WXSIM 2025 Build 1.0 released!

Hi Everybody! I’ve just posted the new version of WXSIM. If you’re an existing user, get the UPGRADE (www.wxsim.com/wxsimupg.exe or www.wxsim.com/wxsimupg.zip … and of course you may want to save your existing .exe files in another folder just in case any issues pop up - which I doubt, as I’ve tested it a good bit). If you don’t have WXSIM, I invite you to give it a go! In that case get the DEMO (www,wxsim.com/wxsimdem.exe or www.wxsim.com/wxsimdem.zip). Thanks, as always, for your interest and support! Here’s what’s new in this version:

The ‘Anthropegenic and Solar Forcings’ algorithms were changed slightly in response
to updated research - in part done in support of and resulting from output of my
Clima-Sim program, but also from further observations over time and more extensive
study of the rate of warming (whatever the causes) in different geographic regions.
The biggest change was a considerable reduction in the effect of the solar (11-year
sunspot) cycle. I had previously used direct calculations of rather immediate
effects of changes in the solar constant, which I’d considered could cause an
amplitude (either side of the mean) of almost 0.09 C. Now, both my modeling in
Clima-Sim and my very careful study of NASA’s GISTEMP data shows this amplitude to
be almost undetectable (I think this is largely a result of “thermal inertia, so to
speak), so I’ve reduced this amplitude to about 0.01 C. My recalibration of the
warming (which is a somewhat complex pattern, based on historical trends) results
in a slightly smaller warming that I’d had before for some Northern latitudes, but
slightly more than before over Southern Hemipshere land masses. For future warming,
I’ve adopted a trend close to the “RCP4.5” scenario, widely regarded as a mid-range
and very plausible eissions and warming future.

To allow a finer look at the warming rate, I’ve changed the temperature text box
display to two decimal places instead of the previous one (so now, hundredths of a
degree). This is displayed when you change the date in the text box at top left,
and reflects the “normal” mean temperature for that date. Note that customizations
are based on 1981-2010 climatology, with this factor “correcting” for that.

A header, “UTC Date/Time” has been added for the last column in latest.csv (or saved
versions of that).

The “Import Data” and “Import Local Data” options under the “Import” menu item
on the Data Entry Form now have addiotal wording: “(Access with Auto Run OFF)”. It
has not been obvious previously that Auto Run needs to be temporarily deactivated
(under Start/Auto Run and Other Settings) in order to made changes on those forms.
With Auto Run ON, the forms disappear before manual changes can be made.

The help files have been modified slightly, with additional wording and information
about the ‘Wind gust controls’ (under Preferences/Settings/Tab 1). This issue can
be confusing, so I recommend reading the blue text help there.

wret.exe was modified to correct a problem that occurred with the (relatively new)
Cumulus MX file name format (now numerical months instead of previous first three
letter os the month name). The problems had been that the previous version of wret.exe
would (with this file type only), revert to the first of two months when the analysis
went to the next month. Another, rarely encountered bug, in which an incorrect message
about the time interval not being appropriate for analysis, was corrected, by allowing
anything between 0.48 and 0.52 hours to “count” as 0.5 (some systems seem to have tiny
rounding errors of some sort).

A new messaging tool (msgprojwret.exe) was added for certain errors occurring in wret.exe,
to avoid the program needing to close. Putting the message in a separate program allows
wret.exe operation to continue, while still notifying the user of an anomaly (such as a bad
forecast file).

A further modification was made to the message alerting the user if the file for WXSIM
to import has a different name from the one euromix.exe is set to generate. For users
who do NOT wish to use ECMWF data (through euromix.exe), replying ‘No’ to this message
box will precent the message from appearing again.

A few messages about non-critical items (such as visibility being the maximum reported,
or difficulty reconciling cloud cover from relative humidities with GFS or ECMWF output)
have been reworded to clarify that these are NOT errors, but simply messages about which
the user could take (low priority) actions if desired.

A diagnostic file called wdatrec.txt will be created fresh whenever WXSIM accesses the
file world2.fdt. The reason for this is a ‘input past end of file’ error one user has
been randomly seeing once every few scheduled auto runs.

A very careful reassessment, adjustments, and enhancements were made to the solar
altitude and azimuth, and sunrise and sunset algorithms. Most of the existing routines
actually date back to my initial efforts in writing this program, in 1981, and WXSIM
‘grew up’ around that core. Now, I’ve gone back and very carefully modeled the ‘equation
of time’ (it having changed slightly in the last 44 years), with a greater accuracy than
before, reducing average errors from about 40 seconds to less than 3 seconds. I also
re-did the algorithm for solar declination, making it more accurate and up-to-date.
Futhermore, I properly took the leap year cycle and longitude into account, whereas
previously February 29 had been treated as Marxh 1, the the date was considered to be
GMT (it’s now local noon). A slight change was also made to the amount of atmospheric
refraction for sunrise and sunset calculations (note that, otherwise, solar altitude and
radiation calculations do not include atmospheric refraction, as the effect on temperature
is very small, and – to the extent it mattered at all – would have changed carefully done
calibrations). Sunrise and sunset times now are usually well within one minute of official
sources.

A small informational and default value change (from 55-60% to 50-60%) was made in
the recommended mixing weight for WXSIM-Lite data when including ECMWF with GFS.

2 Likes

Installed - no issues to report so far.

Guy

installed and all good so far…

Running fine here on Linux using Wine.

Stuart

1 Like

Thanks Tom, only catching up now

Now that I want to know how you did it. Do you have a write up anywhere?

I did do a write up some months ago but not in the public domain as it was not pretty and needed some others to test first. I have been running it now for quite some while in Linux and it is reliable but anyone wishing to try this needs some Linux knowledge and ideally knowledge of PHP as that is how I manage the scheduling in Wine as normal schedulers do not work in Wine as it is not Windows internally but mimics the Windows calls from programs.

Now if we could persuade Tom to rewrite WXSim (and the rest of the suite of programs) to run under Microsoft .NET 8 (now supported in Linux) we could run WXSim natively in Linux, as Mark Crossley has done with CumulusMX, which would be the ideal solution with scheduling being done via CronTab entries.

Stuart