Different Long-term Trends of Extra-tropical Cyclones and Windstorms in ERA-20C and NOAA-20CR Reanalyses

Abstract:

Two reanalysis datasets covering the past 100 years, namely ERA-20C and NOAA-20CR (v2) are used to analyse the temporal variability of extra-tropical cyclones and wind storms. Analyses are carried out for both hemispheres and several smaller sub-regions. These datasets allow assessing the long and short-term variability of these events beyond the past 60 years covered by common reanalysis products. Due to the damage-related impact of strong cyclones and wind storms, knowledge about their variability is of high importance for societies and economies. Cyclones and wind storms are identified using two objective algorithms based on six hourly mean sea level pressure fields and near-surface wind speeds, respectively.

Cyclone and windstorm time series are low-pass filtered with a cut-off frequency of 1/31 years. High-frequency time series are obtained by calculating the residuum between the original and the low-pass filtered time series. Additionally, linear trends are calculated to analyse long-term trends for three periods: 1901-1930, 1931-1960 and 1961-1999.

The results show substantial differences regarding the long-term trends of cyclones and wind storms between both datasets, especially during the early 20^th century. In general, better agreement is found for the end of the century and to some extend also for the most severe cyclones. In contrast, higher-frequency variability is in general in good agreement between both datasets regardless of different long-term trends and also has a higher agreement at the end of the century.

The outcome of this study indicates that based on these two datasets no reliable conclusion about long-term trends of extra-tropical cyclones and wind storm events before the mid of the 20^th century can be drawn. However, analyses focusing on higher-frequency variability of these events and its drivers may be useful, if taking into account long-term trends.