Understanding and Predicting Precipitation Characteristics in the United States Through Machine Learning, Numerical Modeling and Measurements

Author

• Cody Luther Ratterman, Utah State UniversityFollow

Date of Award:

8-2026

Document Type:

Dissertation

Degree Name:

Doctor of Philosophy (PhD)

Department:

Plants, Soils, and Climate

Committee Chair(s)

Wei Zhang

Committee

Wei Zhang

Committee

Yoshi Chikamoto

Committee

Jonathan Meyer

Committee

Enrico Zorzetto

Committee

Cenlin He

Abstract

Precipitation is one of the most important yet uncertain variables in the climate system. It varies dramatically in terms of timing, accumulation, rate, and phase, depending on location, season, circulation patterns, and atmospheric conditions. Precipitation forecasts, especially snowfall, are essential to supporting drought mitigation and water management in the Intermountain West. Because rainfall and snowfall lead to opposite effects on snowpack, accurately partitioning rain and snow is important to estimate snowpack levels, winter recreation, mountain ecosystems and runoff. The research findings in this dissertation have advanced the understanding and prediction of precipitation and snowpack in the U.S. by addressing the following questions:

• Whether and to what extent atmospheric patterns are useful to improve seasonalprecipitation predictions?

• Can we improve mountain snowpack forecasts produced by dynamical weathermodels?

• Which meteorological variables are most useful to partition rain and snow?

• How do climate models simulate snowfall, rainfall and snow fraction?

• How can we measure precipitation phases?

Here we leverage machine learning, numerical modeling and measurements to address these questions. Our results have shown that seasonal precipitation forecasts can be improved by combining weather regimes and Gaussian Mixture Models. Snowpack forecasts have been improved by statistically downscaling NOAA’s snow water equivalent forecasts using the analog method. The decision-tree-based machine learning method helps identify surface meteorological conditions distinguishing rain from snow. Historical changes in snowpack and snow fraction can be attributed to temperature change, rather than relative humidity variation. We also developed sensors to measure the phase of precipitation (rain and snow) using heated precipitation detectors.

Recommended Citation

Ratterman, Cody Luther, "Understanding and Predicting Precipitation Characteristics in the United States Through Machine Learning, Numerical Modeling and Measurements" (2026). All Graduate Theses and Dissertations, Fall 2023 to Present. 891.
https://digitalcommons.usu.edu/etd2023/891