Weather Edge
Forecast distributions against Kalshi's temperature buckets.
The tool runs at /weather and needs a $49.99/mo membership. There is no free tier and no trial, so opening it signed out lands here rather than on the board. This page is the open documentation of what it does.
Daily high-temperature markets are one of the few prediction markets with a real number underneath them. A forecast exists, it is free, and it resolves against a published climate report — which makes it possible to build an explicit model and compare it to the price.
Weather Edge does exactly that for seven Kalshi cities, and it is careful about the one detail that decides whether the whole exercise is meaningful: which point on the map you forecast.
Forecasting the settlement station, not the city
Each Kalshi series settles against a named climate report — a specific reporting station, usually an airport. The coordinates used here are that station's, not the city centre's. For a coastal pair like downtown San Francisco and SFO the two disagree by enough degrees to invert a bucket outright.
The seven series covered are New York, Las Vegas, Washington DC, San Francisco, Houston, Boston and Oklahoma City, each pinned to the station its contracts actually settle on.
Turning one forecast into bucket probabilities
The deterministic daily-maximum forecast is taken as the mean, and forecast uncertainty is modelled as a normal distribution whose standard deviation grows with lead time — tight for tomorrow, loose a week out. Integrating that normal over each market's temperature bucket produces a model probability for that contract.
Settlement is a whole-degree reading, so every boundary gets a half-degree continuity correction, applied according to which of the three bucket shapes a contract is: a floor only, a cap only, or a range with both. Getting those corrections in the right direction is the difference between a model and a plausible-looking wrong answer.
Reading the edge column
The edge is the model probability minus the market price, and the board sorts by absolute edge with a minimum threshold you set. Markets priced at or beyond a cent from either end are dropped, and the price used is the last trade, falling back to the mid only when the book is two-sided — half of a one-sided quote is not a price anyone traded.
Lead time is capped at seven days, because past that the normal is so wide it stops saying anything. Each row shows the forecast high, the lead time, the bucket, and both probabilities, so you can see the arithmetic rather than a verdict.
Outage versus quiet
If every city's fetch fails, the board raises an error rather than rendering an empty table. An empty table would read as no disagreement anywhere, which is a very different claim from the feed being down.
The count shown is the number of cities that actually answered, so the figure on the page is never larger than the data behind it. Results are cached for fifteen minutes, which is well inside the pace at which a daily forecast changes.
What it cannot tell you
Every tool has a boundary, and knowing where it sits is the difference between using one well and being misled by it. For Weather Edge:
- It is one deterministic forecast with an assumed normal, not an ensemble. The uncertainty curve is a modelling choice and it can be wrong.
- A disagreement is as likely to be the model's error as the market's. Traders in these markets often have better weather models than a free public forecast.
- Daily high temperature only, on seven cities. Other weather contracts are not covered.
- Forecasts revise. An edge seven days out is mostly uncertainty width, which is why the horizon is capped.
- Kalshi's per-contract fee eats small edges, and this board's edge column is before fees. Price the trade properly before acting on a five-point gap.
- Model probabilities are estimates. Nothing here is a forecast of settlement and nothing here is a promise.
What it is built on
- Kalshi trade-api v2, open markets for the seven daily-high series, including strike floors and caps.
- Open-Meteo's free forecast API, daily maximum temperature in Fahrenheit, in each city's own timezone.
- The station coordinates each Kalshi series settles against, as named in the contracts' own rules.
Related reading
- Kalshi Weather Edge: How to Price Daily High-Temperature Markets With a Free Forecast — Guide
- Trading Kalshi Temperature Markets Against a Forecast — Strategy
Tools that pair with it
One market, every signal we hold on it, on one screen.
What resolves when, across the markets people are actually trading.
Fee-aware position sizing, hedges, and a drawdown simulation.
WhaleTracks is informational analytics, not financial advice. Market data comes from Polymarket, Kalshi and Manifold's public APIs; WhaleTracks is not affiliated with any of them. Past performance does not guarantee future results.