Theses and Dissertations

ORCID

https://orcid.org/0009-0004-9174-9515

Advisor

Mercer, Andrew E.

Committee Member

Brown, Michael E.

Committee Member

Gutter, Barrett F.

Committee Member

Thompson, Daniel B.

Date of Degree

12-12-2025

Original embargo terms

Visible MSU Only 1 year

Document Type

Graduate Thesis - Campus Access Only

Major

Geosciences (Applied Meteorology)

Degree Name

Master of Science (M.S.)

College

College of Arts and Sciences

Department

Department of Geosciences

Abstract

Dewpoint bombs are a mesoscale, rapid-onset drying pattern which can impact fire weather predictions yet forecasting them remains challenging. This study uses surface observations to investigate dewpoint bomb occurrence in the National Weather Service’s Marquette Weather Forecasting Office (WFO) Fire Weather Area of Responsibility (AOR) between March 1 and November 30 from 2006 to 2024 across 14 sites. Sounding data is used to classify all bomb events as synoptically benign or not. Dewpoint bombs from high-fire risk days are modeled using logistic regression and support vector machine models and random forests. Results show that dewpoint bombs occur on 52% of days that meet relative humidity criteria for Red Flag Warning within the study period and that 13.4% of dewpoint bombs occur during synoptically benign conditions. Model results indicate that resolving lower free atmosphere, entrainment zone, and planetary boundary layer characteristics throughout bomb evolution is key to accurately forecasting dewpoint bombs.

Download

Share

COinS