Policy-SH2
RMNP-divideD-IMG_9423
IMG_1590
AstridLaughing
2022 Draft Project Descriptions

Last Updated: March 24, 2022

Keep in mind that research projects can change quite a bit and that is part of the nature of research.

These are linked to a recent REU project for that mentor, when possible. That will give you a peak at 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 don't hesitate to ask.

 

The following projects will be funded through the NWC REU:

 

1. Investigating the Variability in the Number of Tornadoes Among Landfalling Hurricanes

Mentors: Dr. Ben Schenkel (OU/CIWRO)

Description: There is an order of magnitude variation in the number of tornadoes spawned by hurricanes with similar landfall locations and intensities. Recent work has suggested that both the hurricane and its environment may be key in determining the number of tornadoes spawned. However, we lack a complete understanding of the factors controlling tornado variability in hurricanes, particularly for periods with tornado outbreaks. Hence, this project will investigate the factors that are associated with enhanced numbers of tornadoes in landfalling hurricanes.

Desired skills:Programming experience is preferred, but not required

Most applicable majors: Meteorology, science, or mathematics

Recent REU projects with this mentor:

 

2. Honey, I Shrunk The Forecast: What Severe Thunderstorm Forecasts Might Look Like on Different Time and Space Scales

Mentors: Dr. Harold Brooks (NOAA/NSSL) & Dr. Burkely Twiest Gallo(CIWRO/SPC)

Description: Currently, the probability of severe thunderstorms in the Storm Prediction Center’s convective outlooks is expressed as the probability of a hazard within ~40 km of a point. This probability can also be thought of as the fractional coverage of storms (the fraction of roughly 80 km square boxes with a storm during a 24 hour period). As we explore ways to produce probabilistic forecasts on finer space and time scales, we run into a problem in defining things consistently across a range of scales. As a simple example, the probability of a particular square meter being hit by a severe thunderstorms during a particular second of the day is extremely small even in the vicinity of a large, violent tornado, while the probability of a tornado occurring somewhere in the US during the calendar year is 1. For operational forecasting, we care about forecasts on time scales of minutes to days and space scales from a few kilometers to a few thousand kilometers. This projects looks at how the coverage of events on a grid changes as the size of the grid in time and space changes. This has importance for how forecasts might be made in a meaningful way in the future, how we evaluate forecasts, and how we can tell if forecasts are getting better. As forecasts shift to be increasingly probabilistic at smaller scales than the convective outlook, the theoretical underpinnings of said probabilities become increasingly more crucial. Some of the underlying theory we will test with this work may already be expressed in current operational forecasts (e.g., the difference in predictability at watch time and space scales vs. warning time and space scales). We’ll look at how artificial distributions of storms might be described on different scales. We’ll also look at distributions of reports on real storm days to see how they compare to the distribution from the artificial data. This can provide a foundation for the future of severe thunderstorm forecast, at many different scales of interest.

Desired skills: Some programming and UNIX experience preferred Most applicable majors: Meteorology, computer science

Most applicable majors: meteorology or a related field

Recent REU projects with this mentor:

 

3. Urban and Rural Air Pollution in the US from Satellite Measurements

Mentors: Dr. Ian Chang (SoM), Dr. Lan Gao (SoM), & Dr. Jens Redemann (SoM)

 Air pollution in the US stems from various socioeconomic activities and natural phenomena. Air pollution can cause heart and lung health, resulting in premature death. Small pollutant particles from automobile exhaust and power plants can penetrate deeply into lung tissues and the bloodstream. Approximately 68 million tons of pollution were emitted into the atmosphere in the United States during 2020. Their distributions greatly differ both spatially and temporally. To efficiently examine air pollution patterns over a large spatiotemporal scale, satellite remote sensing is required. The student will study diurnal and seasonal variations of air pollution in terms of concentrations and chemical compositions of a chosen region from novel satellite measurements. Inter-comparison of satellite and an evaluation of satellite against ground-based measurements are done to evaluate data consistency and address their measurement shortcomings. The mentors will offer guidance to process, manipulate, and analyze various satellite data.

