2023 Project Proposals

Last Updated: March 24, 2023

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 a recent REU project for that mentor, when possible. That doesn't mean the proposed project is like that one, but will give you a flavor of that mentor's research. Likewise, some projects have "most directly applicable majors" listed. This does not necessarily limit who can choose the project, but instead is meant to reveal something of the nature of the project. If you have any hesitation or doubt or questions about a project please ask some questions before ranking it.

The following projects will be funded through this REU (first set of projects).

AI2ES projects are spoken for (they do their own selection).

1. Three possible topics: Geographic/Climate Education, Tropical Climatology, or Renewable Energy

Mentor: Dr. Scott Green (OU/DGES)

The mentor will define the project in conjunction with the student, based on their choice from the three topics.

2. Compound heat waves and air pollution episodes in the urban environment

Mentor: Dr. Chenghao Wang (OU/SoM & OU/DGES)

Description: Extreme heat and polluted air are closely associated with a variety of adverse health outcomes. This is especially true in the urban environment, where unique urban climates shaped by anthropogenic activities often amplify heat stress and pollution level. For example, the heat stress during a heat wave can be intensified by the urban heat island effect, while the air pollution caused by a wildfire event can be enhanced by pollutants emitted from traffic and industrial activities. Even worse, concurrent or compound heat wave events and air pollution episodes can further increase the risk of hospital admissions and mortality, threatening the health and well-being of urban dwellers. This project aims to quantify the compound heat and pollution risks for urban areas with distinct geographical controls across the contiguous United States. The project will leverage multiple high-resolution datasets developed from ground-based observations, satellite remote sensing, and numerical models to investigate the frequency and duration of compound heat waves and air pollution episodes for each urban area. Through comprehensive analyses, we will also identify the potential links between these events with large-scale weather systems and evaluate the urban contributions to these events.

Desired skills: Basic MATLAB or other programming experience

Most applicable majors: Meteorology, atmospheric science, geography, Earth science, or related field

3. What differences exist between tornado warnings associated with versus without false alarms in landfalling hurricanes?

Mentor: Dr. Ben Schenkel (OU/CIWRO & NOAA/NSSL), Dr. Kristin Calhoun (NOAA/NSSL), Thea Sandmæl (OU/CIWRO & NOAA/NSSL), Dr. Addison Alford (NOAA/NSSL)

Landfalling hurricanes frequently spawn tornadoes that often occur with other hazards. However, these tornadoes can be difficult to forecast, with most tornado warnings not being associated with a tornado. A greater understanding of how lightning and radar-derived rotation differs between tornado warnings with and without a verified tornado may help improve forecasts. Hence, this project will investigate the differences in the radar and lightning characteristics of tornadoes associated with versus without false alarms in landfalling hurricanes.

Desired skills: None

Most applicable majors: Meteorology, science, or mathematics

Recent REU projects with mentors on this team:

Do Forecasters Add Value to Machine Learning Algorithms of Cloud-to-Ground Lightning?

4. Verification of an FV3-LAM real-time convection-allowing ensemble forecast system during HWT

Mentors: Dr. Nicholas Gasperoni and Dr. Xuguang Wang (OU/SoM)

Understanding error and bias characteristics of numerical model guidance is important to use it effectively and identify areas of improvement in model system development. Efforts are underway in the US meteorological community to transition to the next-generation Finite Volume Cubed Sphere (FV3) model for all scale applications – from coarse global to high-resolution regional systems. In recent years, developments of the limited area model version (FV3-LAM) has been tested within the NOAA Hazardous Weather Testbed (HWT) experiment for convective-scale analysis and forecasting.

This project will examine the performance of high-resolution FV3-LAM-based 10-member ensemble forecasts of convective systems over the continental US run by the OU Multiscale data Assimilation and Predictability (MAP; http://weather.ou.edu/~map/) Laboratory from spring of 2021 and 2022. The specific area(s) of evaluation in the project will be left open to prospective students’ interests, with guidance from mentors in potential topics. Suggestions include (1) evaluation of ensemble spread of convective systems; (2) comparison of performance from different years of experiments; (3) verification of model-derived storm characteristics (mid-level rotation; severe storm surrogate reports); (4) comparison of model climatology (FV3-LAM vs. WRF). This project offers the student an opportunity to learn how to read, analyze, and visualize numerical model output, as well as various observation datasets including radar-derived products and severe weather reports.

