Co.De Triage / pass-quality forecast Operator dashboard →

Built on the SatNOGS open network

Score your
pass.

Before a satellite clears the horizon, its geometry is already known. This is how good the observation will be.

The model runs here, in your browser. No server call to score — nothing leaves your device.

Describe the pass

9 inputs
45°
8.0 min
435 MHz · UHF
Advanced geometry & RF
9600

Expected quality

loading…
···
relative pass-quality score
weakstrong

Loading the forecast model…

A relative ranking, not a probability of success. Use it to compare passes — the training sample is case-control, so absolute rates aren't calibrated.
Score vs peak elevationeverything else fixed

The science underneath

Three things make a pass good — and we proved them causally.

Not correlations from a black box. On 7,499 waterfalls across 402 stations, a case-control causal study estimated the odds ratio of each pre-registered factor, with partial-pooling for station and satellite identity and a battery of refuters. These are the effects the forecast is built on.

← worse odds of a good observationodds ratio (log scale) · better odds →

An honest instrument

What it knows, and what it refuses to.

An accurate-looking model is easy. An honest one is the point. This forecast is deliberately weaker than the numbers a naïve model would show — because most of that strength was an illusion.

0.80 The real accuracy is 0.80, not 0.95.

Measured across unseen stations (grouped cross-validation), the forecast scores AUC 0.806 ± 0.02. The tempting 0.88–0.95 came from letting the model memorise station coordinates and collection dates — skill that vanishes for a new operator. We report the number that survives.

✕ date It won't look at when data was collected.

In a naïve model the collection month alone carried 37% of the predictive power, and dates together 51% — pure artifact of case-control sampling. A future pass has no "collection month", so we removed the whole family: the leaked AUC dropped from 0.88 to 0.66, and the forecast was rebuilt from geometry and RF alone.

↗ prior Higher and longer can only help.

The study found peak elevation and duration improve quality (OR 1.84 and 1.58). We hard-wired that direction into the model as a monotonic constraint — so it can never tell you a higher pass is worse. Science as a guardrail, at no cost to accuracy.

⊘ noise "Daytime is better" — rejected.

Time of day looked significant (OR 0.72), but a three-way placebo refuter still found an "effect" after we shuffled the treatment. It was a structural artifact, not a cause. So the forecast ignores the hour entirely. No evidence, no feature.

7,499
waterfall observations
402
ground stations
696
satellites
9
features · geometry + RF only