Obtaining READY data

Most WXSIM users have been set up (and all registered ones can be, at no charge) for automated download of GFS data from Chris McMahon's weather-watch.com site. He has very kindly set this up for WXSIM users (including himself) at no charge. He downloads large batches of GFS data in GRIB format and has written routines for culling out and interpolating data for each user that I (Tom Ehrensperger) enter in his system.

Understandably, though, there are occasional server problems. These have averaged maybe two or three times a year, and Chris usually fixes the problem within a day or two, but a back-up data source is appropriate. For this, we can turn to the previous (before Chris' system) primary source: NOAA's READY site. WXSIM users routinely used this for years, and now do only in 'emergencies', perhaps, but it's great, free, dependable data. The only problem is that download of it is not automated. In fact, the administrators of the site has a graphical password to prevent totally automated use of the site. This is for two very understandable reasons: (1) heavy use from automated access could crash the site and (2) they are the ones who created the site and have a right to be seen directly by users. It's a great site (and free, courtesy, ultimately, of U.S. taxpayers), with much more there than just the data WXSIM uses. I personally access it often, for a variety of products.

There is an almost automated way to obtain the data, using an AutoIT script developed by WXSIM users Jim McMurry and Stuart Rogers. You still visit the site, but with the script doing the clicking until you reach the password, which you put in yourself. This script pastes the data (optionally including MAPS soundings, for North American users) into a Notepad file, which is then saved as readydat.txt.

This is a good place to mention FOUS data, as well. This is the oldest format of external model data used by WXSIM. It is generated by the North American Mesoscale model (NAM, formerly called Eta) model and also by the Nested Grid Model (NGM). The NGM will actually cease operations March 3, 2009, but the NAM is an excellent model that should stay in use for the forseeable future. FOUS data has a few limitations compared to READY or Chris' GFS model: (1) FOUS is only available in or near the United States, (2) the resolution (temporaly and spatially) is not quite as good, and (3) it goes out to only 48 or 60 hours, so it's good only for fairly short term forecasts.

I present below an illustrated tutorial of gathering the data manually (which I think you should try even if you ultimately use the script) followed by access to the script itself. It's a long-looking tutorial, but once you've tried the process a couple times, you can probably do it in less than two minutes. The data is updated four times a day, so ideally you'd do this every six hours (a bit after 05, 11, 17, and 23 hours GMT), but in practice, you'd still get good forecasts doing it only once a day.

First, go to www.arl.noaa.gov/READYcmet.php. Enter your latitude (north is positive) and longitude (east is positive). If you're really close to an official site or city, you can just choose the identifier or select the city at right. Click 'Continue'.

Now choose 'Meteogram' and select a model. Here, I have chosen the GFS 0-84 hour, 3-hourly model. This is great for a three day forecast, as it has 3-hourly resolution. If you need a longer forecast, you can use the GFS 0-180 hour, 6-hourly, but the time resolution (6 hours) is not quite as good. (The automated GFS data, for which all this is a back-up, is 3-hourly for an enitre 180 hours). North American users can also choose the NAM, which is also 3-hourly for 84 hours, though some of the later options may look slightly different. Actually, I recommend that U.S. users use GFS (either the automated version or this READY data) in combination with NAM FOUS, to get input from both models.

When you selected the model meteogram, click the 'Go' button to the right.

You will then see this screen. Note that I've left 'Default' as the option. This is just the top part of the screen. Scroll down to see the the rest.

Here's the rest of the screen. Note that I've changed the selection to 'Text only' (though the graphics are certainly interesting to look at if you've got the time). Note that I've also typed the graphic password in the box. This is a step you have to do manually even if you are using the script I'll descibe later. After typing it in, press 'Get Meteogram'.

When the data appears, use the mouse to highlight everything starting with 'GFS Meteogram for location ...' down through the last line of the data (reaching the end of the +84 hour line). Some of the highlighted part is not visible in this screenshot. Use either Edit/Copy or (my favorite) <control>c to copy the text.

