Polymarket Weather Markets Explained: How Daily Temperature Trading Works

What Are Polymarket Weather Markets?

Polymarket weather markets are prediction markets where you trade on measurable meteorological outcomes — primarily the daily high temperature in a specific city on a specific calendar date.

Unlike sports or political markets where the outcome is subjective or narrative-driven, temperature markets resolve against a single, verifiable data source: a physical weather station reading, captured once per day, at a named airport. There is no ambiguity about what “wins.” Either the temperature hit that range or it didn't.

That objectivity is precisely what makes them attractive to systematic traders.

As of 2026, Polymarket runs daily temperature markets across roughly a dozen cities: Shanghai, Tokyo, Beijing, Hong Kong, Seoul, Taipei, Wuhan, London, Paris, New York, and Los Angeles are the most consistently active. The city list shifts as Polymarket adds and rotates markets — a market discovery API means new cities appear without announcement.

How Daily Temperature Markets Are Structured

Every daily temperature market on Polymarket is a multi-outcome event, not a simple yes/no binary.

Here's what that looks like in practice. For “Highest temperature in New York on May 15,” Polymarket might list outcomes like:

• Below 60°F
• 60–61°F
• 62–63°F
• 64–65°F
• 66–67°F
• 68°F or above

Each outcome is a separate tradeable contract. You can buy YES (you believe this range will be the winner) or NO(you believe it won't). Every YES token pays $1.00 if it resolves correctly, and $0 if it doesn't. Every NO token pays $1.00 if a different outcome resolves.

The prices of all outcomes reflect market-implied probabilities. If the 64–65°F bucket is trading at $0.38, the market collectively believes there is roughly a 38% chance the day's high lands in that range.

Bucket Width: °F vs. °C

Bucket size varies by city and unit. Most US cities use whole-degree Fahrenheit buckets — typically 1°F or 2°F wide. European and Asian cities usually use 1°C buckets. Since 1°C equals 1.8°F, a 1°C bucket in London is nearly twice as wide as a 1°F bucket in New York. This directly affects how spread out the probability distribution looks and how aggressively the market prices each bucket.

How Resolution Works — The Detail That Matters Most

Every Polymarket temperature market resolves against a specific airport weather station, not a city-center sensor, not a weather app, and not a government average.

This is the single most important thing to understand if you want to trade these markets seriously.

Here are the confirmed resolution stations for the most active markets:

The reason this matters: airport temperatures can differ meaningfully from city-center readings. LaGuardia sits at the waterfront in northern Queens — it runs cooler than inland New Jersey and can differ from Central Park by 3–6°F. Paris-Le Bourget is a general aviation airport north of the city, consistently different from both CDG and the Montsouris meteorological garden.

Retail traders using a weather app set to “New York” will be looking at Manhattan. The market resolves at LaGuardia. That discrepancy is tradeable.

The Rounding Rule

Temperature markets resolve using whole-degree precision in the unit specified by the market. A station reading of 23.4°C resolves to the 23°C bucket. A reading of 22.9°C also resolves to the 22°C bucket. Bots that model this truncation correctly — rather than assuming rounding — make a different set of trades, particularly near bucket boundaries.

Data Source: Wunderground

The resolution data comes from the Weather Underground (“Wunderground”) page for the named station, after daily finalization. Intra-day METAR observations during the day aren't sufficient — the market waits for the final daily record. This creates a brief window between when the temperature “locks in” physically and when the market officially resolves.

Who Resolves the Market? The UMA Oracle

Polymarket doesn't employ humans to sit and verify temperatures. Markets resolve via UMA's Optimistic Oracle, a smart contract system on the Ethereum blockchain.

Here's the process:

1. After the market's close window, a proposer posts a settlement bond (roughly $750 USDC) and asserts a resolution outcome.
2. A two-hour “liveness window” opens during which anyone can dispute the proposal.
3. If undisputed, the market settles automatically. If disputed, a new proposal is requested; if disputed again, UMA's token-holder vote adjudicates.

In 2025, UMA migrated Polymarket to a system called MOOV2, which restricts who can submit resolution proposals to a whitelist of approximately 37 vetted addresses. This dramatically reduced “griefing” and incorrect premature resolutions. For daily temperature markets, resolution is almost never disputed — the data is objectively verifiable. The typical time from market close to payout is two to four hours.

The Market Lifecycle: From Open to Resolution

Understanding the timeline of a daily temperature market helps you understand when and why liquidity behaves the way it does.

