weather

kalshiMarch 10, 202617d ago

Will the world pass 2 degrees Celsius over pre-industrial levels before 2050?

Will the annual global mean surface temperature anomaly reach or exceed +2.0°C above pre-industrial levels (1850-1900 average) in any calendar year before January 1, 2050?

View on kalshi

Signal

BUY

Probability

88%

Market: 78%Edge: +10pp

Confidence

MEDIUM

78%

Summary.

The market is pricing this bet at 78%, implying a roughly 4-to-1 chance that a single calendar year will reach or exceed 2.0°C above pre-industrial levels before 2050. My analysis estimates an 88% probability, suggesting the market is undervaluing this outcome by approximately 10 percentage points. The key insight driving this divergence is that individual years historically cross temperature thresholds 10-15 years before long-term climatological averages reach them, due to natural variability from El Niño events. With CMIP6 models projecting the long-term average crosses 2.0°C around 2052, we should expect individual years to spike above this threshold in the late 2030s to mid-2040s—well within our timeframe. Current data strongly supports this: 2024 reached 1.55°C (the hottest year on record), placing the baseline trend at ~1.35-1.40°C with only 0.60-0.65°C to go over 24 years. Even conservative warming projections (0.20°C/decade) would put the baseline at 1.75-1.85°C by the early-to-mid 2040s, where a single moderate El Niño event (+0.25°C) would breach 2.0°C. With 3-5 expected El Niño cycles between now and 2050, the cumulative probability of at least one coinciding with an elevated baseline is very high. The market appears to be either conflating "single-year spikes" with "long-term averages" or underestimating the cumulative probability over such a long window with multiple natural variability opportunities.

Reasoning.

Step-by-Step Analysis

1. Climatological Base Rate (Starting Point)

The key insight here is the historical pattern: individual years cross temperature thresholds 10-15 years before long-term climatological averages cross them. This is due to natural variability (primarily ENSO cycles) creating spikes above the baseline trend.

CMIP6 models project the long-term average crosses 2.0°C around 2052
Following the historical pattern, individual years should spike above 2.0°C starting in the late 2030s to mid-2040s (10-15 years earlier)
This strongly falls within our timeframe (before Jan 1, 2050)

2. Current State (March 2026)

2024: Hottest year on record at 1.55°C ± 0.13°C (with El Niño boost)
2025: 2nd/3rd warmest year; land temperatures already exceeded 2.0°C (2.03°C over land)
Current underlying trend baseline: ~1.35-1.40°C above pre-industrial
We're currently 0.60-0.65°C away from the 2.0°C threshold on the baseline trend
We have 24 years (2026-2049) for this event to occur

3. Warming Rate Projections

Conservative scenario (CMIP6 median, SSP2-4.5):

Warming rate: ~0.20°C per decade
By 2045: baseline reaches ~1.75-1.80°C
Add El Niño spike (+0.25-0.30°C): could reach 2.00-2.10°C

Accelerated scenario (Hansen et al.):

Warming rate: 0.27°C per decade (aerosol reduction effect)
By 2040: baseline reaches ~1.75-1.78°C
By 2045: baseline reaches ~1.88-1.92°C
Add El Niño spike: easily exceeds 2.0°C

4. Natural Variability Window

El Niño events:

Occur every 2-7 years on average
Add 0.2-0.3°C to global temperatures in peak year
Over 24 years, expect 3-5 significant El Niño events
Probability that at least ONE coincides with high baseline trend is very high

WMO already estimates 1% chance before 2030 (just 4 years). Over 24 years with rising baseline, cumulative probability is substantially higher.

5. Key Calculation

Even with conservative assumptions:

Baseline reaches 1.75°C by 2040 (14 years, ~0.20°C/decade)
Need only +0.25°C from El Niño to hit 2.0°C
Multiple opportunities (3-5 El Niño events) between 2026-2049
Each event has increasingly higher baseline to build from

Probability of NO single strong El Niño in 24 years that adds 0.25°C when baseline is 1.75°C+: Very low (~5-15%)

6. Land vs Ocean Temperature Note

Land temperatures already exceeded 2.0°C in 2025 (2.03°C). The bet asks about "global mean surface temperature" which includes oceans, making it more conservative. However, this demonstrates how close we already are to the threshold on faster-warming components of the system.

7. Market Comparison

Market: 78% (0.78) My estimate: 88% (0.88)

The market appears to be underpricing this bet by ~10 percentage points. Possible reasons:

Market may be conflating "long-term average" with "single year spike"
Underestimating cumulative probability over 24-year window
Not fully accounting for natural variability upside
Conservative interpretation of acceleration in warming rates

8. Confidence Assessment

High confidence (0.78) because:

Recent data is current and authoritative (WMO, Copernicus, Berkeley Earth through 2025)
Multiple independent models converge on similar conclusions
Historical pattern (individual years exceed thresholds 10-15 years early) is well-established
Long timeframe (24 years) provides many opportunities

Moderate uncertainty remains on:

Exact timing and magnitude of future El Niño events (only predictable 6-12 months out)
Future emissions trajectory (though moderate scenarios still support YES)
Potential unexpected cooling factors (volcanic eruptions, though unlikely to persist 24 years)

Key Factors.

