# Climate History AI Agent Connect

> Climate History MCP gives your AI agent direct access to the NOAA NCEI GHCN-Daily database. Instead of relying on weather models or forecasts, you get the actual daily records collected since the 1700s. Your agent can pull specific temperature, precipitation, and wind data for any location using station IDs, providing hard facts for historical climate analysis.

## Overview
- **Category:** weather-climate
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_oE0QLvwzJjJxDbgDO0cCq6qy3EibXOULvzehRN8C/ai-agent-connect
- **Tags:** climate, weather-history, historical-weather, ghcn, ncei, noaa, temperature, precipitation, weather-station, climate-data

## Description

You can stop guessing what the weather was like years ago. This MCP connects your AI client to the National Centers for Environmental Information records, providing raw data from over 130,000 stations worldwide. This isn't a predictive model or a smoothed-out forecast. It is the actual measurement recorded by a thermometer or rain gauge at a specific location. Whether you need to know if a specific summer was unusually hot in Lisbon or how much rain fell during a specific hurricane in Death Valley, your agent can pull the exact numbers. The data coverage is most dense in the United States, but it spans the globe. You can query specific days, entire months, or multi-year spans to find extremes or averages. It turns your AI from a conversationalist into a precise meteorological researcher.

## Tools

### find_weather_station
This tool finds a specific GHCN weather station by name or by latitude and longitude. It returns the station ID required for all other weather queries.

### get_daily_weather_history
This tool pulls day-by-day records for a specific station over a chosen date range. You get max, min, and average temperatures, plus precipitation, snowfall, and wind data.

### get_weather_summary
This tool provides an aggregated overview of weather for a station over a date range. It calculates mean and extreme temperatures, total precipitation, and identifies the wettest or coldest days.

## Prompt Examples

**Prompt:** 
```
What was the weather like in Lisbon in July 2003?
```

**Response:** 
```
The station nearest Lisbon is PO000008535 (Lisbon Geofisica). July 2003 there had a mean daily maximum of 29.4 °C and a mean minimum of 18.1 °C, with the hottest day reaching 36.2 °C on the 18th. Total precipitation was 1.8 mm across 2 wet days.
```

**Prompt:** 
```
Find the weather station for Central Park, New York.
```

**Response:** 
```
The station is USW00094728, NEW YORK CNTRL PARK TWR (US), located at 40.78, -73.97 with an elevation of 12.2 m. This is the longest continuous record in Central Park.
```

**Prompt:** 
```
How much rain did Death Valley get during Hurricane Hilary in August 2023?
```

**Response:** 
```
The station is USC00042319, DEATH VALLEY (US). It recorded 55.9 mm of rain on 2023-08-20, following 0.5 mm the previous day.
```

## Capabilities

### Station Locating
Your agent uses this to find the correct station ID via name or coordinates.

### Daily Data Retrieval
Your agent pulls specific daily values for temperature, wind, and precipitation.

### Period Summarization
Your agent calculates totals and extremes for specific months or years.

### Global Coverage
Your agent accesses records from over 130,000 stations worldwide.

## Use Cases

### Extreme Weather Verification
Confirm the exact rainfall or temperature recorded during a specific historical storm event.

### Climate Trend Analysis
Compare mean temperatures across different decades for a specific geographic location.

### Agricultural Planning
Review historical precipitation and frost dates for specific regions.

### Site Selection
Analyze the historical climate profile of a location before building or investing.

## Benefits

- Accesses raw station measurements instead of interpolated models.
- Eliminates the need for manual API key management.
- Provides direct access to NOAA's massive GHCN-Daily dataset.
- Works with any MCP-compatible client like Claude or Cursor.

## How It Works

Connect your client to Vinkius and start querying NOAA data immediately.

1. Connect your AI client to the Climate History MCP via Vinkius.
2. Ask your agent to find a weather station by name or location.
3. Use the station ID to request daily history or a summary.
4. Receive the raw, verified weather data directly in your chat interface.

## Frequently Asked Questions

**Is this data a forecast?**
No. This MCP provides historical records of what was actually measured by weather stations, not predictions of what might happen.

**Do I need an API key for NOAA?**
No. Vinkius hosts the MCP and manages the connection, so you can use it without your own credentials.

**How many stations are available?**
The MCP accesses data from over 130,000 stations worldwide.

**What kind of weather data can I get?**
You can retrieve maximum, minimum, and average temperatures, precipitation, snowfall, and wind data.

**Can I get data for any location?**
You can search for stations by name or by latitude and longitude. While coverage is densest in the US, stations exist in almost every country.

**Do I need an API key?**
No. This reads NOAA's public GHCN-Daily dataset, which requires no key or account.

**How far back does the data go?**
It depends on the station. The oldest records reach the 1700s; many US stations start in the late 1800s; stations elsewhere often begin in the mid-1900s. If a range returns nothing, the station simply does not cover those dates.

**Is this a forecast?**
No. GHCN-Daily is the observed past — actual thermometer and rain-gauge readings. It says nothing about the future.

**Why does a day show a dash for some values?**
The station did not report that element that day. Some stations measure temperature but never precipitation, and individual days are missing in every long record.