Desired skills: Experience with basic shell scripting and a programming language, preferably Python or NCL.

Most applicable majors: Meteorology or related field

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

5. Applying MYRORSS Azimuthal Shear Observations to the Early Morning QLCS Mesovortices of 27 April 2011

Mentors: Dr. Tony Lyza (OU/CIWRO & NOAA/NSSL) & Dr. Matt Flournoy (NOAA/SPC)

Brief project description: Recently, the tornadic supercells of the 27–28 April 2011 super outbreak were analyzed in detail using the azimuthal shear (AzShear) products in the Multi-Year Reanalysis Of Remotely Sensed Storms (MYRORSS) dataset to document changes in intensity along each supercell’s lifecycle and how those characteristics compare to the evolutions of the supercells, including when they were or were not producing tornadoes, and if they were producing tornadoes, how strong the damage associated with the tornadoes was. While the supercells produced a majority of the damage and fatalities during the outbreak, another high-impact round of tornadoes struck the same areas impacted by the tornadic supercells in the predawn hours of 27 April, about 8–12 hours prior to the supercells. Over 80 tornadoes were associated with an intense quasi-linear convective system (QLCS) across the Southeast U.S., including 21 tornadoes of EF2–EF3 damage intensity. This project will use manually identified mesovortex tracks to develop a database of MYRORSS information for the mesovortices in the early morning QLCS of 27 April 2011, similar to the database developed for the afternoon supercells. Goals of this project will be to test the ability to test for statistical differences in AzShear between nontornadic and tornadic mesovortices and across tornadic mesovortices of different damage intensities, and investigate AzShear differences across different mesovortex structures. The student will gain a valuable background in QLCS tornado understanding while contributing to our understanding of radar detection of QLCS tornado threats and to a more complete documentation of the historical April 2011 super outbreak.

Desired skills: Interest in tornadic storms and some programming experience (preferably Python language)

Most applicable majors: Meteorology (or related major)

Recent REU projects with mentors on this team:

6. Comparison and Impacts of Air Temperature Measurements from Aspirated and Unaspirated Radiation Shields

Mentors: Dr. Brad Illston (Mesonet)

Description: In order to measure air temperature more accurately, temperature sensors are placed within radiation shields to reduce impacts from direct solar radiation. Radiation shields can be either aspirated or unaspirated depending upon power availability at weather stations. Additionally, aspirated radiation shields may operate at differing fan speeds. A comparison facility of a variety of aspirated and unaspirated radiation shields with temperature sensors was installed at a research facility in Norman, Oklahoma. This study will analyze the data collected from this facility to determine implications and characteristics of the measurements from the differing equipment.

Required skills: Basic Unix/Linux, Programming experience (Python preferred)

Most applicable majors:

Recent REU projects with mentors on this team:

7. Investigating Bridge Conditions During Winter Weather

Mentors: Daniel Tripp (OU/CIWRO & NOAA/NSSL), Dr. Michael Baldwin (OU/CIWRO & NOAA/NSSL) & Andrew Rosenow (OU/CIWRO & NOAA/NSSL)

Brief project description: The pavement temperature of a road controls how icy and unsafe travel is during and after a winter storm. The National Weather Service has been testing a new machine learning algorithm that provides guidance on whether or not roads are sub-freezing. The current tool does not discriminate between highways and elevated surfaces such as bridges, which can become icy or snow-covered much quicker than nearby highways. These conditions have led to several accidents and fatalities in winter. This project will involve looking at road sensors and weather observations to understand how well hazardous bridge conditions are captured with our current instrumentation network.

Desired skills: Familiarity with Python (or other programming experience) and basic UNIX commands. Familiarity with machine learning is not required.

Most applicable majors: Meteorology or related field

Recent REU projects with mentors on this team:

8. A Weather-Ready Nation Para Todos: Understanding how Bilingual Communities Respond to Extreme Weather Events

Mentors: Joseph Trujillo-Falcón (OU/CIWRO & NOAA/NSSL), Dr. Justin Reedy (OU Department of Communication) & América Gaviria Pabón (OU/CIWRO & OU/IPPRA)

Known as the fastest growing group in the United States, the Hispanic or Latino population of 59.7 million represents nearly one in five Americans today. Spanish is the predominant language spoken among Hispanic or Latino households, with over 70.6% reporting that they speak it at home. As communities in the United States continue to diversify, life-saving information and infrastructure must address cultural disparities and language barriers experienced during disasters. In this project, you will analyze data coming from a nationwide survey of 1,500 U.S. Spanish speakers that answered questions on extreme weather understanding and response. You will become immersed in bilingual risk and crisis communication theories and apply them to create newfound findings to the weather enterprise. Findings will be forwarded to essential partners in NOAA and NWS headquarters. This summer project will involve 1) gaining an understanding of basic social science and its intersection with underserved communities; 2) using quantitative coding software to conduct statistical analysis on human population data; and 3) understanding how research and policymaking go hand in hand to serve marginalized populations.