Historical pattern: individual years cross thresholds 10-15 years before long-term averages (strongly favors YES within timeframe)
Current state: Already at 1.55°C in 2024 (El Niño year); baseline trend ~1.35-1.40°C; only 0.60-0.65°C to go
Long timeframe: 24 years (2026-2049) provides 3-5 El Niño opportunities, each adding 0.2-0.3°C to annual average
Accelerating warming: Rate increased from 0.18°C to 0.27°C per decade according to Hansen et al. (aerosol reduction effect)
Model consensus: CMIP6, WMO decadal forecasts, and Hansen models all project individual years above 2.0°C before 2050
WMO baseline: Already 1% chance of 2.0°C before 2030; cumulative probability over full period is much higher
Natural variability asymmetry: El Niño spikes are more effective at pushing records higher as baseline rises

Scenarios.

Bull Case (Strong YES)

35%

Accelerated warming continues at 0.25-0.27°C per decade. A strong El Niño event occurs in the 2030s when baseline trend is at 1.65-1.75°C, pushing a single year above 2.0°C. This could happen as early as 2032-2037.

Trigger: Hansen's aerosol-reduction thesis proves correct. Strong El Niño develops (NOAA ONI index >1.5). Monthly temperature anomalies spike above 2.0°C threshold during peak warming months.

Base Case (YES with moderate timing)

53%

Standard warming rate of 0.20°C per decade continues. Baseline reaches 1.75-1.85°C by early-to-mid 2040s. A moderate-to-strong El Niño event (+0.20-0.30°C) coincides with this elevated baseline, pushing annual average above 2.0°C sometime between 2040-2048.

Trigger: CMIP6 median projections hold. El Niño cycle continues normal frequency (1 strong event every 7-10 years). At least one of the 2-3 El Niño events in 2040s coincides with peak baseline warming.

Bear Case (NO)

12%

Combination of factors prevents 2.0°C threshold: (1) Aggressive emissions reductions slow warming to <0.15°C per decade, (2) La Niña conditions or neutral ENSO dominate 2030s-2040s, preventing El Niño spikes, (3) Unexpected cooling factor (major volcanic eruption, rapid policy changes) emerges, or (4) Baseline only reaches ~1.75°C by 2049 and no El Niño exceeds +0.25°C in that window.

Trigger: Paris Agreement targets achieved with rapid decarbonization. Multi-year La Niña episodes in 2030s-2040s. Baseline warming slows measurably below 0.18°C/decade. Strong volcanic eruption (VEI 6+) causes temporary cooling in 2040s.

Risks.

El Niño timing uncertainty: Cannot predict specific El Niño events beyond 12 months; possible (though unlikely) extended La Niña dominance through 2040s
Emissions trajectory uncertainty: Aggressive climate policy could slow warming below model projections, though unlikely to prevent 2.0°C entirely given momentum
Volcanic cooling risk: Major volcanic eruption (VEI 6+) could temporarily suppress temperatures by 0.3-0.5°C for 2-3 years, narrowing window
Model over-confidence: Hansen's accelerated warming scenario may be too aggressive; if CMIP6 models are also running hot, baseline might not reach 1.75°C by mid-2040s
Aerosol rebound: Unexpected increase in aerosol loading (policy changes, industrial shifts) could slow warming
Definition/measurement changes: Though unlikely, changes in how global mean surface temperature is calculated or what constitutes 'pre-industrial baseline'
Ocean heat uptake: Increased ocean heat uptake could dampen surface warming more than expected, though this seems inconsistent with recent acceleration

Edge Assessment.

MODERATE EDGE FAVORING YES

My estimated probability (88%) is 10 percentage points higher than the market (78%), suggesting the market is undervaluing this bet.

Why the edge exists:

Confusion between metrics: Market may be conflating "long-term climatological average reaching 2.0°C" (happens ~2052) with "single calendar year reaching 2.0°C" (historically happens 10-15 years earlier). The bet asks about the latter.
Cumulative probability underestimation: Over 24 years with 3-5 El Niño opportunities and rising baseline, the cumulative probability is very high. Market appears to be pricing this too conservatively.
Recent acceleration not fully priced: Hansen's findings about accelerated warming (0.27°C/decade vs 0.18°C) may not be fully incorporated into market expectations.
Natural variability upside: El Niño events create asymmetric upside risk for temperature spikes, which the market may be underweighting.

Market stability note: The market has been stable at 77-79¢ over the past week, suggesting no recent information flow. This stability at 78% may reflect anchoring or incomplete analysis rather than informed disagreement with the 88% estimate.

Recommended action: There appears to be value in taking the YES side at 78¢, with an estimated edge of ~10 percentage points. The long timeframe (24 years) and multiple pathways to resolution support a higher probability than the market currently prices.

