• English 
  • Spanish 

A new statistical seasonal forecasting method for the winter climate in Europe put to the test

2012 Nov 5 - 00:00

Researchers from the Santander Meteorology Group (CSIC-UC) and from the Atmospheric and Environmental Research (AER), Lexington, USA, have developed a new statistical method for seasonal prediction, which profits from the snow cover advance over Eurasia during October to predict Winter precipitation in Europe one month in advance. The last data available (from the past October) shows a strong signal indicating that the next Winter will be very rainy on the Iberian Peninsula.

Seasonal forecasts of the winter (DJF) climate in Europe are known to have limited skill. In the light of this paradigm, October Eurasian snow cover increase was recently found to be significantly related to DJF-precipitation totals on the Iberian Peninsula and southern Norway, variables known to be driven by the North Atlantic Oscillation (Hurrell 1995). One possible dynamical pathway for this formerly unknown lagged teleconnection is that an above normal increase of October Eurasian snow-cover in October favours a negative Arctic/North Atlantic Oscillation response during the following winter (Cohen & Jones 2011), which in turn enhances/reduces precipitation on the Iberian Peninsula and southern Norway respectively.

Based on this empirical relationship, a one month-ahead statistical forecasting method has been developed which uses October Eurasian snow cover increase as the only predictor variable. For the past 15 DJF-seasons, this one-month-ahead prediction approach has shown to have significant forecast skill (obtained from cross-validation) in the above mentioned regions.

Having this in mind, it is important to note that Eurasian snow cover during this year’s October has increased far more rapidly than it is normally the case (the corresponding index value exceeds two-standard deviations). Hence, for the forthcoming DJF-season 2012/13, the above mentioned statistical forecasting method indicates above normal precipitation totals for the Iberian Peninsula and below normal ones for southern Norway. The detailed forecast values are visualized in the figure shown below. Given the large anomaly for the predictor variable, the following winter will be a good test for assessing if the approach is suitable for real-time seasonal forecasts. It is encouraged to use this lagged teleconnection as an additional source of information for making seasonal forecasts of the winter season in Europe.

Fig. 1) Forecasted precipitation totals for DJF 2012/13, shown as relative anomaly from the 1997/98-2011/12 mean, e.g. a value of 100% means that DJF-precipitation is twice the normal. Dark points indicate regions where the method has significant forecast skill for the last 15 DJF-seasons, obtained by one-year-out cross-validation. The forecast was generated on 01/11/2012.

Brands S, Manzanas R, Gutiérrez JM, Cohen J (2012) Seasonal Predictability of Wintertime Precipitation in Europe Using the Snow Advance Index. Journal of Climate 25:4023-4028

Cohen J, Jones J (2011) A new index for more accurate winter predictions. Geophysical Research Letters 38:L21701. doi:10.1029/2011GL049626

Hurrell JW (1995) Decadal Trends in the North Atlantic Oscillation: Regional Temperatures and Precipitation. Science 269(5224), 676-679