Desired skills: Interest in studying the intersection of social science and meteorology. Experience with quantitative coding in SPSS and/or ArcGIS is an added benefit, but it is not necessary. You do not need to speak Spanish to participate in this project, but being bilingual is a plus!

Most applicable majors: Meteorology, Communication, Geography, Political Science, Hispanic Studies, Spanish, or Related Field

9. Understanding the Coastal Precipitation over West Africa during CPEX-CV

Mentors: Shun-Nan Wu, Dr. Naoko Sakaeda, Dr. Elinor Martin (OU/SoM)

Description: This project aims to analyze coastal precipitation and atmospheric profiles using satellite retrievals from IMERG and sounding data collected during NASA CPEX-CV field campaign at Sal, Cape Verde (https://espo.nasa.gov/cpex-cv/content/CPEX-CV). The West African offshore is known to be where the North Atlantic hurricanes originate and start to develop from tropical waves. However, it is challenging to simulate convective activity over this area due to the scarce observations of atmospheric vertical profiles. Therefore, this project will use valuable data collected during the CPEX-CV field campaign to examine the preferred atmospheric conditions for the coastal precipitation and assess their relationships.

Desired skills: Basic programming experience

Most applicable majors: Meteorology or related field

Recent REU projects with mentors on this team:

10. Public Characterization of Tornado Risk

Mentors: Dr. Harold Brooks (NOAA/NSSL) and Dr. Joe Ripberger (OU/IPPRA)

Risk communication in the severe weather domain is rapidly evolving. Forecasters are now considering the probability and potential severity of severe weather events when communicating information about risk. Historically, the categories forecasters use to communicate risk (e.g., high risk, moderate risk) are defined internally, not by the information users (such as the public). This project will flip the script. It will use data from the Extreme Weather and Society Survey to explore public definitions of tornado risk. The exploration will address questions like: How do members of the public balance information about the probability and severity of tornadoes when characterizing the risk of a given threat? What counts as low, moderate, high, or extreme risk? Is a 50% chance of an EF0 or EF1 tornado a high risk? What about a 20% chance of an EF2 or EF3 tornado? Answers to these questions will provide important insight into ongoing conversations about how to improve risk communication in the future.

Recent REU projects with mentors on this team:

11. Evaluating Severe Convective Storms using Rapid-Update Data from Dual-Polarization Phased Array and Conventional Radars

Mentors: Charles Kuster (NOAA/NSSL), Dr. Addison Alford (NOAA/NSSL), Dr. Terry Schuur (OU/CIWRO &NOAA/NSSL), and Dr. Vivek Mahale (National Weather Service Norman)

Previous research has demonstrated potential benefits of rapid-update, single-polarization radar data for issuing severe-weather warnings, but more work is needed to examine the potential benefits of rapid-update dual-polarization radar data. Therefore, this research project will focus on an analysis of tornadic and nontornadic supercells as well as downburst-producing storms using rapid-update dual-polarization radar data collected by two research radars in central Oklahoma. One radar, KOUN, is a conventional radar with a dish antenna, and the other, the Advanced Technology Demonstrator (ATD), is the first dual-polarization phased array radar (PAR) developed for weather applications. We aim to examine potential operational benefits of various dual-polarization signatures, such as ZDR arcs, ZDR columns, and KDP cores observed with rapid volumetric updates. This analysis will include comparing rapid-update and traditional-update radar data using a variety of approaches, quantifying the differences in radar signatures of severe and nonsevere thunderstorms (i.e., “null cases”), and examining rapidly-evolving trends in these radar signatures that might provide helpful information to warning forecasters. Opportunities may also exist for becoming immersed in ATD and KOUN radar data collection activities and shadowing a National Weather Service forecaster.

