Fisheries — Open Science / Run History

"Converged" here means the data-harvest pipeline completed without an unhandled error — it is not a claim of real model convergence. Reproducibility metadata is captured per run for transparency, not as an official scientific record.

Draw a New Region

Click "Start Drawing", then click-drag a rectangle on the map to define a new region's bounding box. This becomes a geographic filter applied to OBIS species-occurrence records — not an official survey stratum.

Show species occurrences:

No rectangle drawn yet.

Name:

Run a New Pipeline

Region: Models: FIMS WHAM Demo (fitted stocks)

EwE and SS3 aren't available from this form — EwE needs a paid AI step to build its diet matrix, and SS3's output has never been checked against real data. Both still work from the command line (see README.md).

Include iNaturalist community observations (inventory only) Use community observations as a proxy survey index for species with no real survey data — citizen-science bias caveat applies

Include Reef Life Survey diver-transect data (inventory only) Use RLS transect counts as a structured-survey proxy index for species with no real survey data or iNaturalist proxy fit — still not a calibrated survey

Data Source Coverage by Region

Row counts from each region's most recent pipeline run. A source showing 0 across every region for several runs usually means that source doesn't actually cover that geography (e.g. a survey dataset scoped to one coast) rather than a transient failure.

Pipeline Runs

	Region	Model(s)	Run Date	Git Commit	Converged	Files	Species
Loading…

Comparison

Metadata differences

Field	Run A	Run B	Changed?

Data source row-count deltas

Source	Run A	Run B	Δ