Monarch Castle Technologies

Climate Security Index

Loading·—

0–90 day public-data outlook

Climate-security pressure

Agroclimate stress, multi-hazard impact, persistent natural events, and global food-price pressure. Not a conflict, famine, displacement, or casualty forecast.

—/100LOADING
Confidence—
Forecast horizon—
Cross-domain concurrence—
Data health—
Agroclimate Sentinel Radar
Latest 30 days against five matched annual windows
Warning History
Accepted autonomous runs · 0–100 scale

Reproducible by design

Methodology

The index estimates public climate-security pressure over the next 0–90 days. It does not predict conflict or humanitarian outcomes. Inputs are public and rules are fixed before each run.

35%

Agroclimate stress

Thirty-day rainfall and maximum-temperature anomalies at eight exposed agricultural/pastoral sentinels versus five matched annual windows from NASA POWER.

25%

GDACS impact pressure

Green, orange, and red multi-hazard alerts, weighted by level and cross-border reach, against a rolling 12-month robust baseline.

20%

EONET persistence

NASA-curated floods, storms, wildfires, drought and related events; open status, category and reported magnitude shape the score.

20%

Food-price pressure

Three-month changes and robust anomalies in FAO headline, cereals, and vegetable-oil indices, with year-on-year context.

Aggregation. Available components are reweighted to 100%. A five-point bonus requires both elevated climate evidence and the independent FAO price component. This prevents NASA EONET and GDACS overlap from earning multiple bonuses.
Autonomy. Sources cache for no more than 72 hours; recent validated components bridge temporary outages. Total physical-source failure preserves the last-good publication. Compact public snapshots avoid repository bloat.
Reproduction. GitHub Actions runs every six hours, executes tests, commits accepted output, and deploys the same data shown here. Formulae, evidence, scores, and URLs are in data/output.json.
Primary Sources
Official climate, disaster, and food-price data