Desired skills: Knowledge or interest in meteorology including severe weather radar signatures (e.g., ZDR column, tornado vortex signature), basic computer programming skills (any language)

Most applicable majors: Meteorology or a closely-related field

Recent REU projects with mentors on this team:

12. Investigation on polarimetric radar signatures of lightning and storm electrification processes using rapid-scanning radars

Mentor: Dr. David Schvartzman (OU/SoM & OU/ARRC), Vanna Chmielewski (NOAA/NSSL), Tian-You Yu (OU/ARRC), David Bodine (OU/ARRC), and Mike Stock (OU/CIWRO)

Mechanically scanning radars traditionally used to investigate electrification processes in thunderstorms insufficiently sample the evolution of the atmosphere in space and time compared to the time scale of lightning processes. In this project, we will analyze data from a rapid-scanning dual-polarization phased array radar (PAR) with simultaneous vertical sampling (imaging) over electrified regions. These observations are critical to advance our fundamental understanding of the microphysical and kinematic processes surrounding electrification in a cloud, and the relationship of lightning channels to radar-observed features. High temporal resolution data are critical to characterizing the structures which promote lightning initiations near storm updrafts, describing the interplay of lightning channels with microphysical features such as ice crystal alignment signatures, and exploring the direct scattering of radar signals from channels within radar-sampled volumes. Detections from lightning mapping arrays will be used to correlate data to lightning discharges at fine time scales.

Desired skills: Interest in radar meteorology and lighting/electrification processes in thunderstorms, and computer programming skills (Python or MATLAB preferred)

Most applicable majors: Meteorology, electrical engineering, or a closely-related field

Recent REU projects with mentors on this team:

13. Assessing How Timing Information is Communicated During Severe Weather Events Using New Storm Prediction Center Experimental Outlook Products

Mentors: Dr. Michelle E. Saunders (OU/CASR), Dr. Elizabeth H. Hurst (OU/CASR), and Dr. Daphne S. LaDue (OU/CAPS)

This project will examine how new Storm Prediction Center (SPC) experimental Outlook timing tools/graphics were used by NWS forecasters to communicate timing information about severe weather events to Emergency Managers during a NOAA Hazardous Weather Testbed (HWT) experiment. The student researcher will work on identifying if there were any differences in the content of forecasts, messaging to the public, and formal briefings provided to deep core partners associated with including new SPC experimental Outlook products. This project will involve: 1) gaining an understanding of basic social science concerns associated with contemporary operational meteorology; 2) performing text analysis of forecaster decision-making during the HWT experiment to identify patterns of communication; 3) analyzing which tools/graphics were communicated during the experiment, what communication channels were used, identifying reasons for why specific tools/graphics were communicated, and understanding how the tools/graphics were used/incorporated during the forecast process.

Desired skills: Interest in learning how to use qualitative methods and analysis, no prior experience necessary.

Most applicable majors:

Recent REU projects with mentors on this team:

14. Who does what and when as tornadoes occur: validating and visualizing data from Tornado Tales web application

Mentor: Dr. Justin Sharpe (OU/CIWRO & NOAA/NSSL)

Tornado Tales is a new Citizen Science tool aimed at soliciting responses from individuals who experience tornadoes about their protective actions. The survey (https://apps.nssl.noaa.gov/tornado-tales) has been collecting data on tornadoes that have happened in the past as well as more recently with the recent Norman tornadoes and Kansas City tornadoes last Spring.

Currently this outputs the survey responses to an Excel spreadsheet where we can see what actions were taken in watch and warning phases, what structures individuals were in and how safe they felt. There is also a space on the form where individuals can share their story.

We would like to work with a student to verify tornado reports against real events (using SPC database) and then analysing the current data in terms of the responses recorded, so that we could start to see what actions were/were not taken and at what timescales. We are currently building a Tornado Tales 2.0 in which it will be easier to sort this data, iterating on the first version but we don't want to lose what we have recorded thus far.

Once this initial phase has been completed, the student would be tasked with creating interesting data visualizations that we could then share with the public as well as co-authoring a technical report and potentially a manuscript on the overall results this far. This would be dependent on progress with the data sense-making and analysis.

You would be working with the project lead for Tornado Tales, a Research Scientist who coordinates social science research for VORTEX-USA. His expertise lies in disaster geography, social science and communications.

