ORCID

https://orcid.org/0009-0008-8886-8125

Date of Award

Summer 2026

Language

English

Embargo Period

8-3-2026

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

College/School/Department

Department of Atmospheric and Environmental Sciences

Program

Atmospheric Science

First Advisor

Ryan Torn

Committee Members

Kristen Corbosiero, Robert Fovell, David Novak, Brian Tang

Keywords

tropical cyclone, precipitation, forecast verification

Subject Categories

Atmospheric Sciences

Abstract

Precipitation from landfalling tropical cyclones (TCs) poses a significant risk to life and property in both coastal and inland communities. Early warning systems can help mitigate the impacts of TC-induced flooding; however, this requires accurate quantitative precipitation forecasts (QPFs). Forecasting precipitation from landfalling TCs is a challenge, as the intensity, duration, and location of precipitation can be dependent on numerous storm-related and environmental factors. Previous studies have attempted to identify and correct NWP model deficiencies and reduce forecast uncertainty by applying traditional and advanced verification metrics to TC QPF. However, these studies have mostly used gridpoint verification metrics, limited samples of forecasts, and older versions of NWP models. Therefore, this dissertation aims to evaluate the precipitation forecasts of landfalling TCs in CONUS using object-based verification methods within the Model Evaluation Toolkit (MET) to investigate what aspects of the precipitation forecast are driving storm-total and 6-h TC QPF errors.

Storm-total and 6-h TC QPFs from global and regional forecast models are verified for TCs that made landfall in CONUS from 2018 to 2024. Overall, there are minimal signs of storm-total QPF error improvement as forecast lead time decreases in any 1-inch threshold or 5-inch threshold cluster verification metric for either the National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS) or the European Centre for Medium-Range Weather Forecasts (ECMWF); however, the GFS does consistently under-forecast all amounts of storm-total precipitation within 1-inch threshold clusters compared to the ECMWF. In addition, when compared to the GFS, the Hurricane Analysis and Forecast System (HAFS) produces 6-h precipitation forecasts with fewer location errors, but over-forecasts all amounts of 6-h precipitation, especially for the highest magnitudes of 6-h precipitation. However, HAFS tends to over-forecast the amount of precipitation and place the precipitation in the wrong location, while the ECMWF forecasts too little precipitation over too large of an area.

Lastly, ensemble-based sensitivity analysis is used to investigate the sensitivity of two high-error precipitation forecasts to the uncertainty in upper-tropospheric features and lower-tropospheric thermodynamic boundaries at earlier lead times. The 1200 UTC 22 August ECMWF EPS forecast for Hurricane Henri (2021) exhibited high overlap errors in the storm-total precipitation forecast and is sensitive to the location and shape of the nearby upper-level low over the Mid-Atlantic CONUS. The ensemble members with the best precipitation overlap forecast appear to produce a good precipitation forecast despite incorrectly forecasting the shape and position of the upper-level low relative to ERA5. Meanwhile, the 1200 UTC 29 August GEFS forecast for Hurricane Idalia (2023) under-forecasts all amounts of precipitation in numerous verification blocks valid after Idalia made landfall and is sensitive to the location of an upstream trough, the existence and magnitude of downstream ridge amplification, and the development and location of nearby lower-tropospheric baroclinic zones, which are all associated with Idalia’s extratropical transition. A comparison of the best and worst precipitation intensity ensemble members to ERA5 reveals that the precipitation intensity forecast errors are more likely driven by localized errors in the mesoscale environment affiliated with Idalia’s extratropical transition instead errors of the synoptic environment.

License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Download

Share

COinS