Collaborative research: advanced data assimilation applied to the study of nocturnal thunderstorms

The National Science Foundation Division of Atmospheric and Geospace Sciences awarded $174,789 to the University Corporation For Atmospheric Research (NCAR) on August 15, 2026, under the Geosciences program (CFDA 47.050) to develop and apply advanced data assimilation techniques to improve forecasting of nocturnal thunderstorms.

The research focuses on implementing novel non-Gaussian data assimilation algorithms to incorporate ground-based remote sensing measurements from boundary layer profilers into numerical weather prediction models. Key objectives include determining which observed temperature, moisture, and wind variables are most critical for predicting nocturnal convection initiation, evaluating whether quantile conserving ensemble filters can better constrain unobserved variables and reduce model imbalances, and assessing how synoptic forcing and initiation mechanisms influence the predictability of these high-impact weather events. The work will support graduate student workforce development.

Performance occurs in Boulder, Colorado, with a completion date of July 31, 2029. No sub-awards are planned for this project.

Abstract

Thunderstorms that develop at night, also known as nocturnal convection, are a difficult forecasting challenge. As the surface cools after sunset, the temperature and moisture in the lowest layer of the atmosphere can sort into complex layers, which can either favor or suppress the potential for thunderstorms. This project will apply new techniques for incorporating ground-based remote sensing measurements into numerical models to better understand nocturnal convection and whether additional measurements can help forecasters predict these high-impact weather events. The project also will contribute to workforce development through the training of a graduate student. This project will focus on data assimilation (DA) techniques applied to the problem of simulating nocturnal convection initiation (CI), particularly related to the implementation of novel non-gaussian DA algorithms for boundary layer profiler data. Key questions to be addressed include the relative importance of different observed temperature, moisture, and wind variables for estimating the nocturnal CI environment, the ability of quantile conserving ensemble filters (QCEF) to better constrain unobserved variables and reduce model imbalances, and the dependence of nocturnal CI predictability on the underlying synoptic forcing and initiation mechanisms. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the foundation's intellectual merit and broader impacts review criteria.- subawards are not planned for this award.