Research 2

Broadly, my interests are in physical oceanography and biogeochemistry. I use statistical methods paired with models, observations, and theory to gain dynamical information about the upper and coastal ocean.


Biases in surface drifter statistics

ROMS_500m_L_C_W_Trajectories_short_745_785

One component of my PhD work focuses heavily on the impacts of using semi-Lagrangian drifters to estimate Eulerian structure functions. This involves using ROMS simulations at 500m resolution alongside the LTRANS v2b. particle tracking algorithm to isolate the biases in second and third order statistics of the northern Gulf of Mexico. Above is a gif showing a convergent ‘zipper’ structure in the upper right corner, as well as a persistent eddy southeast of the Mississippi river outflow area sampled by the drifters. These drifter trapping features prevent the floats from adequately sampling the entire domain, and the results of our work show clear biases at the submesoscale, but agreement with the Eulerian statistics at scales above 10km.

Accepted Publications: 

Pearson, J., Fox-Kemper, B., Barkan, R., Choi, J., Bracco, A., & McWilliams, J. (2019). Impacts of convergence on structure functions from surface drifters in the Gulf of Mexico. Journal of Physical Oceanography., 0, 0
https://doi.org/10.1016/j.physd.2017.02.004

 


Traffic Modeling

Screen Shot 2018-08-15 at 2.34.35 PM.png

My research interests also include using data assimilation schemes to improve macroscopic traffic models and estimate difficult parameters to observe. My undergraduate research at Brown’s REU in Dynamics & Stochastics focused on utilizing ensemble Kalman and particle filters to provide more accurate estimates of traffic flow states. This collaborative effort developed these schemes to incorporate both Eulerian and Lagrangian traffic flow observations, as well as estimate key parameters. Above is a figure displaying the root-mean-square-error of our state estimates over time  with and without  data assimilation techniques. When we incorporate  both Eulerian and Lagrangian observations, denoted in green, we find a substantial reduction of error regardless of the data assimilation scheme implemented.

Publications: 

Xia, C., Cochrane, C., DeGuire, J., Fan, G., Holmes, E., McGuirl, M., Murphy, P., Palmer, J., Carter, P., Slivinski, L., & Sandstede, B. (2017). Assimilating Eulerian and Lagrangian data in traffic-flow models. Physica D: Nonlinear Phenomena, 346, 59-72. https://doi.org/10.1016/j.physd.2017.02.004