Forecast Validation

Step 1 — Record. Every day we save what 6 weather models predict for 80 cities across India.

Step 2 — Check. Next day, we check what actually happened using global weather records (ERA5 reanalysis — the same data weather agencies worldwide use).

Step 3 — Score. We compare every prediction against reality and update the scores on this page.

Ranking (#1, #2, ...)

Models are ranked by overall accuracy — balancing rain detection and false alarms. #1 is the most accurate model for the selected region and time period.
Technical: Ranked by Critical Success Index (CSI)

“Catches rain?”

When it actually rained, how often did this model predict it? Higher is better. 47% means it correctly predicted 47 out of 100 rain hours.
Technical: Probability of Detection (POD)

“False alarms?”

When this model predicted rain, how often was it wrong? Lower is better. 72% means 72 out of 100 rain predictions didn’t happen.
Technical: False Alarm Ratio (FAR)

Rain events

“Caught” = model predicted rain and it rained. “Missed” = it rained but model didn’t predict it. “False alarms” = model predicted rain but it stayed dry.

• Ground truth data has ~25 km resolution — can't verify street-level accuracy.
• Archive data is 24-48 hours delayed.
• MausamGram (IMD) is only available ~64% of the time.
• Pre-monsoon thunderstorms are inherently hard to predict — low scores don't mean models are broken.
• First season of data collection — scores become more meaningful over time.