Last Updated: March 7, 2025
This is our holding place for project descriptions while we get them all identified. *These are draft descriptions.* Keep in mind that research projects can change quite a bit. That is part of the nature of research. You don't know how things will work out when no one has done your project before.
We put a link to recent REU projects for that mentor, when possible. That doesn't mean the proposed project is like the others, but will give you a flavor of that mentor's research. Likewise, some projects have "most directly applicable majors" listed.
Students applying to the program can mention any particular projects of interest in their essays if they wish, but that is not required and we may not have all projects identified during the application period.
Students who are selected to the program will be asked to rank the projects. This does not necessarily limit who can choose the project, but instead is meant to reveal something of the nature of the project.
The following projects will be funded through this REU.
1. Understanding winter precipitation in urban areas
Mentors: Dr. Heather Reeves (OU/CIWRO & NOAA/NSSL)
Description: An open question about winter precipitation (i.e., snow versus rain versus freezing rain, etc) is whether the sensible heat flux can cause warmer forms of precipitation to fall in urban areas versus surrounding rural areas. It has long been presumed that this happens, but clear and irrefutable evidence of it has yet to be discovered. The aim of this project is to better understand the influences of urban heat on winter precipitation type and better characterize the conditions under which this may occur through inspection of various observational datasets including aircraft observations, upper-air observations, surface observations, and model data. These results will then be used to better understand the impacts of winter precipitation in urban areas on aviation interests such as aircraft deicing and flight restrictions on unmanned aircraft. This project affords the opportunity to work as a part of multi-agency collaboration with partners at Mitre Corp and Florida State University in addition to the mentors located at OU/CIWRO and NOAA/NSSL. The research will involve Python programming.
Desired skills: Python programming knowledge or a quick study to learn it; this project relies on Python for success
Most applicable majors:
Recent REU projects with mentors on this team:
- Traffic Fatalities in Winter: An Evaluation of Weather Regimes and NWS Guidance During Killer Storms
- Building a Climatology of Snow Squall Conditions
- Mapping Precipitation Intensity to Radar Observations and Derivatives
2. Evaluating the Impact of Vertical Mixing Schemes in the Ocean Surface Boundary Layer on Intensification Forecast of Hurricane Fiona (2022)
Mentors: Dr. Yue Yang (OU/SoM/MAP), Prof. Xuguang Wang (OU/SoM/MAP)
Description: The turbulent mixing in the ocean surface boundary layer (OSBL) can modulate the exchange of heat and momentum between the atmosphere and ocean. Such air-sea fluxes contribute to the energy budget of tropical cyclones (TCs) and affect the storm intensity. Consequently, accurate parameterization of the OSBL turbulent mixing in the vertical mixing schemes is essential for modeling TC intensification. In addition, establishing a physically consistent ocean-atmosphere background ensemble across the OSBL and atmosphere planetary boundary layer (APBL) is crucial for the air-sea coupled ensemble data assimilation (DA) system. Under the Unified Forecasting System (UFS) framework, the self-cycled Hurricane Analysis and Forecast System (HAFS) has been developed to implement the eddy-resolving regional Modular Ocean Model version 6 (MOM6) ocean coupling capability (HAFS-MOM6) by the OU MAP lab. This project aims to evaluate the impact of vertical mixing schemes and associated parameters in the OSBL on the ocean-atmosphere background ensemble of HAFS-MOM6 for Hurricane Fiona (2022) using novel observations at the air-sea interface. The ocean components to be tested include the vertical mixing scheme in the OSBL (KPP vs ePBL) and the critical Richardson number in the KPP. Novel observations were collected from a field campaign using instruments such as saildrones, paired dropsondes and Airborne Expendable Bathy-Thermograph (AXBT) probes, and gliders. This project will provide the student with an opportunity to diagnose and visualize model outputs, gain familiar with novel observations at the air-sea interface, and develop a foundational understanding of DA.
