PixelGust Environmental Data MCP Server

Monthly time series 2015-2025 (about 130 monthly values) for one variable at a point, plus the 12-month climatology. Accumulation variables are monthly totals: precipitation_mm and total_evaporation_mm in mm/month (evaporation positive), snowfall_cm in cm/month, net_solar_radiation_mj_m2 and net_thermal_radiation_mj_m2 in MJ/m2/month. Net thermal is negative when the surface loses heat. Sum 12 months for an annual total. Other variables are monthly means (temperatures in C, pressure in hPa, wind speed in m/s, humidity and cloud cover in %, soil water as volume fraction, ndvi unitless, fire_weather_index as FWI). The response carries a unit string. Use for trend questions (is it getting hotter or drier here?). Wind direction is not available as a time series.

Current weather conditions for a point: temperature, humidity, dewpoint, wind speed and direction, precipitation, pressure, cloud cover, solar radiation, plus terrain, land cover, NDVI, population density, and soil context. Source: NOAA GFS 0.25 degree, updated every 6 hours. The response's units object gives each value's unit and time basis.

Environmental data for a point: NDVI vegetation index (300 m), annual evapotranspiration (MODIS, mm/year), and ESA WorldCover land cover class.

Natural hazard indices for a point: Fire Weather Index (current danger), soil erosion risk (RUSLE, t/ha/yr), and flood susceptibility (Topographic Wetness Index).

10-year historical climate averages (2015-2024, ERA5 reanalysis) for a point. Pass month 1-12 for that month's climatology with min and max ranges: accumulations (precipitation, snowfall, evaporation, radiation) are then totals for that month, for example precipitation_mm in mm/month. Omit month for the annual climate: accumulations are then annual totals (mm/year, MJ/m2/year). Evaporation is positive (water evaporated). Use this for questions about typical climate, growing seasons, or what weather to expect at a location and time of year. The response's units object gives each value's unit and time basis.

Aggregate statistics (mean, min, max, std) over a polygon area: weather, terrain, hazards, environment, soil, and population. Coordinates are [longitude, latitude] pairs forming a simple ring (no self-intersections). Maximum area depends on plan (free tier: 100 km2, measured on the sphere). Weather units follow the chosen mode, as in get_current_weather, get_weather_forecast and get_historical_climate. Use for area questions such as average slope of a farm or population inside a zone.

Straight-line distance in km from a point to the nearest feature of six layers: coastline, highway (motorway or trunk road), airport (large or medium), major port, hospital or clinic, and school. Each layer is returned with its display name and distance_km.

Soil properties at a point from SoilGrids 2.0 (ISRIC - World Soil Information, CC BY 4.0), 250 m mean predictions: clay, sand and silt (%), bulk density, coarse fragments, organic carbon (content, density and 0-30 cm stock), total nitrogen, cation exchange capacity and pH, at six depths from 0-5 cm to 100-200 cm. Built-up areas, water and permanent ice have no prediction, so the nearest mapped pixel within a few kilometres is used and reported in pixel. Model estimates, not measurements: ISRIC advises against using SoilGrids at the local or farm level, and they are not for engineering design. Units are in soil.units, and the source block holds the credit to pass on with the values.

Terrain for a point: elevation (m), slope and aspect (degrees) at 30 m resolution (Copernicus DEM), plus Topographic Wetness Index (flood susceptibility indicator).

Daily weather forecast for a point. day is 0-based: day 0 is the first forecast day, the day after the GFS run date (UTC), not today. forecast_dates in the response lists all available days (currently 6 or 7 depending on the model run). The response gives the calendar date of the day returned, and daily totals or mean, min, and max per variable. The response's units object gives each value's unit and time basis.