I don’t think you have anything to work with yet. You need to have it just collecting forecasts for at least a couple weeks before trying any analysis. Are you doing that yet? If you are, you should have several files with names starting with gfsdat0… in the wxsimlitedata subfolder of c:\wxsim.
Hi Tom,
I get this every morning in the first run of the day at 08:00. Something to do with the download time of the file? There no errors in the log files that’s for sure. Here is the last fdata and the error screen.
Thanks
Peter
I notice that the date of the last WXSIM-Lite “forecast” is 4 days ago, which is what I would expect to see if the last thing WXSIM-Lite did (before the WXSIM forecast) was its analysis run. Please check to make sure you have a forecast run (in WXSIM-Lite) set to go after the analysis run (which should be in the early morning hours), AND before the first WXSIM forecast of the day. You might also look at the red status updates in the upper right corner of WXSIM-Lite to see whether or not the latest forecast was right before the WXSIM run.
The first time it did go away after a couple of days, but now I get the same since I moved to a new PC, for the last 5 days I get this every morning. Here are my schedules for all 3 apps, the log files are ok.
Any suggestions to where I look for any other settings or corrections to files, anything.
Petros
I have wxlite downloading and forecast twice in the morning just in case it misses the first time.
I’m not sure this will solve the problem, but it’s something that needs fixing in any case. The forecasts in WXSIM-Lite really need to be done AFTER the analysis run (which you have after 8 AM). If you are using WeatherLink , then it is good to have a WXSIMATE run (NOT a WXSIM run) at say, 1 AM to collect the previous day’s data. Then, WXSIM-Lite should probably run about 2 AM, and then your WXSIM forecasts should all be later (at the same hours as the WXSIM-Lite ones). You time gaps are appropriate (among the programs), but I do wonder why you do them so long after the hour. I think things would work better if you ran shortly after the hour (like WXSIM forecast about 12 minutes after). This is for coordination purposes with METAR data. The advection routine should be a bit more reliable that way.
Let me know if this helps the actual problem you are seeing.
Hi Tom,
Did all the time adjustment, but the problem was back this morning, did a manual download and forecast and then a full scan, the error showed up again. Any suggestions where to look next?
Petros
Hmmm … You need to do the Full Scan before the forecast, though I’m not sure that’s the problem. What I’m seeing on the screen suggests some very strange values in cormod0.txt, so maybe you could send me that. Also, try a Full Scan, with doing a forecast first. Again, I’m not sure that’s really a problem, but I’d like to be sure.
Ok Tom, what I did so far is, first I removed old the gfs data prior to the 1/8/2015 just to have enough data (2 weeks) so the analysis would run, did a foul scan before the forecast and after the forecast, the results are the same, removed cormod0.txt and fdata.txt so they can be recreated, still the same. put all the above files back in place tried again, still the same. Moved all files back to the old PC (winXP) just in case something is wrong with the new one, still the same. I don’t know what files are used for all this, but my logfiles are ok. Here are my files from the last run. Running on win7 64bit. All wxsim programs are run as administrator.
Thank you
Petros.
Hi Tom, any other suggestions as to where the problem could be? still running the same way as before. Can you outline the best time schedule for all 3 programs so I can try them?
Petros
There’s still something strange about your data. It’s a legitimate cormod0.txt file, but it just doesn’t look realistic. I’m not clear about how much data you tried removing … are you saying you removed all but two weeks, dating back to January?
I have two suggestions (let me know via email which you prefer):
(1) You could package up the last couple months of log files, along with a sample of gfsdat0…txt files from that period (maybe 50 or so of them, scattered through that time period, and send me those (along with sitelist.txt and wxlini.txt). I could then try to solve the problem here.
(2) You could allow me to remote in on your computer using Team Viewer. This might be in the middle of the night or something, as I’m back teaching again now and can work only in the evening here.
I sort of prefer #1, but am fine with #2 if you like. Just let me know!
I’m sorry, I can’t seem to find that email (and I don’t remember it - it was a busy week, though!). Is the problem still happening? Can you try sending the files again?
OK, just had a look. Very unusual situation here, it appears:
You are on an island. I looked at the raw GFS forecasts and they show almost no diurnal variation in temperature at all (whereas there is really at least several degrees difference at your station), so it seems GFS “sees” only water there! Basically, the raw GFS forecasts are so bad that WXSIM-Lite is reluctant to believe it really has to correct them that much! So, it’s giving a warning message. The subset of the data I looked at so far (part of August) would suggest that WXSIM-Lite will not be helpful there. I need to look at a larger stretch of data to see if it’s just a fluke, or a real problem with the island location. For now, I recommend letting WXSIM-Lite keep collecting data, but don’t try to use it just yet. We need to see if this is a permanent feature of GFS in your location.
Ok so you suggest not to use the wxsimlite for now? I’ll let it collect data and not use it with wxsim. I’ve always had a bit of a problem during summer months especially July and August the forecasts where 2-4 deg. out, the rest of the year they are very good. Thank you for looking at the files, let me know if there is anything else that you need to look at.
Petros
