Date of Award
5-1-2024
Language
English
Document Type
Master's Thesis
Degree Name
Master of Science (MS)
College/School/Department
Department of Atmospheric and Environmental Sciences
Dissertation/Thesis Chair
Brian H Tang
Keywords
Convective, HRRR, Mesonet, Verification
Subject Categories
Atmospheric Sciences
Abstract
Forecasting can be more challenging in areas of complex terrain, such as in and around the valleys and mountains of Upstate New York. Biases in numerical weather prediction models, related to the lack of resolution of the terrain and errors in physical parameterizations, can affect forecasts. The main goal of this research is to assess biases in the High-Resolution Rapid Refresh (HRRR) model that could contribute to deficiencies in forecasts for days with and without convection. We investigate how the HRRR model performed in the New York Capital Region for 4 clear days, 4 isolated convective days, and 4 widespread convective days. These days were selected via a Clear Sky Index (CSI), in addition to a subjective analysis using observed NEXRAD reflectivity. HRRR model output, from the 06 UTC model run for each of the selected days, was compared to New York State Mesonet (NYSM) standard-site and flux-site observations during the period from 06 UTC through 00 UTC. Surface moisture, temperature, incoming shortwave, accumulated precipitation, and heat fluxes were compared. Results reveal that the HRRR has an overall afternoon warm and dry bias across all three CSI groups, with the clear group having a stronger bias in the morning. The HRRR on average predicts too positive of a sensible heat flux and too negative of a ground heat flux. Soil moisture at 5 cm is consistently drier than observations for all groups. Additionally, the NYSM standard sites have a consistent negative incoming shortwave (SW_down) error. In some locations, the measured SW_down for clear days was 15-20% lower than the calculated clear-sky SW_down using the Ineichen model, indicating caution should be exercised when using the standard-site SW_down data. Knowledge of these HRRR biases is important for informing forecasters using the model, explaining biases in convection and precipitation, and informing future NWP development.
Recommended Citation
Eldridge, Rachel A., "Examining The Performance Of The High-Resolution Rapid Refresh Model During Clear, Isolated, And Widespread Convective Days In The New York State Capital Region" (2024). Legacy Theses & Dissertations (2009 - 2024). 3309.
https://scholarsarchive.library.albany.edu/legacy-etd/3309
