Summary of Working Paper No. 125-1999

I.6.3 “Prediction of Ice Conditions along the NSR Using Satellite Data”

By Chang Kyu Rheem and Hajime Yamaguchi, University of Tokyo, Japan.

A new system for sea ice condition prediction has been develiped. The SSM/I satellite ice distribution data is used as an initial data for prediction computation by linking the SSM/I data with the prediction program. The SSM/I data are in digital form and can be obtained almost real time. The prediction computation is made by a dynamic sea ice model, Distributed Mass / Discrete Floe model. The computation time takes only 20 minutes for eight days simulation for the area of 1000 km, the Sea of Okhotsk. It takes about 4 hours for the whole Arctic Ocean.

The Distributed Mass / Discrete Floe model is a numerical model proposed for the computation of pack ice movement. This model possesses the advantages of both the continuum and the discrete element models: it can express the discrete nature of pack ice which is difficult by using a continuum model, and requires a much shorter computation time than a discrete element model. In the present work, ice concentration dependency of the form drag of an ice floe is newly taken into account. As the result, the over estimation of ice zone has been improved and the prediction has become more similar to observations.

The present system has been applied for the sea ice motion predictions in the Sea of Okhotsk and the Arctic Ocean. The present system is still in the developing stage. More improvements of the computation model and scheme are needed, and also more comparisons with the observation data are required by acquiring more natural condition data. The present results, however, have showed favourable agreement with the observations, demonstrating the high potential of the present system.