Name
Frederik Schenk
Department of Mechanics, Royal Institute of Technology (KTH), Stockholm, Sweden
Institute of Coastal Research, Helmholtz-Zentrum Geesthacht (HZG), Germany

Title
Analog-Reconstruction of Northern European Storminess since 1850

Time and place
Tue 1 Oct 2013, 11.15 2013
Room C609, Arrhenius Laboratory, 6th floor

(This event has teken place.)

Abstract
Storminess over the Euro-Atlantic regions shows mostly positive trends since the 1960s accompanied by a NE-shift of the storm tracks towards Scandinavia. This pattern appears to be consistent with simulations under increasing greenhouse gas emission scenarios for the end of the 21st century. There is, however, no scientific consensus on the existence of long-term trends in storm statistics since the 1880s or earlier.

In the absence of reliable wind observations, long-term storminess is usually derived from indices based on more homogeneous pressure data or quasi-homogeneous reanalysis wind fields. As shown by previous analysis (Krueger et al., 2013), low-frequency variations and trends of geostrophic wind indices are not consistent with the output of the 20th Century Reanalysis (20CR) since 1871. This is most likely due to spurious trends caused by a varying number of assimilated observations in 20CR i.e. in data sparse regions.

In this talk, the analog-method is introduced as an alternative strategy to derive physically consistent meteorological fields from sparse observation data (Schenk and Zorita, 2012). Storminess derived from the reconstructed daily wind fields since 1850 are compared with geostrophic wind indices, a storm flood index and the 20CR output. While the analysis of deep cyclones (< 980 hPa) suggests an unprecedented high spatial impact of extreme cyclones in recent decades, annual wind statistics since 1850 point towards a rather stationary wind climate in agreement with global model simulations of the last millennium. The results of the analog-reconstruction are consistent with geostrophic wind indices confirming also that 20CR cannot be used for the evaluation of long-term trends since 1871 over data sparse regions.

Welcome!