Forecasting sea ice trajectories in the Arctic

While the drifting of sea ice in the Arctic has been subject of many studies, this new publication by Sukun Cheng, Laurent Bertino (both NERSC), Pierre Rampal (Université Grenoble Alpes and NERSC) and others, is adding valuable findings to forecasting drift trajectories.

 

The article “Probabilistic Forecasts of Sea Ice Trajectories in the Arctic: Impact of Uncertainties in Surface Wind and Ice Cohesion” was published in December 2020 in Oceans, you can find it here.

Cheng, Bertino, Rampal, and their co-authors set out to study how our sea ice model neXtSIM responds to uncertainty of two important factors: wind at the sea surface and sea ice cohesion. These two factors affect how accurate forecasts of sea ice trajectories can be.

Why is this relevant? Well…

 

If you are lost on sea ice, around where should rescuers look for you?

Can you imagine yourself drifting on sea ice in the middle of the Arctic? Fridtjof Nansen was the first person ever confronted with that situation, though voluntarily during the Fram expedition. He observed that the sea ice was generally drifting at a fraction of the wind speed (2%) and at a 10 degrees angle to the right of the wind direction. He had no operational support to utilize that information in case of trouble, though. 

More than 100 years later, the authorities of Arctic nations would deploy airplanes for search and rescue, but these may take up to 5 days to reach you wherever you have drifted. And the same old science would still be used today to track your position: computing the ice drift as a fraction of the surface winds.

 

Ensemble simulations of search area averaged over the Arctic Oceans.: Solid lines indicate the mean and error bars indicate the spatial standard deviation. Inset: the spatial-temporal averages over lead time. 13 consecutive ensemble runs simulate how wind, ice cohesion, and both effects jointly affect the search area size. Wind has the largest effect, but the strong winds in the period 13th -22nd March enhance the effect of cohesion. Source: Cheng et al. 2020Ensemble simulations of search area averaged over the Arctic Oceans.: Solid lines indicate the mean and error bars indicate the spatial standard deviation. Inset: the spatial-temporal averages over lead time. 13 consecutive ensemble runs simulate how wind, ice cohesion, and both effects jointly affect the search area size. Wind has the largest effect, but the strong winds in the period 13th -22nd March enhance the effect of cohesion. Source: Cheng et al. 2020

But where can you possibly be after 5 days?

How far and where should rescuers search for you? To address this problem – forecasting sea ice drift in the Arctic and its uncertainties – Cheng and his co-authors conducted the recently published study. Their findings indicate that both wind and sea ice cohesion can have significant effects on the quality of a forecast for the sea ice drift in the Arctic, especially during winter. They compare how much the uncertainties both in surface winds and in the intrinsic sea ice mechanics contribute to the uncertainties in the end position of the drift. They confirm that the wind is usually the main source of uncertainty but that accounting for uncertainties in the sea ice cohesion parameter does help in cases of strong wind events and makes the search areas more asymmetric. Both sources of uncertainties should thus be used jointly, according to their study.

 

The three main points the study highlights are the following:

  • Perturbations of the surface wind forcing and sea ice cohesion have a significant effect on the ensemble forecast of sea ice drift in the Arctic during winter.
  • Wind perturbations make out most of the ensemble spread in ensemble forecasts.
  • The inhomogeneities of ice cohesion significantly increase the anisotropy of ensemble sea ice drift.

Considering these points, forecasting sea ice trajectories will be improved. The authors however cannot guarantee your survival if you step too close to the footsteps of Fridtjof Nansen. Drift responsibly.

  

Reference:

Cheng, S., Aydoğdu, A., Rampal, P., Carrassi, A., and L. Bertino. Probabilistic Forecasts of Sea Ice Trajectories in the Arctic: Impact of Uncertainties in Surface Wind and Ice Cohesion. Oceans 2020. 1. 326-342. https://doi.org/10.3390/oceans1040022  

Add comment

Login to post comments