on Army Installations” [White, K.D. et al., USACE Washington, DC 2018] delivered to Congress in January 2019 in the “Report on Effects of a Changing Climate to the Defense Department,” from the Office of the Under Secretary of Defense for Acquisition and Sustainment, to specifically incorporate CMIP6 model results.

Back to Top

Points of Contact

Principal Investigator

Daniel Feldman

Lawrence Berkeley National Laboratory

Phone: 510-495-2171

Program Manager

Resource Conservation and Resiliency

SERDP and ESTCP

Share