Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Renewable Natural Resources

First Advisor

William J. Wiseman, Jr


Present-day, operational, upper-ocean, thermal-structure forecast models consist of mixed-layer models with local wind-generated horizontal and vertical advection. To extend their applicability into dynamically active regions, e.g. western boundary current regions, the next generation models are envisioned to include mesoscale advection provided by high horizontal resolution circulation nowcast and, eventually forecast models. In this study, I consider the impact of this additional component of advection in a representative dynamic ocean region. I perform four experiments using a modified version of an operational, upper-ocean, thermal prediction model. Each of the four experiments consists of a series of daily, 72-hr-duration, upper-ocean hindcasts. They were each conducted for four weeks during the warming season in the Sea of Japan. The first experiment uses an Nx1 dimensional mixed layer model with no horizontal processes included. The second experiment uses the same model with the addition of horizontal diffusion and local wind-generated horizontal and vertical advection. This model is comparable to present-day operational models. The third experiment repeats the second with the addition of a fixed geostrophic component to the horizontal advection. The fourth experiment allows daily variation of the geostrophic component through each three day forecast. A suite of statistical measures applied to the results indicates a small but statistically significant increase in forecast skill due to the addition of the nowcast mesoscale advection. The additional analysis of a representative individual forecast strengthens this result. The statistical plus individual analyses together lead to three conclusions. First, the addition of geostrophic flow can have a statistically significant impact, especially in frontal regions. Second, global statistical measures alone are not sufficient model comparison criteria since they can mask specific regions or times of significant change. Third, the use of forecast mesoscale circulation in future upper-ocean thermal forecast models will require care due to the potential for artificial cross-frontal advection.