Desired skills: Basic MATLAB or other programming experience

Most applicable majors: Meteorology or related field

Recent REU projects with this mentor:

 

4. Analysis of Tornadic Direct Hits and Near Misses of Mesoscale Observations Stations (Project assigned)

Mentors: Dr. Brad Illston (Oklahoma Mesonet) and Megan Schargorodski (Kentucky Mesonet)

Description: Kentucky and Oklahoma both operate world-class mesoscale networks (e.g., Mesonets) of atmospheric surface observations. Data from these networks provide valuable information to forecasters, numerical models, and emergency managers during severe weather. Over the past 25 years, eight stations have been directly hit by tornados in Kentucky and Oklahoma with many other stations recording near misses. This study will analyze high temporal resolution observations from each station to compare each event to determine similarities and differences. Additionally, analysis of the characteristics of near-misses of tornadoes to mesoscale in-situ observing stations will be conducted. These research quality weather stations provide a unique insight into the conditions occurring during one of the most extreme weather events on the planet.

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

Most applicable majors: Meteorology or related field

Recent REU projects with this mentor:

 

5. Building a Climatology of Snow Squall Conditions

Mentors: Dr. Andrew Rosenow (CIWRO/NSSL) & Dr. Heather Reeves (CIWRO/NSSL)

Description: A "snow squall" is a quick onset burst of snow that causes a rapid loss of visibility, which can be hazardous to drivers. Snow squalls have quickly become a focus of wintertime operations in the past five years with the introduction of Snow Squall Warnings by the National Weather Service. However, what constitutes a "snow squall" is somewhat ill-defined. This project will look back over data from Automated Surface Observation System (ASOS) sites to determine the historical location and frequency of sudden, blizzard-like conditions to better understand the threat these events pose to the traveling public. This snow squall climatology will help bring to light the types of meteorolgical phenomena that lead to snow squall conditions.

Desired skills: Familiarity with Python (or other programming experience) and basic UNIX commands

Most applicable majors: Meteorology or related field

Recent REU projects with this mentor:

 

6. Investigating Extended Periods of Severe Thunderstorms and Tornadoes

Mentors: Dr. Kim Hoogewind (CIWRO/NSSL) & Dr. Harold Brooks (NOAA/NSSL)

Description: Extended periods (1–2 weeks) with multiple days of severe thunderstorm and tornado events are possible “forecasts of opportunity” for making skillful predictions of severe weather at lead times of two weeks or longer. A recent example includes the second half of May 2019, and a much more significant example occurred during 14-27 April 2011 period. Understanding these relatively rare events and their potential linkages with modes of subseasonal climate variability remains difficult, in large part due to the limited temporal record of severe weather observations and 3D atmospheric data. Recently, version 3 of the 20th Century Reanalysis (20CRv3) has been released which provides an estimate of the state of the atmosphere on sub-daily time scales back into the early 19th century. In this project, the student will explore methods to identify active severe weather periods, examine their climatological characteristics, and analyze associated large-scale environmental conditions using 20CRv3 data.

Desired skills: some Python programming experience, familiarity with gridded datasets, basic statistical knowledge

Most applicable majors: Atmospheric science, meteorology, physics

Recent REU projects with this mentor:

 

7. Analyzing intra-supercell warning performance

Mentors: Dr. Matt Flournoy (OU/CIWRO) & Dr. Kenzie Krocak (CRCM)

Description: This project will build off of a prior REU project to examine tornado and severe thunderstorm warning performance along tornadic supercell paths. This research is motivated by the fact that, although our basic understanding of processes influencing tornado behavior in supercell thunderstorms has increased during the last several years, tornado warning performance has remained somewhat steady. In particular, this work will examine warning performance before, during, and after tornadic periods of individual supercell thunderstorms during tornado outbreaks. Some relevant questions that will be answered include: how long is a to-be-tornadic supercell tornado-warned prior to the formation of the first tornado? How long may tornado warnings extend on a previously tornadic supercell that has already produced its last tornado? How might these warning traits be related to characteristics of the background environment? The student working on this project will use a combination of tornado-warning datasets, including observed tornado characteristics and warning statistics, and a radar-derived supercell-tornado database developed by a prior REU student (see the link below) to address these questions during the summer REU.

Desired skills: interest in storms and tornadoes, programming experience (Python preferred)

Most applicable majors: Meteorology, a related science, or math

Recent REU projects with this mentor:

 

8. Understanding the Diurnal Precipitation over Western Puerto Rico during CPEX-AW (Project assigned)

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

Description: This project investigates the diurnal evolution of precipitation and atmospheric profiles using MPAS simulations and sounding data collected at Mayagüez, Puerto Rico during CPEX-AW field campaign. The diurnal cycle explains a large fraction of total precipitation variability in the tropics. However, it is often difficult to predict the diurnal precipitation over tropical islands and coastal regions. Therefore, this project will use valuable data collected during the CPEX-AW field campaign to examine the atmospheric conditions leading to the diurnal evolution of precipitation and evaluate its representation in a model.

