Author(s): Robert S. Pritchard; Guang Li; Gordon F. N. Cox
Linked Author(s):
Keywords: No Keywords
Abstract: A simple, blended deterministic-statistical ice motion model is presented. The model requires regular wind and ice velocity observations as input. Wind observations and forecasts are needed as input to the deterministic component of the model. Ice velocity observations are needed as input to the statistical component of the model. The model is intended as an ice forecast tool for offshore operators. We anticipate this model will be useful for short term forecasts up to roughly five days. Ice velocity is estimated and forecast as the sum of a quasisteady free-drift model component that includes wind-driven drift and slowly varying deep ocean currents and a statistical component that minimizes errors between the deterministic components and recent observations. The free-drift model includes effects of sea surface tilt as a geostrophic current balance. The quasi-steady free drift model neglects inertia of the ice and ocean mixed layer and describes the effect of wind forcing. Tidal velocity contributions are also neglected so that daily averages of ice motions are considered. Deep ocean currents are estimated as the average of recent wind-driven model errors. Effects of ice stress divergence and other un-modeled physics are approximated using optimal estimation. This statistical component is estimated from the difference between recent ice motion observations and the deterministic components of the model. Its mean and variance are then held constant for the forecast period. The blended model is more accurate than the free-drift model that comprises its deterministic component for the test case presented here.
Year: 2010