Desired skills: Experience with coding and programming
Most applicable majors: Meteorology or related field
Recent REU projects with mentors on this team:
- Comparing OU MAP Real-Time Convection Allowing Ensemble Forecasts Produced During 2021 & 2022 Using Neighborhood & Surrogate Verification Methods
- Regime Dependent Verification and Calibration of a 10-Member Convection-Permitting Ensemble during the 2019 HWT SFE
- Impact of AQUA Satellite Data on Hurricane Forecast: Danielle 2010
3. Identifying Emergency Manager Workflows for a Variety of Extreme Weather Events
Mentors: Elizabeth Meister (OU/IPPRA), Anna Wanless (OU/IPPRA), and Sam Stormer (OU/IPPRA)
Description: This project will focus on an analysis of Emergency Management workflow interview data. Data were collected for a nationwide project where interviews were conducted with various types of emergency managers and hazards. Specifically, students will analyze interview data from emergency managers in urban and rural areas, for different kinds of municipalities and/or private industry. The hazards studied are tropical cyclones, severe weather, winter weather, wildfires, extreme heat, and flooding. The goal is to evaluate what information EMs are looking for, what specific decisions are made based on forecast information, what sources or channels they rely on, when and how they share hazardous weather forecast information. This project will focus on qualitative analysis of interview data (i.e., finding themes across hazards and municipalities, etc.) and require minimal coding in R (which the mentors can provide guidance on as needed).
Desired skills: Familiarity with R programming language
Most applicable majors: Meteorology, Emergency Management, communication, other related social sciences
Recent REU projects with mentors on this team:
4. Verification of AI-generated Severe Weather Forecasts across North America
Mentors: Aaron Hill (OU/SoM)
Description: Artificial Intelligence and Machine Learning (AI/ML) are becoming commonplace in severe weather forecasting. The Global Ensemble Forecasting System Machine Learning Probabilities (GEFS-MLP) modeling system forecasts severe weather hazards from 1—8 days into the future across the U.S. to support forecast operations at the National Weather Service. The forecasts are skillful and reliable, and they have demonstrated value in the operational forecast process. Recently, the forecast systems were extended to Canada to support severe weather field programs that lacked operational forecast guidance. Despite not being trained to predict severe storms in the region, the forecasts are qualitatively skillful for a variety of severe weather hazards at different times of the year. This project will formally evaluate the skill of the GEFS-MLP forecasts across Canada using data from the Northern Tornado Project and Norther Hail Project and quantify how transfer learning approaches can leverage existing AI forecast systems for different geographical regions.
Desired skills: Python programming, forecast verification, severe weather forecasting
Most applicable majors: Meteorology, computer science, or related field
5. Convective Cloud Evolution near Coastal City Houston, TX
Mentors: Yongjie Huang (CAPS) and Dhwanit J. Mise (CAPS & CIWRO)
Description: Convective clouds are vital to Earth’s weather and climate, as they transport matter and energy through the troposphere and influence large-scale atmospheric patterns. However, the sensitivity of convective clouds to aerosols and meteorological factors, and the lack of high-resolution observational constrains, introduce significant uncertainties in numerical weather perdition and Earth system models. The TRACER field experiment, conducted near Houston, Texas, provides a unique dataset of isolated convective cells under diverse meteorological and aerosol conditions. This study leverages TRACER observations to investigate how environmental factors affect convective cloud properties and their evolution, ultimately enhancing our understanding of convective cloud evolution under various environmental conditions.
Desired skills: Experience in programming
Most applicable majors: Atmospheric sciences, Meteorology, or related fields
6. Wildfire risks and their environmental controls over the US
Mentors: Xiaodong Chen (OU/SoM & OU/CEES) and Wenjun Cui (OU/CIWRO)
Description: Wildfire is a significant global natural hazard and has been responsible for billion-dollar losses in the US each year. Improving our skills in predicting wildfires can lead to better preparedness and emergency response during these challenging events. However, a lack of high-quality data has always been an issue as most wildfires occur in mountainous regions, where high-resolution modeling is needed to reflect the local hydro-meteorological conditions accurately. Fortunately, this gap is being filled by the recent advances in high-resolution atmospheric/climate modeling. In this study, we will:
- Use some recently produced regional climate modeling data to derive the fire risk metric, Fire Weather Index (FWI), over the contiguous US (CONUS) during 1981-2020.
- Evaluate the connections between FWI variability to several circulation indices (like El Nino and Southern Oscillation, Madden-Julian-Oscillation).
Through this hands-on work, you will gain experience in developing codes to process large-amount climate model data, make statistical analyses and visualizations. Meanwhile, you will also get some first-hand wildfire risk information in our region!
