Weather Display version: 10.37S Build 134.
OS Weather Display is running on: Windows 10 Home
Weather station hardware: Davis VP2+
Not sure if this is really a bug but might be an artifact of graphing? The highest wind gust recorded so far today was 27 mph at 0719 but on the graph it plotted the gust speed at 22 mph as shown in the attached image. Any thoughts on why the discrepancy between the measured gust speed and graph plot?
Still looking for suggestion/solution for this issue. Again, yesterday, the WD graph did not display the maximum gust. The maximum gust was 33 mph at 2035 but as shown in the attachment, the gust at 2035 was plotted as 27 mph. There was also a gust of 32.5 mph at 1825 that was plotted as 26 mph.
Not sure what would be causing that, but the fact that the tabular gust value is correct shows that WD has got the correct value. Its just the graph which is not displaying it.
I have this happening with my wind graph, but that is because I use the ‘Blocked In’ wind speed graphing mode, which clearly you are not using. In my case the correct gust appears every other minute as the graph updates. I take it that your gusts are permanently missing ?
Its possible the new compiler causes individual gusts to be missed, although I have not seen anyone else reporting this. Is it possible that your graph is not updating every minute and thus missing gusts ?
There are a few things to remember about WD and wind speed. Whatever units the station sends to WD, WD converts them to decimal knots. Then WD might have to convert back to decimal mph for the current/extreme conditions, but for the graphs it only converts to mph from integer knots. This means that max gust mph in extremes may well be different to max gust mph on the graphs. And a Davis VP only sends integer mph to WD in the first place. . .
That said, I agree that your differences look bigger than expected :?
Thanks for the replies. I can understand how converting units could have some small errors but, as bitsostring said, the differences are bigger than what would be expected.
Normally, if there are errors with WD, I usually assume it’s something I did wrong or I made a change with unintended consequences. But I haven’t made changes to WD in a long time, except to keep up with the current versions. As shown in the attached April 30, 2021 WD graph, the maximum wind gust of 43 mph was properly plotted.
I will double-check WD settings to make sure the graph is updating every minute per westhillman’s suggestion.
Must have been 42.6 mph = 37 knots or 43.8 mph = 38 knots . . . you can’t get 43.0 from integer knots 
You could try going back temporarily to build 120, the last build before the new compiler, to see if the problem was present then. (Build 120 is still available on the download page.)
Yup, tabular shows: 42.6 mph from 055°
I’ve attached my WD Data Logger, Wind Graph, and Scale (showing Wind Speed) Setup screen captures. The Data Logger is set to 1 minute. Any other location that needs to be checked?
Build 120 was created on 25 Sep 2020. I was using Build 125, created 14 Apr 2021, when WD correctly plotted the 43 mph gust on 30 Apr 2021, so I think that would rule out the new compiler after Build 120 as the issue.
Sorry, didn’t realise 120 was that old :oops:
I don’t know, I didn’t think there were any graph plotting settings: AFAIK it just plots every minute. Maybe the current/extreme conditions change more rapidly?
Yeah, I was thinking that if the gust occurred at a specific time within the one-minute plotting interval then maybe it wouldn’t display it. But if that’s the case, maybe it’s actually a software issue that could be fixed, which is why I posted in this Bug Reports section of the forum. I would think that regardless of when within a one-minute interval any of the measured phenomena occurred that the max or min would be plotted for that one-minute interval.
Looking at the one-minute data file for December, 2021lgcsv.csv, the gust speed at 2035 is 25 mph and at 2036 also 25 mph. So, it appears the graph is connected to the one-minute file? I’m guessing if WD correctly captured the 33 mph gust at 2035, maybe it occurred at just before or after 2035 and therefore was not included in the one-minute data file nor the graph?
The csv file for April shows 43 mph at 1635 and 1636, so maybe the gust needs to occur longer than a single one-minute interval in order for it to be plotted on the graph?
I think during the WD minute change WD is missing data because that takes usually some seconds…
You can probably see that in the WD data received or time below which freezes some seconds or missing counts.
Yes, I think that you are closing in on the core issue. WD plots the graph every minute, but also plots any higher gust that occurred during the past minute. However, if the data coming in from your Davis VP is only showing that minutes value, that is all that will be plotted.
With my own system, the La Crosse WS2800, its own computer screen software only shows gusts every minute, but the system itself reports every 15 seconds or so. This data is picked up by WD and is thus plotted.
But there should be a way to plot the max gust as long as WD received it, as in my case. WD recorded the gust of 33 mph, so it had the data but didn’t plot it.
I’ve worked with data from a network of wind sensors at Kennedy Space Center and it was critically important for space launch to record and plot the peak wind speed. The wind sensors produced 1-second data and reported every minute in which it derived the mean wind speed from an average of 60 1-second observations and the peak wind speed reported was the maximum of the 60 1-second observations. The peak was not dependent on when it occurred within the 60-second time period. The software simply had a line of code to find and report the max observed within that 60-second period and also plot it graphically.
I also just looked at Weather Underground (see attached) and it showed the correct max gusts of 32.5 and 33 mph that WD sent it. So why can’t WD correctly plot the max gusts?
Yes, these are all valid points. Unfortunately Brian, the developer, is not currently answering queries on these pages. I suppose you could email him, but an answer might not be forthcoming.
Have you verified the correct value is in the logfile 122021lg.txt if so try converting the logfile to datafile/graph then recreate the last 31 days graph to see if that updates the wind graph
Good suggestion, but the logfile, which is the text version of the 2021lgcsv.csv file I mentioned in my earlier post, shows 25 mph.
IIRC in the past this topic has come up when graphs do not exactly match the data and the reason is because the graph plots every other minute. I don’t remember why but it may have something to do with the component used in WD or a limitation of that component. I could be totally off the mark, but this is what I remember of past discussions. It’s also possible that over the years Brian made changes to the code…
I remember something like that, too, when sun/moon rise/set appeared and disappeared every other minute. . . see here.
But build 125 was OK, apparently. . . :?
However, the 33 mph max that WD showed is not reflected in the one-minute data at all at the time of the gust nor 1-minute before or after. So, the recorded 33 mph never got to the one-minute data at all and therefore wasn’t plotted on the graph. In this case, at least, it didn’t matter if WD plots every other minute. Good idea though and might be true but doesn’t solve this issue.
