GFS based weather charts

I’ve made some pages of animated weather charts from GFS data. I download a lot of GFS data every day to produce the GFS data feed to WxSim and I wondered what other things I could do with the data without adding to the download volumes.

Firstly, two warnings…

  1. Unless you have a huge phone then the images won’t work for you. Seriously…don’t even bother trying! It might be OK on a tablet but this is really intended for desktop/laptop screens.
  2. Don’t use this if you have significant Internet bandwidth limitations. Each animated chart requires about 15MB or more to download. That’s per variable, so if you change from Temperature to Pressure that’s about 30MB. If you change from NW Europe to World, it’s going to take 15MB to download the first World chart.

After a lot of experimenting with scripts I’ve managed to create Weather Charts. There are currently 22 charts showing different weather variables or combinations of variables selectable by links down the left hand side. You can also select between 3 regions (NW Europe, World and N America) using the drop down menu above the variable links. In the current instance of the page, charts for all variables exist for every region, but that may change in future. If you hover your mouse over the links you’ll get a slightly longer description of the variable as a tooltip.

The controls at the top allow you to stop/start the animation and move one frame forwards or backwards (when stopped). You can also move the Frame# slider left and right to scroll through the animation more quickly in either direction. The UTC time for the current frame is displayed at the top of the page (to the left of the Frame# slider) and also in text at the bottom of the animated image.

Some of the variables are pretty obvious, e.g. Temperature at 2m. You might need to do some research about what others mean, e.g. 4 Layer Lifted Index (LI). So the charts can be educational and help you to learn more about meteorology…that is if you’re like me and don’t understand it yet!

The animations run from hour 6 of the forecast out to hour 240, so about 10 days in total. The 0 and 3 hour frames aren’t produced because they will already have passed by the time the charts can be produced.

I’m finding it fascinating to watch weather patterns move over time on the charts and to see the impact of different weather patterns, e.g. where there’s often precipitation near an area of low pressure.

I’ve got some other ideas for different charts which I’ll try to create. If you see any other ideas you think it would be useful to see then please suggest them. I’ll also probably create charts for different areas.

I’m also considering if it would be possible to make copies of the files available for others to download for use on your own web sites. I don’t know how feasible this is because you’d need plenty of bandwidth to make them available and I know many people are on limited bandwith plans. To download the current 22 variables for three regions would be about 4GB per day, so I suspect this would need to be a limited number of variables for a single region to avoid overloading your bandwidth (and mine to a lesser extent).

Chris I just tried the charts on my tablet and they work surprisingly well, although I do need my reading glasses on :wink:


1 Like

I suspected it might not be too bad, but my tablet is out of action at the moment so I couldn’t test it.

The big problem with mobiles is the fixed width menu and fixed height top action bar take up most of the screen. The page is a horrible hybrid of a HTML5 animation script built into a Bootstrap page. The two scripts seem to fight each other for screen space so it took a lot of experimentation to get it working at all!

Nice! These are really good.


1 Like

Nice job Chris! They really look great - thanks for sharing.


1 Like

Hi Chris,

Great job. I’m creating my own charts from GFS with the OpenGrads scripts you designed earlier GFS maps - Denmark. Is it these scripts you use to create the maps?

You mentioned that [quote=“administrator, post:1, topic:72682”]
After a lot of experimenting with scripts I’ve managed to create Weather Charts

My script are failing everytime - after some time when running - with a message that data download takes too long. Can’t remember the exact wording.

But - in case - is there an update of the scripts that I missed or is there a workaround to prevent the OpenGrads script to fail?

BTW - I have the same challenge with the Hanis - not mobile friendly. Havent solved it yet to get the maps to resize and adapt to the size of the screen.


Hi Soren

The plotting part of the scripts is pretty much the same as previously posted here, but that isn’t the part that’s causing you problems.

My OpenGrads scripts are a modified version of the original scripts. I already download a lot of GFS data as GRIB2 files for processing to provide the WxSimate data. So I also use that local data as input to the Charts scripts rather than downloading the data using OpenDAP/DODs (?) as the published scripts use. I have an extra set of unpublished scripts which download the GRIB2 data and process it into a database for WxSimate to download. I don’t intend publishing these because they’re not simply download and run type scripts. They require additional components/utilities installed/configured on the server to work and they’d also need a lot of work to remove the WxSim/database functionality.

I seem to remember someone else mentioned that they were having problems with the OpenDAP data but I haven’t looked at that because I’m not longer using that connection.

My Hanis problems are made worse because I’m embedding the Hanis component inside a Bootstrap page. Bootstrap wants to control the screen layout so convincing both components to play nicely together was a big part of the work. I think at best I’ve created an uneasy truce between them, but I think I’ll need to go back again sometime to try to improve on what I’ve created so far.

Hi Chris,
Thank you for clarifying this. To avoid that my scripts are failing I only plot temperature and wind.

You are doing a great job on this - thank you