Subseasonal Prediction of Land Cold Extremes in Boreal Wintertime.

Autor: Xiang, Baoqiang, Sun, Y. Qiang, Chen, Jan‐Huey, Johnson, Nathaniel C., Jiang, Xianan
Předmět:
Zdroj: Journal of Geophysical Research. Atmospheres; 7/16/2020, Vol. 125 Issue 13, p1-18, 18p
Abstrakt: Subseasonal climate prediction has emerged as a top forecast priority but remains a great challenge. Subseasonal extreme prediction is even more difficult than predicting the time‐mean variability. Here we show that the wintertime cold extremes, measured by the frequency of extreme cold days (ECDs), are skillfully predicted by the European Centre for Medium‐Range Weather Forecasts (ECMWF) model 2–4 weeks in advance over a large fraction of the Northern Hemisphere land region. The physical basis for such skill in predicting ECDs is primarily rooted in predicting a small subset of leading empirical orthogonal function (EOF) modes of ECDs identified from observations, including two modes in Eurasia (North Atlantic Oscillation and Eurasia Meridional Dipole mode) and three modes in North America (North Pacific Oscillation, Pacific‐North America teleconnection mode, and the North America Zonal Dipole mode). It is of interest to note that these two modes in Eurasia are more predictable than the three leading modes in North America mainly due to their longer persistence. The source of predictability for the leading EOF modes mainly originates from atmospheric internal modes and the land‐atmosphere coupling. All these modes are strongly coupled to dynamically coherent planetary‐scale atmospheric circulations, which not only amplify but also prolong the surface air temperature anomaly, serving as a source of predictability at subseasonal timescales. The Eurasian Meridional Dipole mode is also tied to the lower‐boundary snow anomaly, and the snow‐atmosphere coupling helps sustain this mode and provides a source of predictability. Key Points: The wintertime land extreme cold days can be largely predicted by the ECMWF model 2–4 weeks in advanceThe physical basis in predicting ECDs is mainly rooted in predicting a small subset of leading EOF modes of ECDs identified from observationThe predictability source for the leading EOF modes originates from atmospheric internal modes and the land‐atmosphere coupling [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index