Quantifying local versus remote sources of polar predictability
Abstract
The predictability of the polar ice-ocean-atmosphere environments has been the subject of growing research over the last decade, driven by the rapid changes taking place and increased stakeholder interest in the polar regions. At the sub-seasonal to seasonal (S2S) timescale, there is indication from both models and observations that both local (such as sea-ice thickness or area) and remote predictors (such as tropical SSTs) can offer predictive information for the polar ice-ocean-atmosphere system. Quantifying the partition and pathways of local and remote contributions to polar predictability has however received less attention. In this work, we quantify the contribution of predictive information from the tropics and mid-latitudes to the poles at S2S timescales, essentially posing the question 'How much does a polar forecast improve given a perfect forecast of tropical/mid-latitude conditions?'. To do this, we employ a perfect-model GCM framework in which we compare the predictability characteristics of a forecast ensemble in which the atmosphere over various tropical and mid-latitude domains is nudged to a parent control run to the predictability characteristics of a free-running forecast ensemble. We find that the tropics only offer modest S2S predictability to either pole, while the mid-latitudes offer significant S2S predictability to Antarctica, but less so to the Arctic, revealing a polar asymmetry in the partition of local and remote contributions to polar predictability.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFMGC51O0976B
- Keywords:
-
- 3305 Climate change and variability;
- ATMOSPHERIC PROCESSESDE: 0798 Modeling;
- CRYOSPHEREDE: 1878 Water/energy interactions;
- HYDROLOGYDE: 4207 Arctic and Antarctic oceanography;
- OCEANOGRAPHY: GENERAL