DEMONSTRATING HYDRODYNAMIC DATA ASSIMILATION WITH OPENGL ANIMATIONS

We have developed OpenGL animated demonstrations for a hydrodynamic data assimilation project. The project uses the Princeton Ocean Model to simulate wind-driven circulations in Lake Michigan. Hourly advection observations at 10 locations are assimilated into the model under mass conservation and coastal constraints. The demonstration employs a straight-forward particle tracking algorithm and is designed such that (1) it clearly shows both the magnitude and the direction of the advection; (2) it clearly compares the observed and modeled advections at the data sites; (3) it supports zooming, panning, fast forwarding/backwarding, pausing, and on-the-fly adjustment of most parameters that control the appearance of the animated advection field; (4) the demonstrated advection field is switched easily at any time between different scenarios (without assimilation, with assimilation assuming exact data, with assimilation considering measurement error).