Convection-Permitting Modeling for Intense Precipitation Processes (NUREG/CR-7290)

On this page:

• Publication Information
• Abstract

Download complete document

• NUREG/CR-7290 (PDF – 7.11 MB)

Publication Information

Manuscript Completed: September 2021
Date Published: May 2023

Prepared by:
Andreas F. Prein
Jordan G. Powers
Erin L. Towler
David Ahijevych
Ryan A. Sobash
Craig S. Schwartz

National Center of Atmospheric Research (NCAR)
3090 Center Green Dr.
Boulder, CO 80301

Elena Yegorova, NRC Project Manager

Office of Nuclear Regulatory Research
U.S. Nuclear Regulatory Commission
Washington DC 20555-0001

Availability Notice

Abstract

This report documents work sponsored by the U.S. Nuclear Regulatory Commission (NRC) and conducted by the National Center for Atmospheric Research (NCAR) as part of the RES project, "Convection-Permitting Modeling for Intense Precipitation Processes." This project was undertaken as part of the Probabilistic Flood Hazard Assessment (PFHA) Research Program. The objective of the PFHA Research Program is to develop tools and guidance on the use of PFHA methods to risk-inform NRC's licensing of new facilities as well as the licensing and oversight of currently operating facilities as they relate to flooding hazards.

Many flooding scenarios of interest to nuclear power plant (NPP) licensing and oversight involve extreme precipitation events occurring at the plant site or within the watershed of the plant. Generating probabilistic assessments of extreme precipitation within a catchment is challenging due to typically short observational records, insufficient data coverage, and climatic variation. Furthermore, traditionally used estimators of extreme precipitation, such as Probable Maximum Precipitation (PMP), do not allow for the quantification of uncertainties in hazard estimates in either a physical or a risk sense. The application of numerical atmospheric models to the problem offers a way forward. State-of-the-art convection-permitting models (CPMs) can explicitly simulate deep convection and can accurately represent orography on fine scales, and thus they present powerful tools for investigating extreme precipitation events. Using convection-permitting models is, therefore, a valuable alternative approach in the provision of more physically-based and probabilistic flood risk assessments.

This report includes a thorough literature review and analysis that summarize the state of the science in simulating extreme precipitation events with convection-permitting models while outlining key challenges, opportunities, and promising research areas. Based on the literature review, we assess the ability of CPMs to capture extreme precipitation in recent flood events in the contiguous U.S. (CONUS) east of the Rocky Mountains by leveraging three existing convection-permitting ensemble datasets. These cover 10,570 36-hour simulations/forecasts at 3-km horizontal grid spacing (Δx=3 km) and 810 36-hour simulations at Δx=1 km spacing. Additionally, we analyze the impact of observational uncertainties on the results by using a selection of high-quality multi-sensor and gauge-based rainfall datasets.

The central finding is that numerical weather prediction models configured with convection permitting resolutions can capture heavy precipitation events in the Eastern U.S. Many aspects of the simulations are verified by multi-sensor observational datasets, and the precipitation output from the convection-permitting (CP) model configurations shows less error than precipitation estimates based on station data. This demonstrates the potential value of incorporating CP model outputs in flood risk assessments. However, CP model configurations are not perfect and sometimes show systematic biases, as revealed in case examples of the underestimation of event peak accumulations by up to 30%. While one way to reduce these biases is by statistical post-processing, over the long-term model development to improve fidelity is the preferred way to address the issue.

The report closes with two recommendations for future work on the usage of CPM output for probabilistic flood risk assessments. First, targeted downscaling of heavy precipitation events in global climate models with CPMs would allow the building of a catalog of heavy precipitation events that are physically plausible, but unprecedented in the observational record. Second, events from this catalog and existing CPM heavy precipitation simulations could be used in combination with statistical approaches, such as stochastic storm transposition, to generate a large set of plausible heavy rainfall events to generate input for hydrologic models (e.g., WRF-Hydro).