Tag Archives: metar statistics

We’re excited with Flight condition Statistics

By now, you should know we’ve been working on metar statistics (there are 3 other posts, or so)

The fact  is that lots of information can be gathered from METARs through time (temperature, wind, dew point and cloud base profiles…)
At one point, we were even asked to evaluate forecast reliability (through METAR-TAF difference)

But for now, we elected to keep it simple and monitor how much an aiport is in VFR conditions.

My initial intention was for the stranger pilot to be able to get an idea of how ‘bad’ an airport could be in a particular month, or time of day.
So we started saving flight condition statistics back on Dec the 1st, 2008. So far, we’ve stored over 30 Million flight condition reports !!!

Well, let me tell you that we’ve found database and server load challenges… But we’ve kept up.

And recenlty, we’ve started to analyze whether all of that storage made any sense. Let me show you…

Below we have a chart for LELL flight condition statistics from Dec’08 to Jun’09:

LELL flight condition statistics
LELL flight condition statistics

Here you can easily see what a nice airport LELL (Sabadell in Spain) is to fly to in VFR conditions:
The percentage represents the number of VFR, MVFR, IFR or LIFR METAR reports (according to this flight conditions definition)
On the horizontal axis you have each month (with time of day inside). So we can see that back in February, the morning and night METAR reports had less VFR conditions than those around noon. You can also see that the closer you get to the summer months, the more VFR conditions existing at LELL.

Let’s see what this looks like at the famous low visibility LIMC (Milano Malpensa in Italy):

Flight condition statistics at LIMC
Flight condition statistics at LIMC

What a difference!

Here, we can see that back in the December-January period on average only about 40% of the METARs where reporting VFR conditions.
Here too, the closer we get to summer time, the better it gets.

What we see seems now worth the time and effort we put into gathering statistics.

We’re not quite ready to share this sort of reports with the general public, but have started working on it.

As usual, please let us know how you like it.

Promising future for METAR statistics

As I write this post 8,611,198 METAR reports have been analyzed since we started storing metar statistics back on Dec 21st 2008.
As you can imagine, computing metar statistics on real time for each METAR that we serve is loading our server a bit.

We are working on ways of doing that without such a huge server load.

In the mean time, we are taking the middle path. Meteo·Mobile will show a link to the statistics right below the METAR report itself for those of you who enjoy knowing how many times the METAR has shown VFR, MVFR, IFR or LIFR conditions.

Since only a fraction of the pilots will probably ask for the statistics, this should take care of the server load thus improving the response time (that had been slowed down a bit) while we work on the faster statistics calculation.

I really hope you are enjoying the statistics.
In the near future, I hope to be able to expand its capabilities by being able to show what percentage of the time a particular field is enjoying VFR conditions on the month of July, for instance.

As usual, I will ask you not to hesitate to comment on this.