On the future reduction of snowfall in western and central Europe

Abstract

Large parts of western and central Europe face a 20–50 % future reduction in snowfall on Hellmann days (days with daily-mean temperatures below freezing). This strong reduction occurs in addition to the expected 75 % decrease of the number of Hellmann days near the end of the twenty first century. The result is insensitive to the exact freezing-level threshold, but is in sharp contrast with the winter daily precipitation, which increases under most global warming scenarios. Not only climate model simulations show this. Observational records also reveal that probabilities for precipitation on Hellmann days have been larger in the past. The future reduction is a consequence of the freezing-level threshold becoming a more extreme quantile of the temperature distribution in the future. Only certain circulation types permit these quantiles to be reached, and it is shown that these have intrinsically low precipitation probability.

This is a preview of subscription content, log in to check access.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11

References

  1. Cattiaux J, Douville H, Ribes A, Chauvin F, Plante C (2012) Towards a better understanding of wintertime cold extremes over Europe: a pilot study with CNRM and IPSL atmospheric models. Clim Dyn. doi:10.1007/s00382-012-1436-7

  2. Cheng X, Wallace JM (1993) Cluster analysis of the Northern hemisphere wintertime 500-hPa height field: spatial patterns. J Atmos Sci 50:2674–2696

    Article  Google Scholar 

  3. Croci-Maspoli M, Schwierz C, Davies HC (2007) A muli-faceted climatology of atmospheric blocking and its recent linear trend. J Clim 20:633–649

    Article  Google Scholar 

  4. Frei C, Schöll R, Fukutome S, Schmidli J, Vidale P (2006) Future change of precipitation extremes in Europe: intercomparison of scenarios from regional climate models. J Geophys Res 111:D06105. doi:10.1029/2005JD005965

  5. Haren van R, van Oldenborgh GJ, Lenderink G, Collins M, Hazeleger W (2012) SST and circulation trend biases cause an underestimation of European precipitation trends. Clim Dyn 1–20. doi:10.1007/s00382-012-1401-5

  6. Haylock M, Hofstra N, Klein Tank A, Klok E, Jones P, New M (2008) A european daily high-resolution gridded dataset of surface temperature and precipitation. J Geophys Res 113:D20119. doi:10.1029/2008JD10201

    Article  Google Scholar 

  7. Joshi MM, Gregory JM, Webb MJ, Sexton DMH, Johns TC (2008) Mechanisms for the land/sea warming contrast exhibited by simulations of climate change. J Clim 30:455–465

    Google Scholar 

  8. Kew S, Selten F, Lenderink G, Hazeleger W (2011) Robust assessment of future changes in extreme precipitation over the rhine basin using a gcm. Hydrol Earth Syst Sci 15:1157–1166. doi:10.5194/hess-15-1157-2011

    Article  Google Scholar 

  9. Klein Tank AMG, Können GP, Selten FM (2005) Signals of anthropogenic influence on european warming as seen in the trend patterns of daily temperature variance. Int J Climatol 25:1–16

    Article  Google Scholar 

  10. Lenderink G, van Meijgaard E (2008) Increase in hourly precipitation extremes beyond expectations from temperature. Nat Geosci 1:511–514. doi:10.1038/ngeo262

    Article  Google Scholar 

  11. Meehl GA, Covey C, Taylor KE, Delworth T, Stouffer RJ, Latif M, McAvaney B, Mitchell JFB (2007) The WCRP CMIP3 multimodel dataset: a new era in climate change research. Bull Am Meteorol Soc 88:1383–1394. doi:10.1175/BAMS-88-9-1383

    Article  Google Scholar 

  12. O’Gorman PA, Schneider T (2009) The physical basis for increases in precipitation extremes in simulations of twenty first century climate change. Proc Natl Acad Sci USA 106:14773–14777. doi:10.1073/pnas.0907610106

    Article  Google Scholar 

  13. Pall P, Allen MR, Stone DA (2006) Testing the ClausiusClapeyron constraint on changes in extreme precipitation under co2 warming. Clim Dyn 28:351–363. doi:10.1007/s00382-006-0180-2

