Weather Underground used to let you publish forecasts.
Not just upload weather station data. Not just read model output someone else interpreted. Actually publish your own predictions — write up a forecast discussion, call the event, stake your name on what you thought was coming.
It wasn't a polished system. WU's forecast community lived largely in discussion threads and the hazard reporting system, not in any dedicated publishing tool. But it was something. Forecasters in the WU community built real reputations there. Regular readers came back before major events to see what specific community members were calling. People kept track, informally, of who was usually right.
That's gone now.
What Weather Underground removed — and when
The degradation happened in stages. By 2022, the personal forecast discussion community had largely gone quiet — not through a single announced shutdown, but through neglect, redesigns that buried community content, and a platform increasingly optimized for consumer ad impressions rather than forecasting discussion.
By June 2025, Weather Underground had removed its hourly, daily, and weekly forecasts entirely. Its webcam network was shut down. Customer service went silent. The platform that Francisco Partners (the private equity firm that acquired WU in February 2024) is running now shows current conditions. The community — including the forecasters who made it worth visiting — is simply no longer there.
If you were one of those forecasters, you've been looking for a weather underground forecaster alternative for a while. You probably haven't found one that actually fits.
Here's why — and what does.
What forecasters specifically lost (different from what station owners lost)
It's worth separating two distinct Weather Underground communities, because the alternatives are different for each.
PWS station owners lost a place to upload and share their personal weather station data — a community of observers contributing to a hyperlocal observation network. Their alternative needs are around data upload, station visibility, and community among other hardware owners. (If that's your situation, weather underground pws alternatives for station owners covers that workflow specifically.)
Forecasters lost something different: a place to publish predictions. If you were writing forecast discussions — calling the track before the storm, posting your snowfall map interpretation, debating model solutions in the days leading up to an event — the data upload platforms do nothing for you. Your problem was never uploading observations. Your problem was having a place to publish what you thought was going to happen, with your name on it, visible to people who cared.
The forecaster's loss was about output, not input. And that's a completely different problem to solve.
Why the obvious substitutes don't work
When WU's community faded, forecasters scattered to the usual platforms. None of them are a real weather underground forecaster alternative.
Social media
X, Facebook, Instagram — real distribution, zero structure. You can share a forecast graphic or a text prediction, but there's no permanent URL for it. There's no concept of a forecast event window or a predicted region. When the storm is over, your call exists somewhere in your feed history if you remember to search for it. The algorithm has no idea whether you were right, and neither does your audience in any systematic way.
More fundamentally, social media is optimized for engagement, not accuracy. The forecaster who issues the most dramatic prediction gets the most reach, not the forecaster who is most often correct. That's a poisoned incentive structure for anyone trying to build a reputation on actually being good.
Substack and blog platforms
Better than social in important ways — you own the audience, posts have permanent URLs, you're building toward something. Several serious forecasters publish on Substack, and it works well for longform weather writing.
But a Substack post doesn't understand what a forecast is. You're describing regions in prose and hoping readers can follow along. Your prediction for "the I-95 corridor from DC to New York" is however many sentences you wrote about it, not a drawn polygon with a value range attached. After the event, there's no system that matches what you called against what actually happened. You're manually posting a recap and hoping people believe you were right.
That's a meaningful improvement over social, but it's still not purpose-built for forecasting.
Windy and model visualization apps
These fill a completely different need. Windy is a model data visualization tool — a very good one. It shows you what the GFS, Euro, and NAM are saying. It doesn't let you publish a forecast or build a forecaster identity. It has no concept of a prediction you've committed to, a subscriber list that follows your work, or a verification record.
If you thought the Weather Underground forecaster community was centered on reading model data, Windy would serve you. It wasn't. It was centered on people making and sharing predictions. That requires a publishing layer that visualization apps don't have.
What a real weather underground forecaster alternative looks like
Here's what forecasters who were active in the WU community actually need:
A way to publish a structured prediction. Not a blog post describing a forecast — an actual forecast object with defined regions, predicted accumulation ranges or temperature parameters, and a specific event window. Something a reader can look at after the event and compare to what happened.
A named forecaster identity. A profile that's yours. Not a username on a forum thread, not a social media handle, but a forecaster page that aggregates your forecast history, your track record, your verified outcomes.
A subscriber list. People who followed your WU contributions weren't following WU — they were following you. A real alternative keeps that relationship between you and your audience, not between your audience and a platform.
An after-the-event record. The thing the WU community did informally — keeping track of who was right — needs to exist formally, with real observation data to back it up. A verified prediction is worth more than an unverifiable one.
ForecasterHQ is built to solve exactly this. Draw your forecast regions on an interactive map. Assign accumulation ranges, timing windows, temperature parameters by region. Publish — your forecast gets a permanent URL with your name on it. Readers can subscribe to your profile and get notified when you publish. After the event, NWS observation data is pulled automatically and matched against your drawn regions. Your profile shows the outcome: verified or missed, per region, with the full strip-plot visualization showing where observations actually fell relative to your prediction.
For more on the full publishing workflow, see how to publish a weather forecast online — it covers the step-by-step process from first draft to public URL.
The verification piece matters more than it sounds
When Weather Underground was active, forecaster reputations were built on informal memory. People in the community remembered who had been right about last year's ice storm, who called the March 2019 bomb cyclone track correctly, who had been consistently overblowing events for engagement.
That memory lived in the community. When the community scattered, the reputation didn't transfer.
ForecasterHQ's verification system is the formal version of that informal reputation. Every published forecast is timestamped. After the event window closes and enough observation data is available, the system matches what you predicted against what was recorded. Your profile page shows your cumulative track record — not just "I posted 47 forecasts," but "I published 47 forecasts and here's the verification record for each of them."
That's a portable reputation. It doesn't depend on a forum community to remember your calls. It's in the data.
The community hasn't vanished — it relocated
The people who made Weather Underground's forecast community worth visiting are still out there. They're on X, on Facebook groups, on Substack, in a few Discord servers that formed around regional weather watching.
What they don't have is a place designed for what they do. ForecasterHQ is building that.
If you were a WU forecaster looking for an alternative — specifically for the forecast publishing and verification use case, not the station data upload use case — this is the platform that fits what you were doing there.
For the broader picture of where the WU community went across all its different use cases, see the Weather Underground alternatives overview, which covers what happened to each part of WU's original community.
Getting started
Creating a ForecasterHQ page takes five minutes. Publishing your first forecast takes less than ten.
There's no learning curve specific to ForecasterHQ if you already know how to read a weather map and interpret model data. The platform is designed for forecasters who know what they're doing — the tools get out of the way and let you focus on the forecast.
Publish your first forecast free →
No credit card. No setup fee. Your forecasts get a permanent URL, a subscriber list, and a verification record that builds over time.
The WU forecasting community deserved a real home. This is it.