Desired skills:

Most applicable majors: Meteorology or related field

Recent REU projects with this mentor:

 

9. Assessing the Mesocyclone Characteristics of Miniature Supercells in Landfalling Hurricanes

Mentors: Dr. Addison Alford (OU/CIWRO)

Description: Tornadoes are often produced by so-called “miniature supercells” associated with landfalling hurricanes and have been found to be particularly numerous within the first 25 miles of the coast. It is unclear if the transition of miniature supercells from ocean to land causes changes to the intensity of individual mesocyclones. Recent research suggests that the change in the low-level wind field augments the background vertical wind shear available to miniature supercells, but the response of individual storms has yet to be documented. This project will investigate the storm-scale characteristics, including low-level mesocyclone intensity, of miniature supercells as they move from ocean to land using radaar-based analyses.

Desired skills: Interests in storm-scale dynamics and/or radar analysis; familiarity with Python, a similar language, or basic Unix is welcome but not necessary.

Most applicable majors: Meteorology or a related field

Recent REU projects with this mentor:

 

10. Understanding the Impact of Increased Vulnerability Awareness on NWS and Emergency Manager Decision-Making

Mentors: Dr. Jack Friedman (CASR) & Dr. Michelle Saunders (CASR)

Description: This project examines the impact of increased awareness of vulnerabilities — provided through the experimental Brief Vulnerability Overview Tool (BVOT) — on how NWS meteorologists issue warning products, provide briefings to their emergency management (EM) partners, and shape their messaging on social media and NWSChat throughout the evolution of a severe weather threat. This study draws on data that was collected in experimental conditions as part of NOAA’s Hazardous Weather Testbed (HWT), and it considers when and why operational meteorologists provide increased details in or frequency of messaging due to increased awareness of on-the-ground vulnerable people, places, and things. Data includes both the recording of the meteorology of the severe weather cases as well as audio/visual recordings (and transcriptions) of meteorologists and emergency managers making decisions and communicating with each other throughout the evolution of a severe weather threat — from days in advance, to hours in advance, to storm-on-the-ground. This summer project will involve 1) understanding the meteorology of the events used in the experiments; 2) gaining an understanding of basic social science concerns associated with contemporary operational meteorology; 3) coding decision-making throughout the event to identify patterns; and 4) analyzing what impact increased awareness of vulnerabilities has had on decision-making. Click here for more BVOT information.

Desired skills: Interest in studying the intersection of social science and meteorology. Experience with qualitative coding and/or ArcGIS is an added benefit, but it is not necessary.

Most applicable majors: meteorology, geography, social science, or related field

Recent REU projects with this mentor:

 

11. Analysis of Tornadic Storms Using High-Resolution Radar

Mentors: Dr. David Bodine (ARRC), Rachael Cross (SoM), & Morgan Schneider (SoM)

Description: The dynamics and formation of tornadoes in quasi-linear convective systems (QLCSs) are generally less understood than supercells. This lack of understanding leads to shorter lead times for QLCS tornado warnings which exacerbates the impacts of tornadic QLCSs, especially in the Southeastern US where there are a higher percentage of QLCS tornadoes. Past studies have correlated polarimetric radar signatures (such as ZDR columns) to tornadic intensity in supercells, however these relationships have not been assessed in QLCSs. This project will look at features such as ZDR columns prior to tornado genesis using radar (NEXRAD) data and correlate this with tornadic intensity. Analysis will focus on QLCS events in the Southeast.

Desired skills: Basic coding experience or an interest to learn (Python)

Most applicable majors: Meteorology, mathematics, physics, earth science

Recent REU projects with this mentor:

 

12. Assessing Lidar Scanning Strategies for Wind and Turbulence Retrievals

Mentors: Dr. Josh Gebauer (OU/CIWRO) & Dr. Jeremy Gibbs (NOAA/NSSL)

Description: Doppler wind lidars are often used to obtain wind and turbulence profiles of the atmospheric boundary layer. As lidars are only able to directly measure radial velocity, wind profiles are often obtained using scans that allow for Doppler beam swinging (DBS) or velocity azimuth display (VAD) retrievals and other more advanced scans have been proposed for accurately measuring turbulence. These retrieval techniques rely on the assumption that the environment is horizontally homogenous, which is often not the case, and the amount of error in the retrievals due to this assumption being violated may very between techniques. This project will make use of a Doppler wind lidar simulator and Large Eddy Simulation (LES) output to assess the accuracy of wind and turbulence retrievals using DBS, VADs or other types of scans in different boundary layer flow regimes.

