Adding sea ice effects to a global operational model (NEMO v3.6) for forecasting total water level: approach and impact

<p>In operational flood forecast systems, the effect of sea ice is typically neglected or parameterized solely in terms of ice concentration. In this study, an efficient way of adding ice effects to the global total water level prediction systems, via the ice–ocean stress, is described and eva...

Full description

Bibliographic Details
Main Authors: P. Wang, N. B. Bernier
Format: Article
Language:English
Published: Copernicus Publications 2023-06-01
Series:Geoscientific Model Development
Online Access:https://gmd.copernicus.org/articles/16/3335/2023/gmd-16-3335-2023.pdf
Description
Summary:<p>In operational flood forecast systems, the effect of sea ice is typically neglected or parameterized solely in terms of ice concentration. In this study, an efficient way of adding ice effects to the global total water level prediction systems, via the ice–ocean stress, is described and evaluated. The approach features a novel, consistent representation of the tidal relative ice–ocean velocities, based on a transfer function derived from ice and ocean tidal ellipses given by an external ice–ocean model. The approach and its impact are demonstrated over four ice seasons in the Northern Hemisphere, using in situ observations and model predictions. We show that adding ice effects helps the model reproduce most of the observed seasonal modulations in tides (up to 40 % in amplitude and 50<span class="inline-formula"><sup>∘</sup></span> in phase for <span class="inline-formula">M<sub>2</sub></span>) in the Arctic and Hudson Bay. The dominant driving mechanism for the seasonal modulations is shown to be the under-ice friction, acting in areas of shallow water (less than 100 <span class="inline-formula">m</span>) and its accompanied large shifts in the amphidromes (up to 125 <span class="inline-formula">km</span>). Important contributions from baroclinicity and tide–surge interaction due to ice–ocean stress are also found in the Arctic. Both mechanisms generally reinforce the seasonal modulations induced by the under-ice friction. In forecast systems that neglect or rely on simple ice concentration parameterizations, storm surges tend to be overestimated. With the inclusion of ice–ocean stress, surfaces stresses are significantly reduced (up to 100 % in landfast ice areas). Over the four ice seasons covered by this study, corrections up to 1.0 <span class="inline-formula">m</span> to the overestimation of surges are achieved. Remaining limitations regarding the overestimated amphidrome shifts and insufficient ice break-up during large storms are discussed. Finally, the anticipated trend of increasing risk of coastal flooding in the Arctic, associated with decreasing ice and its profound impact on tides and storm surges, is briefly discussed.</p>
ISSN:1991-959X
1991-9603