Desired skills: Data sorting, collation and visualization, understanding of tornado watch and warning continuum, strong communication skills

Most applicable majors: Geography, data science, social science, communications

15. Storm Displacement Errors in the Warn-on-Forecast System

Mentors: Dr. Derek Stratman (OU/CIWRO & NOAA/NSSL) and Dr. Corey Potvin (NOAA/NSSL)

A feature alignment technique for storms (FAT-S) has been developed for forecast systems like the experimental Warn-on-Forecast System (WoFS) to correct storm displacement errors in forecasts prior to data assimilation. Storm displacement errors can result in suboptimal assimilation of observations and thus negatively impact analyses and forecasts of storms. However, before testing the FAT-S with the WoFS, a more thorough understanding of current storm displacement errors within the WoFS’s forecasts is needed. The WoFS is a regional storm-scale ensemble data assimilation and forecast system designed to provide short-term (<6 h) probabilistic guidance of severe weather to end users, such as NWS forecasters. This study will use several years of WoFS forecast output to explore storm displacement errors using object-based techniques with the goal of understanding how storm displacement errors develop and evolve during the free forecast periods.

Desired skills: Experience with basic Unix/Linux commands and a programming language (preferably Python)

Most applicable majors: Meteorology or related field

Recent REU projects with mentors on this team:

16. Testing Precipitation Nowcasting Capabilities with Tropical Cyclones

Mentors: Steven Martinaitis (OU/CIWRO & NOAA/NSSL), Jackson Anthony (OU/CIWRO & NOAA/NSSL), and Dean Meyer (OU/CIWRO & NOAA/NSSL)

New efforts are underway within the Multi-Radar Multi-Sensor (MRMS) system to create a 60-minute precipitation forecast to help with flash flood prediction and improve flash flood warning lead times. The MRMS team is exploring ensemble nowcasting methodologies within the open-source pySTEPS framework. Current nowcast testing for the MRMS system are focused on typical flash flood events to help guide the setup of the nowcasting system; however, the nowcasting project must also work with events that have complex motion vectors, such as tropical cyclones. This project will examine and evaluate the performance of two ensemble-based nowcasting techniques along with various iterations of the nowcasting parameters for two recent tropical cyclone events. The project will specifically evaluate forecast instantaneous precipitation rates to see how they compare to observed rates.

Desired skills: Programming experience (Python preferred)

Most applicable majors: Meteorology or related field

Recent REU projects with mentors on this team:

The following projects will be funded through AI2ES:

AI2ES-1. Robust predictions for weather phenomena

Mentors:

Description: We will explore directions of semi-supervised learning in order to enhance the predictive accuracy of machine learning models that are trained on meteorological data (most likely: winter-weather data). This exploration may involve tabular data, image data, or both, depending on the direction that we will decide to pursue. Our approach will have an eye on methods related to trustworthy machine learning and for this reason we may touch upon issues related to explainability or adversarial robustness of the learnt models.

Desired skills: Python programming, an interest in using machine learning for weather prediction, understanding basic Unix/Linux

Most applicable majors:

Recent REU projects with this mentor:

AI2ES-2. Forecasts of opportunity for extreme weather over North America

Mentors:

Description: This project will employ artificial intelligence (AI)/machine learning (ML) methods to assess periods of enhanced forecast skill (forecasts of opportunity) at subseasonal to seasonal timescales related to weather extremes. Such periods can arise when the climate signal from processes like the El Niño–Southern Oscillation and the Madden-Julian Oscillation is large compared to unpredictable weather noise. Depending on the interest of the student, this project can focus on model development and evaluation or the use of explainable AI to better understand the physical processes associated with forecasts of opportunity.

Desired skills: Python programming, an interest in machine learning, basic knowledge of weather and climate dynamics.

Most applicable majors:

Recent REU projects with this mentor:

AI2ES-3. Estimating Convective Updraft Characteristics for the Storms within the GridRad Severe Archive

Mentors: Dr. Randy Chase (AI2ES) & Dr. Cameron Homeyer (SoM)

Desired skills: Basic Unix/Linux, programming (python preferred)

Most applicable majors: Meteorology, Computer Science, Physics, Applied Mathematics

Recent REU projects with this mentor:

2022 Title: “Estimating Convective Updraft Characteristics from Radar”; Student: Kayla Hoffman
2022 Title: “Tornado nowcasting with Machine Learning”; Student: Alex Nozka
2021 Title: “Hail nowcasting using machine learning”; Students: Jordan Robinson and Lydia Spychalla