Real-time forecasts for your holidays in Switzerland. Keep tuned and give feedback!

Real-time forecasts for your holidays in Switzerland. Keep tuned and give feedback!

In Switzerland the 1st of June represents the unofficial start of the flash-flood season. Against this, in 2013 we selected June 1st to launch our platform on early recognition of critical droughts (see this post here). At that moment, a flood situation was going on in large parts of Switzerland (see our archived forecast here), and so less attention was given to that new product of us.

This year I was planning to write sometimes here a blog post about an ongoing flood event. I discovered now that in the coming days a small heatwave is meant to hit Switzerland and so I changed my mind and decided to present you a very experimental product that we are evaluating before launching it for real-time assessment of disposition to hydrological extremes in Switzerland. This product gives a glance on both hydrological extremes, so there is no risk that I focus on the wrong one.

In the following, I will present you the products and then give you a link were you will be able to see each day at about 11:00 (central European summer time) an updated forecast for the next 5 days. I anticipate that this product is so far rather simple and the consideration of uncertainties is still rudimentary.

How it works:

  • We run a real-time 5-days forecast at 500×500 m resolution for almost entire Switzerland and propagate an ensemble of 16 COSMO-LEPS members to obtain forecasts for the next 5 days as presented in Zappa et al. (2014).
  • We subdivided Switzerland into 57 sub-areas according to tributaries having an areas of about 1000 km2 each.
  • For each area and each variable in our hydrological model, we have a 32-year seasonal climatology (a 30-days centered cumulative distribution function CDF for each day of the year) for evaluating the current anomalies (Fig. 1).
  • Having this 16-member information for each sub-area of Switzerland and for each variable of the model, we create a composite map with the goal of identifying areas that might be prone to develop drought or be affected by flood situations in the next few days (Fig. 2).
Figure 1: Operational forecasting of total natural runoff the Thur basin (Fundel and Zappa, 2011). The real-time five-days probabilistic forecasts forced by COSMO-LEPS are embedded into the long-term seasonal climatology (0.5% to 99.5% percentiles of the CDF) for easy assessment of water shortages or flood disposition. The visualization also includes information on the recent past (CTRL-Run)
Figure 2: Operational assessment of drought (left) and flood disposition (right) in Switzerland (here for a lead time of 3 days in case of a forecasts started on June 4th 2015) based on total natural runoff (RGS). Colouring is assigned according to the number of COSMO-LEPS members exceeding or not-exceeding a certain percentile of the climatology

So far the assignment of the level of flood or drought disposition is guided by the median of our 5 days COSMO-LEPS and the respective CDF of each sub-area. The three levels concerning the assessment of drought are related to the 2.5%, 10% and 25% percentiles, while the assessment of flood disposition relates to the 75%, 90% and 97.5% percentiles of the CDF.

In the figure above, there are some regions where at least 50% of the COSMO-LEPS show natural discharge below the 25% percentile, while other regions are currently assessed to have a low disposition to flood because the natural runoff exceeds in over 50% of the members the as 75% percentile of the climatology.

This split behaviour is rather typical for situations linked to (small) heatwaves. Regions without snow and glaciers suffer from water deficit and enhanced evapotranspiration, while the alpine areas have still snow cover and being to some degree glaciated experience large melt rates from the cryosphere. We described this kind of situation in a paper published some years after the 2003 European Summer Heatwave (Zappa and Kan, 2007).

Figure 3: Same as Figure 2, but for the daily snowmelt amount (SWA)

Where are the forecasts? [Here!]

5-days forecast of the simulated natural discharge anomalies (RGS)

5-days forecast of the simulated snowmelt anomalies (SWA)

5-days forecast of the simulated soil water resources anomalies (SSM)

5-days forecast of the simulated groundwater storage anomalies (SLZ)

And now?

I decided to open this product for the HEPEX community for the duration of this summer (JJA). Maybe someone of you plans holidays in Switzerland for a 5 days trip. This might be useful to you. Maybe other among you have some ideas and feedback for that. It would be very welcome in order to finalize the products (after verification) and begin with the realization of similar products for the hydropower sector including also extended range forecasts (see my previous blog on the “Böögg Bang Theory”). My plan is to report on this at the workshop on seasonal forecasting at SMHI in September (abstract to be submitted soon!).

I will put a comment now and then on current situation to keep this post alive until the end of the summer. Thanks in advance for your feedback.

I wish all of you a summer with interesting situations to be forecasted and managed in your country and hopefully no severe events affecting infrastructure and population. For those among you travelling to Prague for IUGG, see you there at the HEPEX session.


  1. Zappa M, Bernhard L, Spirig C, Pfaundler M, Stahl K, Kruse S, Seidl I, Stähli M. 2014. A prototype platform for water resources monitoring and early recognition of critical droughts in Switzerland, Proc. IAHS, 364, 492-498, doi:10.5194/piahs-364-492-2014
  2. Zappa M, Kan C. 2007. Extreme heat and runoff extremes in the Swiss Alps. Natural Hazards and Earth System Sciences, 7:375-389. [Get it!]


The forecasts linked to this blogpost are EXPERIMENTAL. They are provided to get feedback from a scientific community until the end of August 2015. They are NOT meant to provide official support for decisions concerning any event in any area of Switzerland during this time. If you are anyway willing to take a decision based on the information presented in the links, then you make this at your own risk.

2 thoughts on “Real-time forecasts for your holidays in Switzerland. Keep tuned and give feedback!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.