mud2dust publishes a SpatioTemporal Asset Catalog (STAC) of every layer the platform produces. Today the catalog has 10 collections registered in pgstac on Aurora. The public STAC API server (stac-fastapi at https://stac.mud2dust.com) lands in P3.2 iter 2 — until then this page is a static rendering of the same JSON the API will serve.
mud2dust-clay-0-5Surface clay content (% by mass) from ISRIC SoilGrids 2.0. Static prior used as a retention-floor feature in the fusion model.
CC-BY-SA-4.0mud2dust-demSurface elevation in meters from ESA's GLO-30 global DEM. Static prior — recomputed on schema upgrade.
CC-BY-4.0mud2dust-lstBrightness temperature in °C from HLSL30 v2.0 B10 (Landsat 8/9 TIRS), atmospherically corrected.
mud2dust-n-stress-proxyPer-pixel classification raster — directional nitrogen-stress proxy. Pixel codes (uint8): 0=inconclusive, 1=canopy_too_sparse, 2=water_stress_first, 3=heat_stress_first, 4=healthy, 5=likely_n_stressed. Default-deny decision tree fuses NDRE + NDVI + NDWI + calibrated VWC + LST to flag the pixels worth walking for N after ruling out water / heat / sparse-canopy drivers. Honest about being a proxy, not a tissue test — tightens with contributor lab-N + tissue-test + fertilizer-event contributions.
CC-BY-4.0mud2dust-ndreNormalized Difference Red-Edge index from HLSS30 v2.0 — saturates later than NDVI at high LAI.
mud2dust-ndviNormalized Difference Vegetation Index from HLSS30 v2.0 surface reflectance.
CC-BY-4.0mud2dust-ndwiNormalized Difference Water Index (McFeeters) from HLSS30 v2.0.
mud2dust-precip-1h1-hour precipitation accumulation from NOAA HRRR forecast model. v0 — full 7-day rolling lands with Lambda automation.
mud2dust-sigma0C-band SAR backscatter, terrain-corrected gamma-nought from NASA OPERA RTC-S1, 30 m. Per-region per-day mosaic.
CC-BY-4.0mud2dust-vwc-rootzoneVolumetric water content (m³/m³), per-pixel blend of σ⁰ + NDVI + clay% via mud2dust v0 fusion. Literature-prior placeholder until ground-truth calibration data from partner-farm anchor stations lands.
CC-BY-4.0