Research - Matthew Bray

Evaluating the RAP/HRRR NWP Model in the Arctic

Summary

Errors in the RAP model relative to observations at Utqiagvik, AK as a function of height and forecast lead time from Bray, Turner, and de Boer (2021).

Although the Arctic is home to indigenous communities and economic interests that are highly sensitive to weather conditions, relatively few high-resolution operational models cover the region, and even fewer comprehensive evaluations of real-time model performance in the region have been conducted. One such numerical weather prediction (NWP) model that has covered the Arctic (specifically Alaska) for an extended period is the Rapid Refresh (RAP) / High-Resolution Rapid Refresh (HRRR) modeling system. As a part of my NOAA Hollings scholarship, I conducted an evaluation of two years of RAP/HRRR model output against in-situ observations from two stations on the northern coast of Alaska, with a focus on key Arctic phenomena like mixed-phase clouds, stable boundary layers, and ice-ocean surface effects.

Persistent seasonally varying temperature and humidity biases near the surface emerged that varied with wind direction, which was likely related to the model's handling of turbulent fluxes over sea ice (newer model versions with better fractional sea ice representation improved on these biases). Moreover, strong surface inversions were routinely under-represented, affecting temperature and cloud forecasts and likely stemming from limited vertical resolution in the boundary layer. Additional surface temperature biases appeared related to the occurrence of mixed-phase clouds, which were modeled in the RAP with relatively too little liquid water, leading to errors in downwelling longwave radiation calculations. Notably, the RAP is built on the basic infrastructure of the Weather Research and Forecasting (WRF) model, making these findings broadly applicable to many previous NWP experiments in the Arctic.

Publications

Bray, M. T., D. D. Turner, and G. de Boer, 2021: Evaluation of the Rapid Refresh Numerical Weather Prediction Model Over Arctic Alaska. Wea. Forecasting, 36, 1061-1077, https://doi.org/10.1175/WAF-D-20-0169.1.

Conference Presentations

Bray, M. T., D. D. Turner, and G. de Boer, 2020: "Evaluating the Rapid Refresh Numerical Weather Prediction Model in the Arctic." 100th American Meteorological Society Annual Meeting; 30th Conference on Weather Analysis and Forecasting (WAF)/26th Conference on Numerical Weather Prediction (NWP), Boston, MA. Poster Presentation.