a national resource for computational science education

HOME BWSIP Shodor Blue Waters

View Position
Position TitleTornado Outbreak Modeling Student Experience
SummaryA prior Blue Waters intern has analyzed the importance of lead time on non-tornadic severe weather outbreak uncertainty and developed an error function representing the modification of uncertainty in simulation output for non-tornadic outbreaks. In this project, the student will be tasked with developing the same error growth function for a series of major tornado outbreaks. Upon completion, results from the first internship will be compared with results from this project to assess error growth as it relates to lead time in tornadic and non-tornadic severe weather outbreaks. Such knowledge is of utmost importance for severe weather forecasters issuing forecasts for outbreaks multiple days in advance and deals directly with limitations of medium-term predictability of outbreaks.
Job DescriptionThe goal of this project is the development of an error growth function for forecasts of tornadic severe weather outbreaks using a high-resolution mesoscale model. The student will be working with the WRF-ARW forecast model at a high spatial (12 km) and temporal (time steps of 30 seconds) resolution for 5 major tornado outbreaks. Each outbreak will be simulated in initial condition ensemble mode, yielding 10 forecasts for each case at 5 varying lead times to the outbreak (24-hour, 48-hour, 72-hour, 96-hour, and 120- hour), in an effort to ascertain the sensitivity of the spread in the 10-member ensemble to lead time. This spread will be used to formulate an error growth function by lead time for numerous common severe weather diagnostic variables. Further, the student will use the output to determine spatial uncertainty metrics in the results using the R software. This project will build upon results from a previous Blue Waters internship studying non-tornadic outbreaks and allow for a direct comparison of each outbreak type in the context of spatial and temporal uncertainty. Ultimately, this project has the potential to help forecasters with outbreak forecasts by providing additional evidence as to the limit of the forecast ability of severe weather outbreaks at lead times in excess of a day and addresses the challenge of medium-term forecasting of outbreaks.
Use of Blue WatersThe Blue Waters supercomputer is of utmost importance for this project. High resolution simulations of outbreaks require significant computational resources, and the scope of this project will include 5 outbreaks, 10 stochastically perturbed initial conditions for each outbreak, and 5 lead times tested, leading to a total of 250 simulations required. This number of simulations requires powerful high performance computing resources. Additionally, large disk space on the order of 10 TB will be needed for storage and analysis of the resulting output.
Conditions/QualificationsThis position is intended for undergraduate meteorology students at Mississippi State University. Basic programming experience is essential for success in this position. The student should have limited supercomputing experience to gain the most from the opportunity. Prior work in the Linux environment will be helpful. The student should have a willingness to learn to use the WRF model and have a solid work ethic, as well as good time-management skills.
Start Date05/01/2017
End Date04/30/2018
Department of Geosciences
Mississippi State University
Mississippi State, MS 39762-5448
Caroline MacDonald