Desired skills: experience with or interest in data processing in Python/R/Matlab, basic working knowledge in Linux.
Most applicable majors: (anyone interested in atmospheric science, natural hazards, or data science)
7. Lightning in severe weather and operations
Mentors: Dr. Vanna Chmielewski (NOAA/NSSL) and Dr. Sarah Stough (OU/CIWRO)
Description: Lightning flashes are a common occurrence in a wide variety of storm modes, which is not only a risk on its own, but it can also give us information on overall storm intensity. We have several different methods available to monitor lightning from satellites to long-range ground networks, but they monitor different parts of the lightning process and provide different information. This project focuses on how we can use that different information to inform us of the complete storm state. Depending on the interests of the student, this project could focus on intelligently merging the different datasets, examining the different detection scales and flash properties in severe weather events, or comparisons to both mid-level, radar-based analyses and resulting surface impacts in the Propagation, Evolution, and Rotation in Linear Storms (PERiLS) Project.
Desired skills: Some experience or willingness to learn Python and Linux
Most applicable majors: Meteorology, Physics, Computer Science or related fields
Recent REU projects with mentors on this team:
- Investigating Lightning Flash Rates in Tornadic Environments
- Investigation of Lightning and Storm Electrification Processes Using a Phased Array Radar and Lightning Mapping Array
8. Evaluation of flow-dependent biases in convection-allowing models with the Random Forest machine learning technique
Mentors: Aaron Johnson and Xuguang Wang, University of Oklahoma, Multiscale data Assimilation and Predictability (MAP) lab
Description: Convection allowing ensemble (CAE) forecast systems play an important role in operational severe weather forecasting. Since the grid spacing in CAEs is fine enough to directly resolve the bulk characteristics of deep moist convection, a cumulus parameterization scheme is not needed. The CAE forecasts can therefore provide detailed guidance on the timing, location and morphology of convective systems.
At lead times less than 1-day, the performance of such CAE-based forecast guidance can depend strongly on the data assimilation methods used to constrain the initial conditions with observational data. Recent work in the OU MAP lab found that model biases in the short-term forecasts not only contribute to CAE forecast error in general, but also interact with and limit the potential benefit of improved data assimilation methods. For example, Gasperoni et al. (2023) focused on a 2m temperature cold bias that varied both regionally and from day to day. There is a need to better understand the meteorological conditions driving such forecast biases in order to both guide future model improvements and to allow for effective removal of time- and space-varying biases during data assimilation.
A project is available to apply the Random Forest (RF) machine learning technique to the prediction of 2m temperature and dewpoint biases in the short-term (1 hr) forecasts used during data assimilation in the OU MAP lab CAE data assimilation and forecast system, which was run in real time during the 2022 HWT SFE. The goals of the project are to identify the predictors that can relate the background flow pattern to the forecast bias, and to evaluate the performance of the RF in predicting such bias.
Desired skills: experience, or interest in gaining experience, with basic programming in Unix/Linux and Python
Most applicable majors: (any)
Recent REU projects with mentors on this team:
- Regime Dependent Verification and Calibration of a 10-Member Convection-Permitting Ensemble during the 2019 HWT SFE
- Comparing OU MAP Real-Time Convection Allowing Ensemble Forecasts Produced During 2021 & 2022 Using Neighborhood & Surrogate Verification Methods
- Impact of AQUA Satellite Data on Hurricane Forecast: Danielle 2010
9. Qualitative Social Science Research in Meteorology
Mentors: Elizabeth Hurst (OU/CASR)
Description: The student will choose between multiple qualitative data sets gathered during NOAA funded research on people, places, and things most vulnerable to impacts from various hazardous weather events. The student will choose to analyze interview transcripts or observational fieldnotes.
Option 1: The student will analyze fieldnotes and artifacts gathered at a tornado readiness integrated warning team workshop as part of the NOAA-Vortex-USA study, “Rural Region Readiness”. Research questions should focus on how vulnerabilities are discussed in the workshop by local leaders. GIS databases will be utilized to compare what was discussed to local infrastructure and vulnerability indexes.
Option 2: The student will analyze interview transcripts with National Weather Service (NWS) meteorologists to gain insight into how forecasters become confident in their forecast and the intersection between the science and art of forecasting. This research is part of NOAA-SBES funded research on the Brief Vulnerability Overview Tool.