Desired skills: Familiarity with Python; familiarity with basic Unix/Linux would be helpful but not necessary

Most applicable majors: Meteorology or a related field

Recent REU projects with this mentor:

 

13. Investigation of the Large-Scale Drivers of 14-Day Extreme Precipitation in the United States

Mentors: Ty Dickinson (SoM) & Melanie Schroers (SoM)

Co-Mentors: Dr. Jason Furtado (SoM), Dr. Elinor Martin (SoM), & Dr. Michael Richman(SoM)

Description: Despite significant interest from various stakeholders, extreme precipitation events on a range of timescales remain as some of the most challenging hazards to forecast. This project is centered on investigating the large-scale characteristics present before and during subseasonal extreme precipitation events. Using a database of 14-day events in the contiguous United States developed for this research, you will take a deep dive into a select number of case studies to find the set of important dynamical and thermodynamical predictors driving these events. These events offer a unique opportunity to identify synoptic- and planetary-scale precursors and drivers since they occur over the course of two weeks. You will have the opportunity to select a region within the contiguous United States (or regions, time permitting) of your choosing and explore events from various seasons. Previous work performed by our PRES2iP team has developed mean patterns that drive these events, and the current opportunity will enhance these bulk representations and potentially uncover multimodal drivers to extreme precipitation.

Desired skills: Programming experience (Python preferred, but anything is acceptable), interest in exploring long-duration extremes.

Most applicable majors: Meteorology or related field.

Recent REU projects with this mentor:

 

14. Identifying Gauge Observations Influenced by Melting Winter Precipitation

Mentors: Steve Martinaitis (CIWRO/NSSL) & Jackson Anthony (CIWRO/NSSL)

Description: The development of a new temporal gauge quality control process for the Multi-Radar Multi-Sensor (MRMS) system showed the challenges with determining if a gauge observation is truly in error or if there was a meteorological influence to the observation. The specific challenge that arose from the study was determining if a declared false non-zero gauge observation (i.e., the gauge value was non-zero but it was not precipitating) was a result of the thawing of winter precipitation or from a non-meteorological error. This study is a proof-of-concept of identifying which gauges might be influenced by the melting of fallen winter precipitation. The study will use a machine learning algorithm that focuses on winter events over the state of Oklahoma using observation from the Oklahoma Mesonet. The goal of this study is to see if we can accurately identify where thawing can occur and how much liquid can we expect to melt in a gauge per hour. Having a more accurate quality control of data can better help determine if a gauge might have a systematic error; moreover, having a theoretical gauge thaw value could benefit observations when thawing and precipitation occur simultaneously by removing the thaw value from the observation.

Desired skills: Basic Linux/Unix skills, Some programming experience, Basic statistical knowledge

Most applicable majors: Meteorology or related field

Recent REU projects with this mentor:

 

The following projects will be funded through AI2ES:

 

1. Estimating Convective Updraft Characteristics from Radar

Mentors:Dr. Randy Chase & Dr. Amy McGovern

Description: The characteristics of a convective updraft, such as the width of the updraft, have been identified by both the modeling and observational research communities as an important property that is related to tornado genesis and tornado strength. In general, storms with wider updrafts are more likely to produce tornadoes, and likely produce stronger tornadoes. Thus, in an operational environment, it would be beneficial to measure the convective updraft characteristics. This has been done in the past by using updraft proxies, such as the overshooting top area or the differential reflectivity (Zdr) columns. This summer project will leverage machine learning to investigate if the three-dimensional (3D) convective updraft characteristics can be estimated from radar. The project will leverage pre-existing convective allowing numerical weather prediction model data to train machine learning models that will estimate updraft characteristics from the 3D radar reflectivity.

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

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

Recent REU projects with this mentor:

 

2. Using Machine Learning to improve thunderstorm prediction from the NOAA Warn-on-Forecast System

Mentors: Monte Flora (OU/CIWRO & NOAA/NSSL; REU14)

Description coming

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

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

Recent REU projects with this mentor:

 

3. Robust Predictions for Weather Phenomena

Mentors: Dimitris Diochnos

Description: We will study how robust current state-of-the-art models on weather prediction are with respect to noisy data. This may involve scenarios where historical data that are used for training have been corrupted by noise, or it may involve scenarios where current measurements that are fed into models used for predicting weather in the near future are affected by noise. In addition, we will try to explore directions by which we can robustify our models and make better predictions under such adversarial situations.

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

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

Recent REU projects with this mentor: