1. Comparison of 4DVAR and LETKF in Assimilating JPSS-Derived Sea-Surface Temperature in the Chesapeake Bay Operational Forecasting System
Funding source: NOAA/University of Maryland
In an effort to improve operational forecasting the Chesapeake Bay we are comparing two state-of-the-art data assimilation systems for use on the NOAA Chesapeake Bay Operational Forecast System (CBOFS) model: 4D-Var and Local Ensemble Transform Kalman Filter. We are testing both systems using simulated data and real satellite sea-surface temperature (SST) data provided by the Visible Infrared Imaging Radiometer Suite (VIIRS).
2. Intramural Forecasting of Cardiac Electrical dynamics
Funding source: NSF
In an attempt to improve understanding of the dynamics of electrical waves in the heart during arrhythmias, particularly in the interior of the tissue, we are coupling a Local Ensemble Transform Kalman filter (LETKF) based data-assimilation system with numerical models of electrical wave propogation. Simulated data is being used for initial testing, before moving onto real data obtained using optical mapping techniques. This work is done in collaboration with Elizabeth Cherry at RIT and Flavio Fenton at Georgia Tech.
3. DDDAS for Object Tracking in Complex and Dynamic Environments (DOTCODE)
Funding source: AFOSR
We are using dynamic data-driven application system (DDDAS) principles to control an adaptive optical sensor for the purpose vehicle tracking. The DDDAS system involves object detection, an adaptive set of vehicle motion models, a background identification and elimination scheme, data assimilation from a Gaussian sum filter, and intersection and road network detection from crowdsourced data. For our adaptive sensing data we are using high resolution video generated by DIRSIG. This work is done in collaboration with Anthony Vodacek at RIT.
1. Mars Atmospheric Data Assimilation and Climate Reanalysis
Funding source: NASA
We coupled the Local Ensemble Transform Kalman filter to atmospheric model of the Martian atmosphere to improve understanding of atmospheric dynamics. We used simuilated data to test the system and then moved on to atmospheric profiles derived from the Thermal Emission Spectrometer on the Mars Global Surveyor spacecraft. We also have used the model to explore bred vectors in the Martian atmophere and compute energy transfer locations and mechanisms. This project is still ongoing and is being led by Steven Greybush at Penn State. It is also a collaboration with my Ph.D. advisor, Eugenia Kalnay, at University of Maryland, as well as John Wilson at the Geophysical Fluid Dynamics Laboratory and Ross Hoffman at AER.
2. Derivation of Chesapeake Bay Surface Salinity from Satellite Radiances
When I was a postdoc at Johns Hopkins, I assisted on Ph.D. work done by Erin Urquhart, now at University of New Hampshire, and her advisor Ben Zaitchik. We tested a number of statistical prediction models to see if salinity measurements could be derived from reflectances observed from MODIS and found that saliniy could be predicted with statistically significant value added by the radiances. Erin has extended this work to look for vibrio in the Bay.
Current Graduate Students:
Burak Uzkent (Ph.D.), DDDAS for Object Tracking in Complex and Dynamic Environments
Former Graduate Students (with thesis title and current positions):
Stephen Scorse (M.S.), An Approach for Applying Data Assimilation Techniques for Studying Cardiac Arrhythmias, Epic Systems Corporation.
Jessica Beiter (M.S.), An SIR Approach to Modeling Business Interactions in a Marketplace, Epic Systems Corporation.