Desired skills: Must have interest in meteorology and social science!
Most applicable majors: Meteorology/atmospheric sciences; social sciences (i.e., sociology, communication, anthropology, psychology…); geosciences
10. Understanding the Quality of Tornado Warning False Alarms
Mentors: Kristin Calhoun (NSSL), Thea Sandmael (CIWRO/NSSL), and Michael Baldwin (CIWRO/NSSL)
Description: The goal of this project is to better understand the causes of false alarms and determine whether there are patterns that may contribute to them or if they are largely similar to warnings with tornadoes. The primary tool used by National Weather Service forecasters for making warning decisions is WSR-88D radar data. This project will analyze both radar data and values from the machine learning tornado probability algorithm (TORP), focusing on the period leading up to and the time of warning issuance for warnings that were not verified.
Key questions include:
- Are false alarms occurring with storms that have similar rotational strength and probabilities to verified warnings?
- Do false alarms tend to occur more often when storms are farther from the nearest radar?
- Are there common radar quality issues, such as sidelobe contamination, contributing to false alarms?
These statistics will be compared with similar radar data from missed tornadic events, and, if time permits, further analysis will explore these statistics relative to hits across the EF scale.
Desired skills: very basic python or desire to learn
Most applicable majors: Meteorology/Atmospheric Science, Mathematics, Statistics, Geography
Recent REU projects with mentors on this team:
- Week 2 Severe Weather Forecasts: What Should We Predict?
- Characteristics and Statistics of Tropical Cyclone Tornadoes and Warning Skill
- Comparing Road and Bridge Surface Temperatures During the Winter Season in Ohio
- An Investigation into the Characteristics of Lightning as it Interacts with Towers
- Do Forecasters Add Value to Machine Learning Algorithms of Cloud-to-Ground Lightning?
- A Comparison of Lightning Flashes Observed by Ground-Based Detection Networks and Video
- Impact of Earth Networks Lightning Data and Dangerous Thunderstorm Alerts on Forecasters' Warning Decision and Confidence
- Verification of Earth Network's Dangerous Thunderstorm Alerts and National Weather Service Warnings
- Evaluation of a Lightning Jump Algorithm with High Resolution Storm Reports
11. Examining Extreme Weather Tendencies in National Parks
Mentors: Jason Furtado (OU/SoM), Emma Kuster (SC CASC), Marcela Loría-Salazar (OU/SoM)
Description:
Desired skills:
Most applicable majors:
Recent REU projects with mentors on this team:
- Synoptic Analysis of Two 14-Day Extreme Precipitation Events in the Northwestern United States
- Preparing to adapt: Are public expectations in line with climate projections?
- Stratospheric Polar Vortex Trends of the Twentieth and Twenty-First Centuries and an Assessment of the CMIP5 Historical Ensemble
12. Aviation Turbulence near Thunderstorms
Mentors: Stacey Hitchcock (OU/SoM)
Description: This project will be related to aviation turbulence near thunderstorms. For more information about the kinds of things I’m working on check out https://journals.ametsoc.org/view/journals/bams/106/1/BAMS-D-23-0142.1.xml.
Desired skills: having taken a coding class is ideal
Most applicable majors: Meteorology/Atmospheric Science, Aviation, aerospace, physics, or similar
13. Rural region readiness: A historical analysis of tornadoes, social vulnerabilities, and infrastructure in the Appalachian region
Mentors: Dr. Harold Brooks (NOAA/NSSL) & Dr. Elizabeth H. Marold (CASR)
Description: As part of a NOAA-Vortex study entitled, “Rural Region Readiness,” a student will analyze tornado trends in the Appalachian Region and compare those trends to available historical data and GIS databases covering local impacts, such as local infrastructure and social vulnerabilities. Historical and cultural contexts will be taken into account.
Desired skills: experience with GIS databases and ArcGIS, must have an interest in social science and meteorology
Most applicable majors: Meteorology/atmospheric sciences; social sciences (i.e., sociology, communication, anthropology, psychology…); geosciences
Recent REU projects with mentors on this team:
- Tornado Warnings, Lead Time, and Death
- Deriving Population Exposure Fatality Rate Estimates for Tornado Outbreaks Using Geographic Information Systems (GIS)
14.
Mentors:
Description:
Desired skills:
Most applicable majors:
Recent REU projects with mentors on this team: