How the AI Conditions Predictor Works

Racked uses an AI-powered conditions prediction engine to estimate whether routes are likely to be in condition, even when no one has submitted a recent report.

The predictor analyzes data from multiple free government sources including NOAA weather stations, SNOTEL stations (which measure temperature and snowpack near climbing areas), and USGS stream gauges. It factors in each route's elevation, aspect (which direction the cliff faces), and solar exposure to model ice formation and melt cycles.

The core metric driving the predictions is Freezing Degree Days (FDD) — a cumulative measure of how much sustained cold a route has experienced. More cold days below freezing means more ice buildup; warm spells mean deterioration.

The AI predictor runs daily and is designed to supplement — not replace — community reports. Think of it as an informed estimate when human reports aren't available.