Cumulus Support

Gentlefolk-
As I start this thread, a lot of people are experiencing what I'd call fragility with Weather Underground. I've been posting to WU most of this decade. For me it begged the question of what are the pros and cons of the alternatives available. Cumulus MX natively supports a variety of options:

• Twitter
• Weather Underground
• PWS Weather
• Met Office WOW
• Weatherbug
• APRS/CWOP
• Awekas
• WeatherCloud

I'd like to start a discussion of the choices CMX provides - and I would hope to stay as objective as possible. Comparisons between various alternatives should be useful. There's considerable Weather Underground bashing I'm reading today, and I get (and feel) the frustration. Still, I'll urge that we try to keep this positive - and help our colleagues who may be looking to make a change.

If I understand the way the forum is sometimes used, I believe I can alter this initial post and add pointers to other sources of information as we come across them. Here's a start...

• Sharing a Station - Brief page touching on several options.
• at System.Net.WebConnection.HandleError (this forum) - the thread in which I started thinking about this one. Maybe this is temporary to get this listing started with something. There PaulMy mentioned his experience briefly with all of these!
• To share or not to share: Drivers and barriers for sharing data via online amateur weather networks - This is really an academic paper, which doesn't actually compare existing options, but lists a number for reference. It may be interesting if you want to take a giant step back from the more specific intent of this thread.

I'll post what I learn as I tinker with some of these options in hopes others may find the discussion useful!

Cheers,
Bob

Hi Bob,
Well probably not what you wanted (a negative) though in my personal experience uploading to all your mentioned sites for many years (aside from Weather Cloud) and (WOW was short lived to WoW AU).
I never received any form of feedback after requesting from many sites, how or even whether they actually used my data.
In addition to that I always wounder if too much data is confusing. By that I mean if the recipients systems are really good at forecasting but are lack luster on filtering out bad injection from poorly managed sites/data (results in bad forecast).
Also on another point, if badly/poorly setup sites are injected the again (bad forecast).
I would also pose the question, how often do PWS owners go to each of these sites and modify previously uploaded data, which they have had to fix on their own site due to an issue for what ever reason the day/week/month before or been down for an extended period etc.
How can software determine that a PWS is positioned correctly, is maintained correctly and is configured correctly ?
I just really question the why/how/should of it all.
Personally as a result of my own Delmar on this subject I elected a while back to not upload to any.
I just use my own onsite WXSIM forecast, which uses (yes) GFS Data, along with my own.
Sorry Bob, Just my personal view. Quality Control and Data Validation I feel is the core issue.
Kind Regards,

You certainly make some valid points, and if negative, you offer a well-reasoned assessment in my opinion. I'm clearly much more more of an amateur than you and many others in this game. From a very quick skim of the "academic" paper I cited above, you might find it interesting. Data quality is decidedly all over the place among this community.

I happen to know enough to know that my sensors are not optimally situated, and probably can't be on my property given the very tall old-growth trees I have. Also, my PWS is decidedly not high-end. That said, I periodically check some of the closest PWSs, and the local airports and have found that generally what my PWS reports is reasonably consistent with what others are reporting. That's something of a benefit, I think, in participating in a network. (Not a proper calibration of course.)

Your comments suggest a possible enhancement to aggregators - they could allow one to at least self-rank the quality of the observations, perhaps based on a questionnaire. Although my data "tracks" well, I would certainly admit I've made compromises in siting and recognize the limited quality of my gear. (I'm also open to learning and improving, and some of these sites do help the community with that.)

My own site has some real rough spots I'd love to work on as time allows, though the essential data is there and updated regularly. I think I've read more negative comments about various sites, but then I suspect folks are more inclined to complain online about problems than cite what is working as expected (I'm guilty of that at times). One undoubtedly has almost immeasurably more control of one's own site.

Cheers!
Bob


- I'll also admit I'm personally more software/hardware engineer than weather enthusiast - focused of late on understanding the Raspberry Pi - where my site could use some work.