Relationship intelligence
The climate has patterns.
Do they still hold?
JADE measures how environmental relationships change through time — what remains stable, what is drifting, and how much confidence the evidence justifies.
Don't assume stationarity. Measure it.
SOI → Townsville wet-season rainfall
- Historical relationship
- Moderate
- r = +0.37 to 1999
- Recent relationship
- Moderate
- r = +0.33 last 20 seasons
- Direction
- NO SIGNIFICANT CHANGE DETECTED
- Confidence
- High
- Structural change
- Not established
Full-period Pearson r = +0.39 over 75 wet seasons. A moderate association remains. The recent window is still moderate. li Jade does not upgrade that into a regime shift.
Open the relationship record →- 01
What do we think we know?
Current state, the relationships we rely on, and the relevant forecasts.
- 02
How reliable is that knowledge?
Confidence, historical skill, relationship stability, and evidence quality.
- 03
What appears to be changing?
Drift, structural-break candidates, analogue mismatch, and changing forecast skill.
Same signal. Different system.
1997 ↔ 2026 analogue
NOAA OISST v2.1, 18 Aug 1997 ↔ 18 Aug 2026, same climatology. ENSO similarity is not whole-system similarity.
Open Analogue Lab →Relationships move.
Temporal graph
Nodes are variables. Edges are relationship objects with strength, direction and confidence. The system is not only changing in state.
Open the graph →Test the pattern.
Townsville Regime Lab
Full-record annual anomaly, 2019 and 2025 marked. The 10–11 year cycle is a test that is allowed to fail.
Open Regime Lab →