Caveats: This is a very long-term bet (resolves 2050) with significant uncertainty in climate dynamics, emissions pathways, and natural variability. While I have higher confidence in YES than the market, the 0.78 confidence level reflects legitimate uncertainty in timing and magnitude of future warming and ENSO events.

What Would Change Our Mind.

Extended multi-year La Niña dominance through the 2030s-2040s preventing El Niño temperature spikes (would require unprecedented ENSO pattern shift)
Observed warming rate falling consistently below 0.15°C per decade over the next 5-10 years, indicating aggressive emissions reductions are working faster than expected
Major volcanic eruption (VEI 6+) in the late 2030s or 2040s causing sustained cooling that narrows the window for threshold crossing
Revision of pre-industrial baseline definition or global temperature measurement methodology that materially changes the effective distance to 2.0°C threshold
Evidence that Hansen's aerosol-reduction acceleration thesis is incorrect and CMIP6 models are systematically running hot, with actual baseline warming tracking below 0.18°C per decade
Unexpected increase in ocean heat uptake that dampens surface warming more effectively than current models project, keeping the 2040s baseline below 1.70°C

Sources.

Market History.

Market has been relatively stable in the last 24 hours (currently 78¢). 7-day range: 77¢ – 79¢.

Get This Via API.

Access real-time prediction market analysis programmatically. Every analysis on this page is available through our REST API.

curl -X POST https://api.rekko.ai/v1/analyze \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"category": "weather", "platform": "kalshi"}'

Get API Key Read Docs

Related Analysis.

weatherkalshi

SELL

California 8.0+ magnitude earthquake before 2035

The market is pricing a California 8.0+ magnitude earthquake at 40.5% probability through December 2035, representing a massive 17.6x overpricing relative to scientific consensus. The USGS UCERF3 model—the gold standard for California earthquake forecasting—estimates just 7.0% probability over 30 years, which adjusts to approximately 2.3% for the remaining 9.8-year timeframe of this bet. My estimated probability is 2.5%, accounting for modest tail risk from the Cascadia Subduction Zone and model uncertainty. The market appears driven by recency bias from yesterday's routine M3.6 Santa Rosa earthquake and February's San Ramon swarm activity, evidenced by the 2.5 percentage point price increase on only 12 contracts of volume. Critically, no M8.0+ earthquake has occurred in California in modern seismological history, the San Andreas Fault's physical maximum is approximately 8.0 (requiring an unprecedented full-margin rupture), and recent seismic activity falls well within normal patterns that do not predict major events. The wide 35¢ bid-ask spread and low liquidity indicate market inefficiency, creating a strong opportunity to bet NO at significant value. The edge is substantial (38 percentage points) and backed by rigorous geological science, though capital will be tied up for nearly a decade.

3%Mar 9, 2026

weatherkalshi

SELL

Will a supervolcano erupt before 2050?

The market prices a supervolcano eruption before 2050 at 17.5%, but my analysis estimates the true probability at only 0.15% — a staggering 117x overpricing. This massive discrepancy stems from critical definitional ambiguity and apparent market confusion. Geological base rates show VEI 8 supereruptions occur globally once every 50,000-100,000 years, yielding just 0.02-0.05% probability over the 24-year timeframe. Current monitoring data from USGS and INGV as of March 2026 shows no precursor signals at any major caldera: Yellowstone operates at background levels with insufficient molten material (6-15%), and even elevated Campi Flegrei activity (Yellow alert, peaked 2023-2025) suggests at most a VEI 3-4 eruption, not a civilization-altering VEI 8. The market appears anchored to an ice core study showing 16.7% probability of a VEI 7 eruption by 2100 — conflating a smaller magnitude event (10x less than VEI 8) with a different timeframe. However, the resolution criteria stating "if a supervolcano erupts" without specifying VEI threshold creates substantial risk: if any eruption at a supervolcano site qualifies, the probability could increase 10-20x to 1-3%. This definitional uncertainty, reflected in my low 0.45 confidence level, is the primary risk factor preventing maximum conviction despite apparent massive mispricing on geological merits alone.

0%Mar 8, 2026

weatherkalshi

SELL

Will a supervolcano erupt before 2050?

The market is pricing a supervolcano eruption at 23.5% probability through 2050, representing a catastrophic ~587x overvaluation relative to geological reality. Based on well-established VEI 8 eruption base rates (once per 50,000-100,000 years globally) and current monitoring data showing zero elevated risk signals, the true probability is approximately 0.04% (1 in 2,500) over the 24-year period. As of March 2026, Yellowstone remains at NORMAL alert level with a "nearly solid" magma chamber, paused ground uplift, and background seismic activity. Recent geyser eruptions are routine geothermal processes being sensationalized by media, driving uninformed speculation (volume spiked 2.2x with 4-point price movement in 24 hours). VEI 8 eruptions cannot occur without years-to-decades of detectable precursors (massive ground deformation, thousands of earthquakes, gas emission spikes)—none of which are present. The 23.46 percentage point edge represents extreme market inefficiency driven by sensationalist headlines, geological innumeracy, and the 24-year capital lockup deterring informed traders.

0%Mar 9, 2026