SPC outlooks

Base URL: https://api.wxalerts.org

The Storm Prediction Center’s convective outlooks: Day 1 to Day 3 categorical, and the Day 1 tornado, hail and wind probabilities. One route lists what is issued, the other returns one outlook’s risk areas.

GET /outlooks
GET /outlooks/day1-categorical/geojson
GET /outlooks/day1-tornado/geojson

No key needed. Browsers block calls from other websites (CORS allows only wxalerts.org), so call it from a server, script or app. Responses carry Cache-Control: public, max-age=300, and SPC is polled about every five minutes, so there is nothing to gain from asking more often. Limits are on the overview.

Which outlooks are issued

GET /outlooks lists the six map products without geometry. Build a product picker from it rather than from a hard-coded list: a product that has never been fetched is absent, and one whose polling is failing says so.

{
  "server_time": "2026-11-09T19:06:41.749666Z",
  "outlooks": [
    {
      "product": "day1-categorical",
      "day": 1,
      "hazard": "categorical",
      "label": "Day 1 categorical",
      "note": "Today's severe risk in the five named categories, from General Thunderstorms through High. Valid to 12Z tomorrow.",
      "issued_at": "2026-11-09T16:09:00Z",
      "valid_from": "2026-11-09T16:30:00Z",
      "valid_to": "2026-11-10T12:00:00Z",
      "forecaster": "Mosier/Moore",
      "areas": 1,
      "empty_label": null,
      "fetched_at": "2026-11-09T19:03:05.075558Z",
      "changed_at": "2026-11-09T16:20:25.151796Z",
      "expired": false,
      "stale": false,
      "last_error": null
    },
    {
      "product": "day1-tornado",
      "day": 1,
      "hazard": "tornado",
      "label": "Day 1 tornado probability",
      "note": "The chance of a tornado within 25 miles of a point, today. The hatched area is where 10% or more of those would be EF2 or stronger.",
      "issued_at": "2026-11-09T16:09:00Z",
      "valid_from": "2026-11-09T16:30:00Z",
      "valid_to": "2026-11-10T12:00:00Z",
      "forecaster": "Gleason",
      "areas": 0,
      "empty_label": "Less Than 2% All Areas",
      "fetched_at": "2026-11-09T19:03:05.075558Z",
      "changed_at": "2026-11-09T16:20:25.151796Z",
      "expired": false,
      "stale": false,
      "last_error": null
    },
    "..."
  ]
}

Outlooks come back ordered by day, then hazard alphabetically.

One outlook’s areas

GET /outlooks/{product}/geojson returns a GeoJSON FeatureCollection with two extra members: server_time, and outlook, the same summary object as above. There is no bounding box parameter: an outlook is a handful of national polygons totalling a few kilobytes, and the point of the layer is the shape of the day across the country.

{
  "type": "FeatureCollection",
  "server_time": "2026-11-09T19:06:50.484971Z",
  "outlook": {
    "product": "day1-categorical", "day": 1, "hazard": "categorical",
    "label": "Day 1 categorical",
    "note": "Today's severe risk in the five named categories, from General Thunderstorms through High. Valid to 12Z tomorrow.",
    "issued_at": "2026-11-09T16:09:00Z", "valid_from": "2026-11-09T16:30:00Z", "valid_to": "2026-11-10T12:00:00Z",
    "forecaster": "Mosier/Moore", "areas": 1, "empty_label": null,
    "fetched_at": "2026-11-09T19:03:05.075558Z", "changed_at": "2026-11-09T16:20:25.151796Z",
    "expired": false, "stale": false, "last_error": null
  },
  "features": [
    {
      "type": "Feature",
      "geometry": {
        "type": "MultiPolygon",
        "coordinates": [[[[-101.06, 30.7], [-99.82, 31.12], [-98.8, 31.09], "...", [-101.06, 30.7]]], "..."]
      },
      "properties": {
        "product": "day1-categorical",
        "level": "TSTM",
        "rank": 2,
        "label": "General Thunderstorms Risk",
        "stroke": "#55BB55",
        "fill": "#C1E9C1",
        "significant": false
      }
    }
  ]
}

Coordinates are rounded to four decimal places (about 11 m) and are not simplified, so a boundary matches the one SPC drew.

Drawing it

Use stroke and fill. The colors of a convective outlook are the product, and readers already know how to read them.

Features come back lowest risk first, which is also the order to draw them: a Slight Risk area sits inside the General Thunderstorms area it is part of, and drawing them the other way round hides it.

On the tornado, hail and wind outlooks, significant: true marks the hatched 10% significant-severe area, and it is always the last feature. Hatch it rather than filling it. It is a second statement about ground the probabilities have already colored (“of those storms, 10% or more would be EF2 or stronger”), and a solid fill reads as a higher probability than SPC issued.

An outlook with nothing to draw returns zero features, with a 200. That is not an error. The summary’s empty_label carries the forecaster’s phrasing, and is how a client tells no risk (areas: 0 with an empty_label) from no data.

Every product

The six map products, the ones /outlooks lists:

The route also serves SPC’s other products, which /outlooks does not list. Read their summary from the outlook member of their own response.

Errors