2006-09-10

Event - Storm Forecast by Windguru.cz

The central issue about the Red Bull Storm Chase is to pick out the biggest storm of the year. Sounds more simple than it is! We asked our partner Vaclav Hornik from windguru.cz how calculable a storm forecast is:

How many days in advance can a storm be predicted?

It depends how a "storm" is defined. For severe thunderstorms in European
conditions, forecasts are issued about 1-2 days ahead (see e.g.
http://www.estofex.org/), but a precise forecast, including the location of a
possible storm, can be issued 6 hours ahead. Other situation is, if a stormy
weather with strong winds occurs in connection with a non-convective weather,
e.g. deep low, this is what storm chase riders are interested in.  Such
conditions are better predictable and 2-3 days ahead we
may have a good forecast.    

How valid are forecasts (5 days before, 3 days before, 1 day before)?

5 days ahead forecast is basically a medium range forecast which has
probabilistic character and the forecast is to be taken for a most likely
development. But other development is well possible. Some centres issue called "ensemble forecasts", giving several possible scenarios. The most
prominent organization for medium range forecast in Europe is the ECMWF
(http://www.ecmwf.int/).
3 days ahead is the limit for short range forecast issued by mesoscale
numerical models and national weather services. The synoptic situation is
usually well forecast but there may be relatively large quantitative
differences in temperature, precipitation and wind strength and direction
between forecast and reality.
1 day ahead is a relatively reliable forecast in the regional scale,
especially if there are large dominant synoptic issues as lows.  

So if you see 10 Bft forecast for 5th day, do not panic, anything can change
later.., 3 or 2 days and the storm is still there? Watch the situation! 10
Bft forecast for tomorrow? Go for it!

How valid are parameters like strength and direction?

Again, it depends on the convective (thunderstorm)/non convective origin of the
storm and the forecast horizon. In addition to what was said above, the
forecast error is higher in convective events which are smaller scale.
Conditions on specific spot may differ from forecast because real wind can be
affected by local topography, thermal winds etc..

Why are there such big differences between forecasts of different
providers?

Some providers just process or visualize one of the few global forecasts,
mostly NCEP GFS, which has a resolution of 0.5 degree (about 50km in mid
latitudes).  Other use one of many mesoscale numerical models to improve
resolution and precision of forecast.  Each big weather service runs his own
model and there are community models like MM5 or WRF. Windguru also runs his
own MM5 model. The mesoscale models can use boundary conditions from
different global models, they are configured different way etc. For the
horizon of one day, however, there can be a remarkable consensus between
different models. It is informative to look at different models. If there is
no consensus, it points at less predictable weather conditions with different
scenarios possible.

What is the difference between windguru and windguru PRO?

The most important difference is that PRO version offers the latest high
resolution MM5 model forecasts for Europe. The free version only provides
forecasts based on global GFS model, you can also see MM5 forecasts but these
are only a "preview" - they are 12 hours (2 model cycles) delayed. As we
mentioned earlier the shorter the forecast horizon the more reliable forecast
we have, that's why access to the latest data is important. Subscription fee
for the PRO version is very reasonable, 490 CZK per year (about 18 EUR). This
is a sum average windsurfer pays for fuel for one day trip to his windsurf
spot so I think it's a real good investment ;)
PRO subscribers can also turn off all banner advertising for more comfort.

What are the advantages of windguru?

Windguru is not an ordinary weather website, it is specialised for windsurfers,
kiters and people interested in other wind related sports. We try to offer
windsurfer optimized, easy to read forecasts with maximum customisability.
We have spot forecasts for thousands of spots, if this is not enough users can
setup their own custom spots for any place on Earth, we have an archive based
on past model forecasts, MM5 forecast maps, WAP and WAP2 forecasts for your
mobile phones, some tide info and other features.