    Article  Google Scholar 

  14. Rex DF (1951) The effect of Atlantic blocking action upon European climate. Tellus 3:1–16. doi:10.1111/j.2153-3490.1951.tb00784.x

    Article  Google Scholar 

  15. Sillmann J, Croci-Maspoli M (2009) Present and future atmospheric blocking and its impact on European mean and extreme climate. Geophys Res Lett 36:L10702. doi:10.1029/2009GL038259.x

    Article  Google Scholar 

  16. Sterl A, Severijns C, Dijkstra H et al. (2008) When can we expect extremely high surface temperatures? Geophys Res Lett 35:L14703. doi:10.1029/2008GL034071

    Article  Google Scholar 

  17. Trigo R, Trigo I, DaCamara C, Osborn TJ (2004) Climate impact of the European winter blocking episodes from the NCEP/NCAR Reanalysis. Clim Dyn 23:17–28. doi:10.1007/s00382-004-0410-4

    Article  Google Scholar 

  18. de Vries H, Haarsma RJ, Hazeleger W (2012) Western European cold spells in current and future climate. Geophys Res Lett 39:L04706. doi:10.1029/2011GL050665

    Article  Google Scholar 

  19. Ward J (1963) Hierarchical grouping to optimize an objective function. J Am Stat Assoc 58:236–244

    Article  Google Scholar 

Download references

Acknowledgments

The paper has benefited from comments by the reviewers. Furthermore, we acknowledge the E-OBS dataset from the EU-FP6 project ENSEMBLES (http://ensembles-eu.metoffice.com) and the data providers in the ECA&D project (http://eca.knmi.nl). The research undertaken in this study has been made possible by GasTerra and NAM.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Hylke de Vries.

Appendix: Projecting onto target patterns

Appendix: Projecting onto target patterns

Here we elaborate on the method how to project the large-scale flow on to target patterns, and how future changes can be attributed to several change mechanisms. The approach is based on Cattiaux et al. (2012).

Basic steps

First, the target patterns (TP) have to be determined. In this study we use Ward clustering of a reference set of Z500-anomaly maps. The TP-structure is denoted as e k . Next, the objective fields a i (the Z500-anomaly fields) are assigned to the TP with which they have the highest correlation cor(e k a i ). A map is not assigned if cor(e k ,a i ) <0.5 for all e k . After the assigning process, the mean of all a i can be written as

$$ \bar{A} = \frac{1}{n}\sum_{i=1}^n a_i=\sum_k f_k d_k \equiv \sum_k\lfloor f d\rfloor_k $$
(3)

where \(\lfloor\rfloor_k\) is shorthand notation, n is the number of fields in the dataset, f k  = n k /n is the relative frequency of the fields assigned to TP-k, and

$$ d_k = \frac{1}{n_k}\sum_{i=1}^{n_k} a_{i} $$
(4)

is called the response pattern (RP), the centroid of all a i assigned to TP-k. Residual maps, that could not be assigned to any TP, are included in (3) and form the “garbage-bin”, RP-X.

Analyzing changes

In the main text, we use 30-winter sliding windows of data and project each of these on to the TPs. As a result, both the frequencies f k and the RPs become functions of the period, which is indicated by t. The first 30 winters are used as a reference period t ref . \(\bar{A}(t)\) can be written as:

$$ \bar{A}(t)=\sum_k f_k(t) d_k(t)\equiv \bar{A}(t_{ref}) + \sum_k\lfloor f(t)\delta(t)\rfloor_k $$
(5)

which defines \(\delta_k(t)=d_k(t)-\bar{A}(t_{ref})\) as the anomaly with respect to the mean field of the reference period. The second step follows by noting that ∑ k f k (t) = 1 for all periods (by definition). Note that according to this definition ∑ k f k (t ref )δ(t ref ) = 0. To simplify notation we now focus on two periods which will be labeled with superscripts PF (“present” and “future”) e.g., \(\bar{A}^{F,P}=\bar{A}(t^{F,P}), \delta_k^P=\delta_k(t^P), f_k^F=f_k(t^F). \) Then we can write for the difference of mean \(\Updelta\bar{A}=\bar{A}(t^F)-\bar{A}(t^P)\)

$$ \Updelta\bar{A} =\sum_k\left\lfloor(f+f')(\delta+\delta')-f\delta\right\rfloor_k=\sum_k\left\lfloor f'\delta+f\delta' + f'\delta'\right\rfloor_k $$
(6)

where we suppressed all labels by using the notation f = f P and f′ = f F − f P (similar notation for δ).

