Skip to content

CSU has published seasonal hurricane outlooks since 1984. Their credibility comes from the archive. Indie forecasters can build the same thing - ForecasterHQ's verification feature is the mechanism.

0-6"
Snowfall forecast of NOV/13-14/2026, 5 observations

Phil Klotzbach and the CSU tropical weather research team have been publishing Atlantic hurricane season outlooks since 1984. The first forecast was remarkably primitive by today's standards. The signal-to-noise ratio was modest at best. But they published it, they named it, and they kept publishing every year after.

That 40-year archive is the reason anyone calls CSU for comment when hurricane season starts. Not the PhDs. Not the research funding. The archive.

The archive is accountability. And accountability, compounded over time, becomes authority.

Most independent forecasters understand this intuitively. They post on social media during every storm, they update their readers as events unfold, they explain their reasoning. What they don't do - yet - is publish forecasts in a way that creates a verifiable record that survives the news cycle and can be scored against real outcomes.

That gap is exactly what Forecaster HQ's verification feature was built to close.


Why Social Media Forecasts Don't Build a Track Record

Here's the problem with informal forecasting: it's not inspectable.

When a storm ends, the social posts that said the wrong thing get buried. The ones that said the right thing get screenshot and shared. Survivors' bias is built into the medium. A forecaster who made 10 calls during the 2024 Atlantic season and got 2 right looks, in retrospect, exactly like a forecaster who got 10 right - if they're selective about what they promote.

This isn't a cynical observation about indie forecasters. It's a structural problem with social media as a forecasting medium. Even forecasters with genuine skill and honest intent can't build a credible track record on platforms that make old content hard to find and provide no mechanism for post-event scoring.

A public forecast track record requires three things:

  1. A forecast that was published before the event, with specific, measurable predictions
  2. An outcome record that captures what actually happened
  3. A scoring mechanism that compares the two

The first condition - publishing before the event - is surprisingly hard to fake at scale. The second and third conditions require infrastructure.


What a Verifiable Forecast Looks Like

On Forecaster HQ, a storm forecast consists of:

Geographic regions. You draw polygons on the map - the area where a specific outcome is predicted. This could be a coastal zone you expect to take the heaviest surge, or an inland corridor where you're calling for the highest snow accumulation. The geographic commitment is what makes the prediction falsifiable.

Quantitative ranges. Each region gets a min and max prediction value. For a winter storm, that's inches of accumulation. For a tropical storm, it might be rainfall totals, wind speeds by zone, or surge heights. Ranges are honest - forecasting is probabilistic - but they're also accountability-creating. If your predicted range is 8-14 inches and 22 inches falls, the record shows that.

A timestamp. The forecast is published before the event. The creation date is on record. There's no retroactive editing.

After the event ends, the verification step begins.


The Verification Workflow: Step by Step

Here's how Forecaster HQ's verification system works in practice.

Step 1: Create and publish your storm forecast.

Use the storm forecast type when creating a new forecast. Draw your regions on the map, assign your prediction ranges, and publish. The forecast is now public and timestamped.

If you haven't done this yet, the hurricane track forecast guide walks through the full creation workflow - same steps apply to any storm forecast, not just hurricanes.

Step 2: Wait for the event to end.

The verification fetch isn't available until the storm's event end time has passed. This prevents retroactive editing or cherry-picking the timing of when you "lock in" your predictions.

Step 3: Fetch verification data.

On your forecast detail page, once the event end date has passed, you'll see a "Fetch Verification Data" button. Click it.

Forecaster HQ pulls NWS station observation data for all stations that fall within your drawn forecast regions. It also queries IEM Local Storm Reports - the verified spotter and automated observation records - for event-specific data like snow totals and wind reports.

This process happens automatically in the background. Depending on event size and station density, it typically returns data within a minute.

Step 4: Read the strip plot.

The verification result shows you a strip plot - a simple linear visualization - for each of your forecast regions. The shaded band represents your predicted range. The dots represent individual observations from NWS stations and storm reports within that region.

If most of the dots fall inside the shaded band, your call was in range. If they fall outside - above or below - the record shows that, too.

Step 5: The "Verified" badge.

When at least five observations fall within your predicted range for a given region, that region earns a "Verified" designation. The badge appears on your public forecast and on your profile.

That badge is your archive. That's your CSU.


What the Track Record Looks Like to Your Audience

Here's the part that matters for credibility-building: the verification data is public.

Anyone who visits your Forecaster HQ profile can see all your past storm forecasts, which ones have been verified, and the strip plots showing predicted vs. actual outcomes. There's no curation possible. The full record is there.

This is uncomfortable for forecasters who've gotten used to informal channels where they can control the narrative. It's extremely valuable for forecasters who are genuinely good and can prove it.

The storm forecast verification tool page goes into more detail on how the NWS station data is sourced and what the verification threshold means in practice. For a broader look at what verification means for an independent meteorologist's credibility - and why it's the most powerful signal you can build for an audience - the forecast verification guide is worth reading alongside this one.

For now, the key point is this: a forecaster with 20 verified storm forecasts on their public profile has something that no social media presence can replicate. They have a record. And records - as CSU has demonstrated for 40 years - are the foundation everything else is built on.


Starting From Zero

You don't need to have been publishing for years to start building a track record. You need to start now.

The 2026 hurricane season is in full swing. The setup for active tropical development is present. Named storms are likely. That means opportunities to publish timestamped, geographic, quantitative forecasts before events form - and to verify them after they end.

Even if your first few forecasts are modest - a single region, a cautious range - the act of publishing creates the record. And the record is what compounds.

See an example of a live storm forecast โ†’

Start publishing - it's free โ†’


The credibility that matters to your audience comes from the record you build before you need it. Start now.

Publish your own storm map, and see the observations land on it

Draw the regions, set the ranges, publish. When the storm is over, the observations are drawn inside your regions.

Publish your forecast