• 5–7 days before the date: The market opens. Early prices are wide and sparse. Volume is very thin.
• 2–4 days out: Liquidity starts building. Forecasts are sharpening; the ensemble spread is narrowing. Algorithmic traders begin taking early positions if edge exceeds their thresholds.
• 24–48 hours out: This is where the market comes alive. Short-range models (HRRR, NBM, NAM) are now the dominant inputs. Volume spikes.
• Final 12 hours: Most of the volume and the largest price moves. Real-time METAR observations from the resolution station are now visible.
• After 2pm local time: The daily high at most mid-latitude stations is typically reached between 2pm and 5pm local time. As this window passes, prices in losing buckets collapse toward zero.
• After station data finalization: UMA proposers submit resolution. Market settles within ~2 hours.

What Drives Prices: Forecasts, Models, and the Crowd

Polymarket weather market prices are driven by a combination of:

1. Numerical weather prediction (NWP) models.ECMWF's IFS model is widely considered the world's best for medium-range temperature. The US GFS updates four times daily. The HRRR updates hourly and dominates inside the 18-hour window. Each model run that shifts the forecast meaningfully creates a window where Polymarket prices haven't yet adjusted.

2. Ensemble forecasts.Rather than one “best guess,” ensemble models run the forecast 31–51 times with slightly perturbed initial conditions. The spread of those runs directly tells you how confident the atmosphere is being — a tight cluster of members in the 64–65°F range is a very different situation than members scattered from 61–70°F.

3. Retail flow. Many casual Polymarket users trade weather on intuition — recent experience, weather app glances, or rough pattern matching. This creates systematic biases that experienced traders exploit.

4. Competing bots. A meaningful fraction of weather market volume comes from algorithmic traders running calibrated forecast models. As more bots enter, edge windows shrink and prices become more efficient over time.

The Negative-Risk Architecture

Polymarket's daily temperature events use a structure called negative risk, which has an important implication for how you can trade them.

Because only one bucket can resolve YES, each YES token and every NO token across all other buckets in the same event are economically linked. Specifically: holding one NO token in any given bucket is equivalent to having a claim on all other buckets resolving YES.

This means you can use $1 of collateral to buy YES in multiple buckets simultaneously, a strategy called “laddering.” Instead of concentrating all your capital into the single most-likely bucket, you spread across the 3–4 most probable buckets. Whichever one wins pays $1; the others pay $0 — but the diversification reduces the variance while retaining the expected-value upside.

It also creates a clean arbitrage: if the sum of YES prices across all buckets drops below $1.00, you can buy all of them for less than a dollar and collect $1 at resolution, risk-free. In practice this is rare on actively-traded markets but appears occasionally on thin international markets.

Why Weather Markets Are Different From Other Prediction Markets

Most prediction markets on Polymarket depend on events that are non-repeating or seasonally concentrated. Weather markets are different in three ways that matter for systematic trading:

Daily frequency. There is a new market for NYC tomorrow, and the day after, and every day. This means a calibrated forecaster gets thousands of trials per year, enabling statistical validation of edge. A political or sports bettor might get dozens.

Mechanistic signal.The driving variable — the future temperature — is physically forecast by some of the most sophisticated models ever built. Those models are public, constantly updated, and their skill is empirically quantifiable. This is very different from “who will win the election.”

Hard resolution.A thermometer doesn't lie. There's no ambiguity, no announcer's judgment, no rule interpretation. The temperature either was in the range or it wasn't.

Who Trades Polymarket Weather Markets?

The Polymarket weather leaderboard is public and on-chain, so anyone can see who the top earners are. Top earners include traders with handles like gopfan2 (reported $343K+ profit), ColdMath ($120K+), and Hans323 — a 23-year-old German law student who earned a reported ~$80K using primarily algorithmic strategies around model update windows. The Insurance Journal profiled several of these traders in April 2026, noting that weather markets had become a quiet but serious category for quantitative traders.

The majority of retail traders lose money — which is the normal equilibrium when algorithmic participants have better data, better probability estimates, and faster execution.

Getting Started: What You Actually Need

To trade Polymarket weather markets, you need:

1. A crypto wallet — MetaMask or any Ethereum-compatible EOA wallet.
2. USDC on Polygon — Polymarket runs on the Polygon PoS chain and settles in USDC.e.
3. Account registration — Polymarket uses an email-linked proxy wallet system.
4. Geographic eligibility — Polymarket blocks approximately 33 countries, including the United States and most of Western Europe.

The minimum practical bet is a few dollars; serious automated traders operate with bankrolls of $1,000–$50,000 and run dozens of simultaneous positions.

The Bottom Line

Polymarket weather markets are not a lottery. They are mechanically well-defined, resolve against objective data, and repeat daily — which makes them unusually suitable for systematic, model-driven trading. The players winning consistently are not guessing. They are running calibrated probability estimates from meteorological models, sizing bets mathematically, and executing faster than the retail traders setting the initial prices.

Related articles

• How to Trade Polymarket Weather Markets: A Complete Step-by-Step Guide
• Polymarket Weather Predictions: How the Odds Are Set and What Drives Them
• The Polymarket Weather Leaderboard: Who's Winning and How They're Doing It