Get climate answers you can trust.
A free, keyless, open-source MCP server for official climate science, IPCC, IPBES and IPOS reports, French DRIAS projections, and global IPCC Atlas data. Exposed through the Model Context Protocol so any compatible assistant can query real data instead of guessing.
You do not need to learn tool names or JSON. Connect this server to your assistant, then talk naturally. It picks the right data source and answers in plain language, with citations or numbers you can verify.
“What are the main drivers of sea level rise according to IPCC reports?”
Your assistant calls retrieve_data_mcp and returns
relevant passages and figures from IPCC / IPBES / IPOS reports, each
with its source and page so you can check the original.
“What will the temperature be like in Paris around 2030?”
Your assistant calls query_drias and returns climate
projection data from the official DRIAS dataset for France —
temperatures, rainfall or extreme events for a city, region or the
whole country.
“How will total rainfall in the USA change by 2050 under different IPCC scenarios?”
Your assistant calls query_ipcc and returns IPCC Atlas
data with historical trends and SSP scenario projections, scoped to
any country or city you name.
Modern assistants are good at reasoning and synthesis, but only when they are grounded in trustworthy inputs. Climate questions deserve better than vague guesses: they need peer-reviewed report excerpts, traceable figures, and curated climate projection tables.
Not ready to connect an assistant? Pick a tool below or switch to Try the tools and run example questions live.
Pick your client below. You only need the endpoint URL — no signup, no API key for this public instance.
ClimateQ&A.https://ekimetrics-climateqa-mcp.hf.space/mcpClimateQ&A · Auth: None.https://ekimetrics-climateqa-mcp.hf.space/mcpRequires ChatGPT Pro, Business, or Enterprise (custom connectors).
ClimateQ&A · Authentication: No authentication.https://ekimetrics-climateqa-mcp.hf.space/mcp.cursor/mcp.json in your project or home directory.climateqa · paste this URL:https://ekimetrics-climateqa-mcp.hf.space/mcp{
"mcpServers": {
"climateqa": {
"url": "https://ekimetrics-climateqa-mcp.hf.space/mcp"
}
}
}
Connect via Streamable HTTP to
https://ekimetrics-climateqa-mcp.hf.space/mcp.
Transport: MCP 2025-03-26. No authentication required on this public instance.
Tool calls on this public server are recorded as JSON files in the
climateqa-mcp-logs
dataset (read access is public). Browse the
mcp/
folder to see which tools were called, with what input, and what was returned.
To find your calls, look at the client_info field
in each log file. It carries the name and version your
MCP client sends when it connects (for example "name": "cursor" for
Cursor IDE).
If you connect with a custom client, please set
clientInfo to something meaningful — a clear
name (team, project, or bot) and a version
— so your usage is easy to identify.
Run real queries against this server. Results appear below each form as readable cards and tables. Expand View raw JSON response when you need the full payload for debugging or integration work.
retrieve_data_mcpSearch official IPCC, IPBES and IPOS reports. Returns the most relevant text passages and figures, each tagged with its source so answers stay traceable.
| Field | Description |
|---|---|
query | Your question in plain language. |
{
"documents": [
{
"content": "Sea level rise is driven by thermal expansion of ocean water and the melting of glaciers and ice sheets…",
"metadata": {
"source": "IPCC AR6 WGI Chapter 9",
"page": 234,
"doc_id": "ipcc_ar6_wg1_ch9"
}
}
],
"images": [
{
"content": "Figure 9.27: Global mean sea-level change relative to 1900…",
"metadata": {
"source": "IPCC AR6 WGI",
"page": 1289
}
}
]
}
query_driasQuery France's official DRIAS climate projections — temperature, rainfall, heatwaves and other indicators for a city, region or the whole country.
| Field | Description |
|---|---|
query | Your question. Mention a place in France and optionally a year. |
full_rows | Optional. Return the full data grid for country-wide snapshot questions (default: summary + sample). |
{
"query": "What will the temperature be like in Paris?",
"error": null,
"results": [
{
"status": "OK",
"table": "mean_annual_temperature",
"indicator_column": "mean_annual_temperature",
"unit": "°C",
"row_count": 102,
"total_row_count": 102,
"truncated": false,
"summary": null,
"rows": [
{"year": 2030, "mean_annual_temperature": 13.5},
{"year": 2050, "mean_annual_temperature": 14.2}
],
"location": {
"location": "Paris, France",
"latitude": 48.85,
"longitude": 2.35,
"country_code": "FR"
},
"year": null,
"sql_query": "SELECT year, AVG(mean_annual_temperature) …",
"data_information": "Yearly values of 'mean_annual_temperature' (°C) at Paris…"
}
]
}
query_ipccQuery the IPCC Atlas for any country or city — historical climate data and future scenarios (SSP pathways) for temperature and precipitation.
| Field | Description |
|---|---|
query | Your question. Always mention a place; optionally a year and/or month. |
full_rows | Optional. Return the full data grid for country-wide snapshot questions (default: summary + sample). |
{
"query": "What will be the total rainfall in the USA in 2030?",
"error": null,
"results": [
{
"status": "OK",
"table": "total_precipitation",
"indicator_column": "total_precipitation",
"unit": "mm/day",
"row_count": 10,
"total_row_count": 12345,
"truncated": true,
"summary": [
{"scenario": "historical", "n": 2469, "mean": 2.51, "min": 0.4, "max": 8.9, "stddev": 1.1},
{"scenario": "ssp245", "n": 2469, "mean": 3.08, "min": 0.6, "max": 10.2, "stddev": 1.3},
{"scenario": "ssp585", "n": 2469, "mean": 3.21, "min": 0.5, "max": 11.4, "stddev": 1.4}
],
"rows": [
{"latitude": 25.5, "longitude": -80.5, "scenario": "ssp245", "total_precipitation": 4.12}
],
"location": {"location": "USA", "country_code": "US"},
"year": "2030",
"sql_query": "SELECT latitude, longitude, scenario, total_precipitation …"
}
]
}