Hey there, somewhat new user but I've been following the software for a minute now before digging in my heels this past severe weather season. I'm seeing this is the next iteration so I figured now is the time to mention things!

I'm curious on your thoughts on being able to bring in community weather stations from somewhere like Wunderground - I know nothing surrounding their API or how easy that'd be, but it could fill in the large data gaps that exist when doing something like a county view. We have two METAR stations in our county - not super impressive on a map.

I'd expect this to be a more manual undertaking client-side, maybe diving into a menu to get the stations setup. I wouldn't expect it to work out of the box. I'm not sure, let me know what you think.

Thanks

    mitchellcrazyeye WSV3 Tactical Mesoanalyst will feature a completely re-engineered surface observation ingest and display engine, some aspects of which I already did some predevelopment work on in early 2022. It will be designed to handle over 100,000 observations per hour and accommodate data flexibly from diverse networks.

    I have to spend a lot of time first creating the rendering engine and data ingest aspects before actually connecting point observation data sources (such as METAR) into the engine. The latter becomes a mere administrative task once I have completed the former.

    I will build an impressive repertoire from public domain data, though likely a data upcharge will be necessary for massive datasets (e.g. MADIS Mesonet obs), and with the possibility of partnering with entities who want to share/input/resell data.

    I don't believe Wunderground publicly shares their PWS data in public domain raw data formats, but this becomes a business relationship opportunity once the actual work is done of creating the engine. The engine is the important part - we can figure out what data is available to tie into it later. That will probably only happen very soon before release.

    This is the opposite of the older, horrible design where everything was client-side and adding new data sources required completely recreating new systems with custom and redundant code. For WSV3 Tactical Mesoanalyst I am creating very versatile high-level ingest engines with a server-side preprocessing component, so that the client PC can display/load data much more quickly, with far less client-side processing, and so I can add new datasets rapidly as an administrative vs coding task.

    I expect this design will come very much in handy for the surface observations system, since one can imagine the likelihood of adding support for many specific networks, depending on what's available.

    In the meantime, if you can let me know of any free, public domain, real-time observation data sources which are in excess of what's in MADIS, then that would be helpful to know about.

      3 months later

      Another update on the topic here of surface observations display is additional R&D I have done into a highly desirable feature. I intend for there to be some way of visualizing interpolated raster fields generated on the client from individual sparse point observations. Local NWS offices put out post-storm graphics doing this with storm reports in some sort of online slow/traditional charting program or web script. I believe it will be possible to do this in real-time on the GPU. There will likely be a tool to draw a box over an area of surface observations and instantly generate the interpolated raster field. The raster engine will support contour lines generically, so that will be very applicable here. This will be a highly valuable feature in some contexts. I invite related ideas/commentary.

      2 months later