WeatherXM FACT — Forecast Accuracy & Ground-Truth Verification

com.weatherxm.factPublished by fact.weatherxm.comOfficial publisherv2.1.0Updated Nov 5, 2026

Scores 45 NWP & AI weather models (ECMWF, GFS, ICON, meteoblue, MOSAIC) against 8,927 stations.

VerifiedStreamable HTTPWeb executableLocation & LifestyleData & Analytics

Overview

AI-generated overview

WeatherXM FACT server that lets an assistant compare 45 weather forecast models against historical station data by location, variable and lead day.

What you can do
  • Find the forecast model with the lowest verified historical error for a coordinate or station, a weather variable and a lead day 0-7
  • Check historical ground-station publication coverage for a coordinate cell or station, with nearest evaluated cell distance
  • Read the global accuracy leaderboard or country-level cell-win summaries across all models and stations
  • Compare up to 10 models head-to-head on global historical error curves across lead days 0-7
When to use it
Useful when choosing or justifying a weather forecast model for a specific site, or when checking whether historical station coverage exists for a location such as a farm or solar/wind asset. Also suited to comparing model accuracy rankings before relying on a forecast source.
Requirements
Remote streamable HTTP endpoint at fact.weatherxm.com; no packages, environment variables or headers declared. No authentication is declared, so network access to the endpoint is the only stated requirement.
Before you install
Results are historical archive data only, not live forecasts or guarantees of future accuracy. Publication gaps do not mean no stations exist, and historical MAE should not be read as a performance guarantee; cohort and station-commissioned (mLM) caveats should be preserved. Read-only lookups, with no writes or credentials declared.

Installation

In SourceWeft

  1. Open WeatherXM FACT — Forecast Accuracy & Ground-Truth Verification in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/com.weatherxm.fact%2Fmcp/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config.

{
  "mcpServers": {
    "weatherxm-fact-forecast-accuracy-ground": {
      "type": "http",
      "url": "https://fact.weatherxm.com/mcp"
    }
  }
}

Tools

4

get_best_forecast_model

Find which weather forecast model had the lowest verified historical error (MAE) for a latitude/longitude coordinate (or station_or_report_id), weather variable (temperature, wind_speed, precipitation, humidity), and forecast horizon (Days 0-7). All 45 FACT models: WeatherXM MOSAIC (MOSAIC), meteoblue (METEOBLUE), meteoblue station-mLM (mLM), ECMWF IFS (IFS025, IFSHRES), ECMWF AIFS AI (AIFS025), NOAA GFS (GFS05, GFS025), NOAA HRRR (HRRR), NOAA National Blend of Models (NBM), NOAA NAM (NAM3, NAM5, NAM12, NAMAK), DWD ICON (ICON, ICONEU, ICOND2), UK Met Office (UMGLOBAL10, UKMO2), ECCC GEM (GEM15, GEM2), Meteo-France ARPEGE/AROME (MFGLOBAL, MFEU, MFFR), COSMO (COSMOI2, COSMOM5), HARMONIE-AROME (HARMONIE), JMA MSM (MSM), WRF (WRFGR, WRFAMS), meteoblue NEMS/NMM (NEMSGLOBAL, NEMSGLOBAL_E, NEMS4, NEMS12, NEMS12_E, NEMSAF01, NEMSAS02, NEMSCAM, NEMSIN, NEMSNZ, NEMSSA, NMM4, NMM22), and regional models (AF01, AS01, SA01). Useful for solar & wind energy forecasting, agriculture & frost/irrigation risk, maritime/port & drone operations, construction planning, and microclimate evaluation. Provide (lat, lon) or station_or_report_id. Missing results stay unknown; not live forecasting.

Parameterslatlonlead_dayvariableexclude_commissionedstation_or_report_id

check_ground_truth_coverage

Check historical WeatherXM FACT ground-station publication coverage for any coordinate (H3 Res-7 cell, ~5.2 km²) or station_or_report_id, plus nearest evaluated reference cell distance and interactive map citation link. Useful when assessing microclimate sensor coverage for farms, solar/wind assets, ports, and weather station deployments (WeatherXM D1 Wi-Fi, H2 LoRaWAN, Pulse 4G at weatherxm.com/stations). Provide (lat, lon) or station_or_report_id. Does not establish current live station presence or absence.

Parameterslatlonlead_dayvariableexclude_commissionedstation_or_report_id

get_fact_leaderboard

Read the global weather forecast accuracy leaderboard (MAE, RMSE, and categorical rain skill HSS, POD, FAR, CSI across lead days 0..7) or country-level published H3 cell-win summaries across all 45 forecast models and 8,927 stations. All 45 FACT models: WeatherXM MOSAIC (MOSAIC), meteoblue (METEOBLUE), meteoblue station-mLM (mLM), ECMWF IFS (IFS025, IFSHRES), ECMWF AIFS AI (AIFS025), NOAA GFS (GFS05, GFS025), NOAA HRRR (HRRR), NOAA National Blend of Models (NBM), NOAA NAM (NAM3, NAM5, NAM12, NAMAK), DWD ICON (ICON, ICONEU, ICOND2), UK Met Office (UMGLOBAL10, UKMO2), ECCC GEM (GEM15, GEM2), Meteo-France ARPEGE/AROME (MFGLOBAL, MFEU, MFFR), COSMO (COSMOI2, COSMOM5), HARMONIE-AROME (HARMONIE), JMA MSM (MSM), WRF (WRFGR, WRFAMS), meteoblue NEMS/NMM (NEMSGLOBAL, NEMSGLOBAL_E, NEMS4, NEMS12, NEMS12_E, NEMSAF01, NEMSAS02, NEMSCAM, NEMSIN, NEMSNZ, NEMSSA, NMM4, NMM22), and regional models (AF01, AS01, SA01). Country counts describe published cell wins, not national matched-sample accuracy.

Parameterscountrylead_dayvariableexclude_commissioned

get_model_head_to_head

Compare up to 10 weather forecast models head-to-head across lead days 0 to 7 on global historical error curves, with explicit cohort and station-commissioned (mLM) caveats. All 45 FACT models: WeatherXM MOSAIC (MOSAIC), meteoblue (METEOBLUE), meteoblue station-mLM (mLM), ECMWF IFS (IFS025, IFSHRES), ECMWF AIFS AI (AIFS025), NOAA GFS (GFS05, GFS025), NOAA HRRR (HRRR), NOAA National Blend of Models (NBM), NOAA NAM (NAM3, NAM5, NAM12, NAMAK), DWD ICON (ICON, ICONEU, ICOND2), UK Met Office (UMGLOBAL10, UKMO2), ECCC GEM (GEM15, GEM2), Meteo-France ARPEGE/AROME (MFGLOBAL, MFEU, MFFR), COSMO (COSMOI2, COSMOM5), HARMONIE-AROME (HARMONIE), JMA MSM (MSM), WRF (WRFGR, WRFAMS), meteoblue NEMS/NMM (NEMSGLOBAL, NEMSGLOBAL_E, NEMS4, NEMS12, NEMS12_E, NEMSAF01, NEMSAS02, NEMSCAM, NEMSIN, NEMSNZ, NEMSSA, NMM4, NMM22), and regional models (AF01, AS01, SA01).

Parametersmodelsvariable

Version history

1
  1. v2.1.0LatestNov 5, 2026

Score

83% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listedRequired15 pts
  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts

    Declare a GitHub repository with a README, or serve resource_documentation in the OAuth protected-resource metadata.

  • LicenseNot applicable

    A hosted service is used under its provider's terms, so no license is expected here. This check is left out of the percentage.