Open up Notepad and either Edit/Paste or (my favorite, again) <control>v to paste the contents into Notepad. If necessary, tap Enter to get the cursor at least one line below the bottom of the text, as indicated by the red ellipse below.

The reason for doing that is the WXSIM really wants more data! So, click your browser's Back button to return to the last screen. This time, though, click 'Choose from below'. The first item to select (using the top drop-down box) is Total Cloud Cover (SFC). Text only' remains selected.

For fields 2 through 8 select Winds Flags (SFC/3D), Temperature (3D), and then five cases of Relative Humidity (3D). Use the second column of drop down boxes to select SFC or the desired pressure level. For most users, this should result in the view below. The exceptions depend on your elevation. The values below are idea for users between about 500 feet (150 meters) and 1600 feet (500 meters) above sea level. Lower elevation users should choose 950 mb instead of 925 mb in the 3rd and 4th fields. From about 1600 feet to 2500 feet (800 meters), 900 mb is a good choice for those fields. Above that, that Relative Humidity should be omitted, and the Temperature (3D) changed to 700 mb (maybe 750 or 800 if you're below 4000 feet or 1200 meters).

The password should still be there in the box, and should still apply. Once again, click Get Meteogram.

Again, highlight and copy the data. Note that the top of the highlighted portion should be the line starting with 'FIELD'.

Paste this into Notepad, at the point where the cursor was from earlier, and save it as c:\wxsim\readydat.txt. Here's the whole text, but with hours 6-81 chopped out to make it shorter (scroll past it to see how to use it):

GFS Meteorogram for location: 33.64 -84.40
Another meteorogram Another product Another location Start over
Latitude: 33.64 Longitude: -84.40 &

LEVEL 2M 2M 850 MB 500 MB 500 MB
+ 0. 1017.8 10.7 10.4 8.5 558.0 572.7 0.45
+ 3. 1018.7 12.7 12.3 8.5 558.3 573.9 0.00
+ 84. 1002.0 9.5 8.5 8.0 558.7 560.7 14.43

LEVEL 10 M 10 M 925 MB 925 MB 850 MB 700 MB 500 MB 300 MB
+ 0. 100.0 190.8 3.6 13.0 97.5 93.8 29.7 23.2 40.0
+ 3. 99.8 193.9 4.8 13.7 93.4 93.8 66.7 15.1 62.5
+ 84. 100.0 30.6 6.6 10.0 86.3 82.4 90.1 97.8 96.8

Now you need to make sure WXSIMATE knows to include this data. First, UNcheck the GFS box. Then, check 'Include this saved file:' box and enter the path to your saved file, which will probably be c:\wxsim\readydat.txt. There's no need to make any changes in WXSIM's settings, as it recognizes READY data. It just doesn't like seeing both READY and Chris McMahon's format at the same time, though. When GFS data beocmes available again (and you can check daily as it's rarely out for more than a day or two), just reverse the checked status of these two boxes.

OK, most of us (including me) have been spolied by the hands-off, automated access to the GFS data, and having to go do anything manaully can seem like a chore, even if it's just two minutes! To ease this pain, Jim McMurry and Stuart Rogers have created a neat AutoIT script, which cuts the time down to probably under 30 seconds, including the time for you to copy the 6-character password and hit Enter. This script also, optionally, allows automated retrieval of MAPS sounding data (for North American users), which is generally better than RAOB, though differences in the resulting forecast are minimal.

Here is a link to a sample of this script. You can download it, read the comments, and modify it slightly as needed (your location, for example!). Some users have encountered bugs with it, which can result from different browser settings and also from differences in the way Notepad documents are titles (such as whether '.txt' explicitly appears in the title bar).

To use the script, you'll also need AutoIT itself. You can get it (there's a free version) from www.autoitscript.com.