Other fields: transfer functions

The next step is to consider the change of a variable x, related to the circulation a by means of a transfer function \(\Upphi, \) such that \(x=\Upphi[a].\) This transfer function selects for example the local temperature/precipitation in the Netherlands, associated with a specific large-scale circulation a. Using the same approach as before we get [cf., (5)]:

$$ \bar{X}(t) = \sum_k f_k(t)x_k(t) \equiv \bar{X}(t_{ref}) + \sum_k\lfloor f(t)\xi(t)\rfloor_k, $$
(7)

with \(x_k(t)=\Upphi[d_k(t),t]\) and \(\xi_k(t)=x_k(t)-\bar{X}(t_{ref})\). Note that \(\Upphi\) generally depends both explicitly and implicitly on time. The change of mean \(\Updelta \bar{X}=\bar{X}^F-\bar{X}^P\) can be written as [cf., (6)]

$$ \Updelta \bar{X} =\sum_k\left\lfloor f'\xi+f\xi' + f'\xi'\right\rfloor_k = \hbox{BeC-CC}+\hbox{WiC}+\hbox{N.L.}, $$
(8)

where we have suppressed labels by writing ξ = ξP and ξ′ = ξF − ξP (similar notation for f).

Term BeC-CC denotes the Between-Cluster Circulation Change, the contribution to \(\Updelta X\) from cluster frequency changes; term WiC denotes the Within-Cluster Changes; N.L is the nonlinear term. Cattiaux et al. (2012) suggest to decompose term WiC into a Within-Cluster Circulation Change (WiC-CC) and a Within-Cluster Response Change (WiC-RC):

$$ \begin{aligned} \hbox{WiC-CC}'=&\sum_k\left\lfloor f^P\big(\Upphi^P[d^F]-\Upphi^P[d^P]\big)\right\rfloor_k,\\ \hbox{WiC-RC}'=&\sum_k\left\lfloor f^P\big(\Upphi^F[d^F]-\Upphi^P[d^F]\big)\right\rfloor_k, \end{aligned} $$

where \(\Upphi^i[d^j]=\Upphi[d(t_j),t_i]\) with ij = (PF). However, it is easily shown that an equally valid decomposition is:

$$ \begin{aligned} \hbox{WiC-CC}''&=\sum_k\left\lfloor f^P\big(\Upphi^F[d^F]-\Upphi^F[d^P]\big)\right\rfloor_k,\cr \hbox{WiC-RC}''&=\sum_k\left\lfloor f^P\big(\Upphi^F[d^P]-\Upphi^P[d^P]\big)\right\rfloor_k. \cr \end{aligned} $$

The simplest way to take the ambiguity of the decomposition into account is by writing

$${\text{WiC-CC}} = (\varepsilon ){\text{WiC-CC}}^{\prime } + (1 - \varepsilon ){\text{WiC-CC}}^{{\prime \prime }} , $$
(9)

where \(\varepsilon \in (0,1), \) and similar for WiC-RC. Two terms require further explanation: \(\Upphi^P[d^F]\) is the present-day response x to future circulations. As in Cattiaux et al. (2012) this term is estimated by searching for the present-day analogues of the future circulations and taking the accompanying present-day x values. The future response to present-day circulations \((\Upphi^F[d^P])\) is evaluated in the same way.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

de Vries, H., Haarsma, R.J. & Hazeleger, W. On the future reduction of snowfall in western and central Europe. Clim Dyn 41, 2319–2330 (2013). https://doi.org/10.1007/s00382-012-1583-x

Download citation

Keywords

  • Snowfall
  • Climate change
  • Circulation patterns
  • Attribution