Dynamical Downscaling of the Arctic Climate with a Focus on Polar Cyclone Climatology

• Published in: Atmosphere-ocean Vol. 57; no. 1; pp. 41 - 60
• Main Authors: Zhang, Minghong, Perrie, William, Long, Zhenxia
• Format: Journal Article
• Language: English
• Published: Ottawa Taylor & Francis 01-01-2019; Taylor & Francis Ltd
• Subjects: Arctic climates; Arctic cyclone climatology; Arctic cyclones; Arctic radiation; Atmospheric models; Basins; Climate; Climatology; Computer simulation; Cyclones; Cyclonic activity; dynamical downscaling; Earth; Environmental modeling; Fields; Induction heating; Long wave radiation; Mathematical models; Moisture content; Polar environments; Polar WRF; Simulation; Solar radiation; Summer; surface radiation; Water; Water content; Water vapor; Water vapour; Weather; Weather forecasting; Arctic region
• ISSN: 0705-5900; 1480-9214
• DOI: 10.1080/07055900.2017.1369390

We present a dynamical downscaling of the Arctic climatology using a high-resolution implementation of the Polar Weather Research and Forecasting, version 3.6 (WRF3.6) model, with a focus on Arctic cyclone activity. The study period is 1979-2004 and the driving fields are data from the Hadley Centre Global Environmental Model, version 2, with an Earth System component (HadGEM2-ES) simulations. We show that the results from the Polar WRF model provide significantly improved simulations of the frequency, intensity, and size of cyclones compared with the HadGEM2-ES simulations. Polar WRF reproduces the intensity of winter cyclones found in ERA-Interim, the global atmospheric reanalysis produced by the European Centre for Medium-range Weather Forecasts (ECMWF), and suggests that the average minimum central pressure of the cyclones is about 10 hPa lower than that derived from HadGEM2-ES simulations. Although both models underestimate the frequency of summer Arctic cyclones, Polar WRF simulations suggest there are 10.5% more cyclones per month than do HadGEM2-ES results. Overall, the Polar WRF model captures more intense and smaller cyclones than are obtained in HadGEM2-ES results, in better agreement with the ERA-Interim reanalysis data. Our results also show that the improved simulations of Arctic synoptic weather systems contribute to better simulations of atmospheric surface fields. The Polar WRF model is better able to simulate both the spatial patterns and magnitudes of the ERA-Interim reanalysis data than HadGEM2-ES is; in particular, the latter overestimates the absorbed solar radiation in the Arctic basin by as much as 30 W m
−2
and underestimates longwave radiation by about 10 W m
−2
in summer. Our results suggest that the improved simulations of longwave and solar radiation are partly associated with a better simulation of cloud liquid water content in the Polar WRF model, which is linked to improvements in the simulation of cyclone frequency and intensity and the resulting transient eddy transports of heat and water vapour.