What you said, next to what happened.

How it works

1. Step 1

   Draw the storm and publish it

   The storm template: regions on a map, each with a range - 8 to 14 inches here, 4 to 8 there - and the hours it covers. Published under your name, on your page.

2. Step 2

   When it is over, pull the observations

   One click fetches every National Weather Service station observation inside your regions for the storm window, plus the Local Storm Reports collected by the Iowa Environmental Mesonet.

3. Step 3

   See them on your map

   Each observation is a dot, coloured and sized by what it measured, on the same map you drew. A strip for each region shows your range as a band and the observations along it.

4. Step 4

   Read it yourself

   A region with five or more observations is marked verified when at least one falls inside your range, and missed when none do. Fewer than five and it stays pending. That is the whole rule.

What it looks like

Say you drew three regions for a Mid-Atlantic storm: 8 to 14 inches across eastern Pennsylvania, 4 to 8 into Delaware, 1 to 4 on the Eastern Shore. Afterwards, the map holds the observations, and each region reads something like this.

• Eastern Pennsylvania, 8 to 14 inches - verified

  Seven observations, five inside the range. Philadelphia measured 9.2.

• Delaware, 4 to 8 inches - missed

  Six observations, all below the range. Wilmington measured 2.6.

• Eastern Shore, 1 to 4 inches - pending

  Three observations. Too few to say either way.

The strip for each region shows where the observations sat against your band - above, inside or below - so a miss tells you which way you leaned, not just that you did.

What it is not

• Not a grade. There is no hit rate, no accuracy percentage and no leaderboard. Every observation is kept, and shown on the forecast it belongs to; nothing is rolled up into a number next to your name.
• Not hidden. Once a storm forecast’s window has closed, the observations on it are public, whoever published it. A reader can see the same map you can.
• Not a model score. NOAA and the Storm Prediction Center verify gridded model output. This is for a forecast a person drew and published, which nothing else checks.

Where the observations come from

Station observations are fetched from the National Weather Service API (api.weather.gov) for the stations inside your regions. Storm reports are the Local Storm Reports published by NWS offices and collected by the Iowa Environmental Mesonet. Nothing is interpolated: if a station did not report, there is no dot.