What it is. An experimental 0–100 summary of whether recent conditions look favourable for grassland fungi in general. It is not a probability and does not predict any species.

Inputs (model conditions-v0.5). Recent rainfall (25%): 45 mm in 14 days scores full marks. Modelled soil moisture at 3–27 cm (20%): 15% volumetric scores 0, 40% scores full marks. Water balance (15%): a 50 mm soil bucket topped up by rain and drained by FAO-56 reference evapotranspiration. Cool nights (20%): the 7-night mean low, best at 8°C, losing 12 points per degree away. Terrain (10%): elevation, slope and aspect from a 90 m elevation model. Habitat (10%): OpenStreetMap land-use mapping, refined by official habitat data — NRW LANDMAP in Wales, Natural England’s Priority Habitat Inventory, the Crop Map of England and Living England in England, NatureScot’s Habitat Map of Scotland — which separates improved (fertilised, reseeded) grassland (scores 30) and arable (10) from semi-natural grassland (100) — or what you set in the field. Each frost night in the last 7 days subtracts 5 points (up to 15).

Land & grazing (Great Britain). Each point is checked against official open habitat data for its nation. Wales: NRW LANDMAP records how each landscape area is managed (stock grazing, mowing, cultivation…), plus NRW’s potential habitat for grassland fungi and open-access land. England: Natural England’s Priority Habitat Inventory (mapped semi-natural habitat), the Crop Map of England 2023 (grass or crop, field by field), Living England (satellite map, used only to flag improved grassland) and registered common land. Scotland: NatureScot’s Habitat Map of Scotland (survey-based; patchy coverage). Only Wales publishes grazing — elsewhere record it in Field notes. Grazing is shown and stored with every find and blank but is not part of the score; Evidence on the History tab tests whether it separates your finds from your blanks. Survey data is often years old and landscape-scale.

Missing data. If an input is unavailable its weight is shared across the others, and the card shows how complete the inputs were. Nothing is filled in with guessed values.

Resolution — read this. Weather comes from models with grid squares a few kilometres across (the historical archive is coarser, roughly 10–25 km). Spots inside the same grid square get identical weather, so differences between nearby spots come only from terrain and habitat. Elevation is 90 m resolution; small banks and hollows are invisible. OpenStreetMap habitat mapping is patchy in places.

How to trust it. Every threshold above is a starting guess. Log finds and blanks, run Evidence on the History tab, and use the separation score (AUC) to see whether the index actually distinguishes them. Around 0.5 means it doesn't; above 0.7 is useful. The thresholds should be re-set from your records once there are enough of them.

Sources. Open-Meteo forecast, archive (ERA5) and elevation APIs; Copernicus DEM; OpenStreetMap via Overpass; Natural Resources Wales LANDMAP, grassland-fungi potential habitat and access-land layers via DataMapWales (contains Natural Resources Wales information © Natural Resources Wales and database right); Natural England Priority Habitat Inventory, Living England and registered common land, and the Rural Payments Agency Crop Map of England (© Natural England / RPA); NatureScot Habitat Map of Scotland (© NatureScot) — all under the Open Government Licence; postcodes.io and Nominatim for search.