Preparing workspace
Bhimashankar Wildlife Sanctuary
19.072 N, 73.536 E · 12 km radius
Loading map
Legend
Every weight and threshold below lives in one file: lib/constants/scoring.ts
lib/constants/scoring.ts
This is a transparent planning heuristic for prioritising where to look first. It is not a validated ecological assessment. The weights below are reasoned choices, not findings: they have never been calibrated against field outcomes, and no part of this application has been peer reviewed. Treat the output as a starting point for field validation, and never as a conclusion about a place.
index = Σ(componentScore × weight) / Σ(weight of AVAILABLE components) climate stress 35% conservation significance 25% habitat disturbance proxy 25% hazard exposure 15% Missing-data rules · An unavailable source is EXCLUDED, never scored as zero. · Remaining weights are renormalised over available weight. · Below 60% available weight, no score is published and the result reads "Insufficient evidence".
Precipitation is weighted below temperature because reanalysis rainfall is less reliable over steep terrain. Departure is scored on absolute magnitude: anomalous drying and anomalous wetting are both disruptive.
Threatened presence is normalised per 1,000 records so heavily-surveyed and lightly-surveyed areas stay comparable. Richness uses a log scale because species accumulation against survey effort is strongly sub-linear.
The land-cover term is down-weighted on purpose: absence of OSM mapping is not evidence of absence of habitat. Road density is a proxy for disturbance pressure, not a measurement of fragmentation.
An empty EONET result scores low but stays available — no relevant catalogued event is a real answer. It is never presented as 'no hazard risk'.
Confidence answers “how much evidence stands behind this score?” It is not a statistical confidence interval and says nothing about how likely the score is to be correct.