Wednesday, September 12, 2012

Why High-Resolution Forecasts?

One thing that sets the SARWeather forecasts apart from other sources of weather information available in various media, and usually for free, is that we generate much higher-resolution forecasts.

But what does this means for me? I don't care about your techno-babble!


Here is a forecast I ran for my native Iceland, where SARWeather is based. Ignoring the horrendous storms around Vatnajökull - we're just not going there this week - there was a bit of a windy patch around Langjökull on Wednesday night.

A storm around the Langjökull ice-cap in Iceland at 9km resolution.

This forecast is run at a resolution of 9km per grid cell, which is about as good as you can expect from traditional sources, and allows us to fit the entire island in a single forecast. We can see the general outline of the worst hit area, but somehow the weather seems to ignore most of the terrain. That can't be right, can it? Let's run a higher-resolution forecast for this area and have another look.

The same storm forecast at 1km resolution.

The high-resolution forecast showing the same storm reveals a lot more structure of high and low wind areas. If you are faced with a decision of where to descend the glacier, or where to set up camp to weather the storm, the second forecast is clearly a much better tool. And either one is most likely better than what you otherwise have access to. How many forecasts can you buy for the price of one hiking tent?

There is an obvious trade-off between resolution, size, and period on one hand, and price and timeliness on the other. This is why we run the large 6-day forecast at our lowest resolution, and then zoom in on an interesting area at a higher resolution. But if you already know that you are going on Mt. Rainier, either for fun or to bring down someone who got into trouble, then you'd got directly for the high-resolution forecast.

1 comment: