Purvi Goel

I am working on a general-purpose computational framework for interactive exploration of massive quantities of spatiotemporal data collected from ensembles of large simulations like climate models. Simulations are often governed by input parameters which must be carefully tuned to produce specific or accurate results. Scientists lack a general-purpose framework to examine outcomes of simulations to determine the best set of input parameters and must examine terabytes to petabytes of data by hand. The goal of my proposed tool is to enable analysts to interactively sift through and visualize members of simulation ensembles, compose spatiotemporal queries directly on simulation data to quickly narrow down the search space, and more easily deal with the massive amounts of data produced in the process of tuning climate models.