The "research" repository contains jupyter notebooks written in the python programming language pertaining to my undergraduate thesis compleated in May of 2023 titled, "Precipitation Type in the High-Resolution Ensemble Forecast System during the February 17th & February 23rd, 2022, Winter Storms" - A Thesis Submitted to the University at Albany, State University of New York In Partial Fulfillment of the Requirements for the Degree of Bachelor of Science.
Winter weather events with near-freezing surface temperatures are often associated with mixed surface precipitation types (p-types) and are shaped by synoptic, mesoscale, and microscale processes. The Winter Precipitation Type Research Multi-Scale Experiment (WINTRE-MIX) focuses on how these processes influence the variability and predictability of p-type and amount under near-freezing surface conditions. Accurately predicting p-type in these conditions is a forecast challenge and can have large implications on aviation, transportation, and utilities. In operational meteorology, the High-Resolution Ensemble Forecast System (HREF) is commonly used, assisting forecasters in probabilistic forecasting. The High-Resolution Ensemble Forecast (HREF) system outputs ensemble products from a collection of high-resolution numerical weather prediction models. Each member of the forecast ensemble is distinct with respect to its combination of dynamical core, physics parameterizations, and initial/boundary conditions.
This study investigates near-freezing precipitation from two winter storms observed in southern Quebec and northern New York by the Winter Precipitation Type Research Multi-Scale Experiment (WINTRE-MIX) to evaluate commonalities and discrepancies between members of the HREF ensemble and to explore what meteorological features contribute to member discrepancies. The WINTRE-MIX campaign deployed field teams to launch radiosondes which provide thermodynamic vertical profiles of the atmosphere. These field teams also manually collected surface p-type measurements. The New York State Mesonet (NYSM), and Automated Surface Observing System (ASOS) surface station networks provide thermodynamic measurements and p-type estimates through their respective multi-sensor diagnostic products.
For this research, a p-type diagnostic product was also developed for the Canadian Fund for Innovation Climate Sentinel (CFICS) network, following a similar approach to the NYSM's p-type diagnostic. Through both near-freezing winter storms, HREF simulations show member disagreement in near-surface and mid-level temperatures resulting in diverse p-type simulations. For the 17 February 2022 winter storm, the NAM-NEST and HRRR provided accurate forecast guidance of multiple transition periods of p-type due to accuracy in simulating northerly low-level terrain-channeled cold air into the Champlain Valley as well as a regime of southwesterly warm air advection aloft. For the 23 February 2022 winter storm, all members performed poorly in the simulation of a shallow surface cold layer within the St. Lawrence Valley. An extended period of FZRA observed in the valley was largely mis-simulated as RA by most members. The NAM NEST showed the best performance as it best forecasted the shallow surface cold layer. Within the HREF ensemble, there is room for improvement in representing p-type, especially along regions of transition and over complex terrain. Farther analyses of such events are required to identify the meteorological features that contribute to member biases.