# nve-mcp [Health: Active]

**Category:** 🌳 Environment & Nature  
**Repository:** https://github.com/3121n/nor-data-nve-mcp  
**GitHub Stars:** 0  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/nve-mcp

## Description
Norwegian flood & landslide hazard data from NVE, by WGS84 coordinate.

## Claude Desktop Quick Installation
Heuristic fallback — verify the package name and runner against the repository README before running it. Uses `npx` (confidence: low):

```json
"mcpServers": {
  "nve-mcp": {
    "command": "npx",
    "args": ["-y","nve-mcp"]
  }
}
```

## Documentation

## What nve-mcp MCP server does

The nve-mcp MCP server provides point-based natural-hazard lookups for Norway. It wraps NVE’s open ArcGIS REST map services and returns structured information for a latitude and longitude supplied in WGS84. The server is intended for applications that need flood and landslide context for a specific location, including address or property-oriented analysis.

The flood lookup covers the nationwide flood susceptibility layer, checks whether the point lies within an area covered by that analysis, and consults the detailed 200-year flood-zone layer where a detailed flood map exists. The landslide lookup checks snow and rockslide susceptibility, debris-flow and landslide susceptibility, and quick-clay hazard zones.

## How it works

`hent_flomaktsomhet(lat, lon)` performs the flood-related queries. Its result can contain a susceptibility-area match, a coverage status, and information about a detailed 200-year flood zone or the mapped area in which that layer is available.

`hent_skredaktsomhet(lat, lon)` combines several landslide-related checks. It queries snow and rockslide susceptibility, soil and debris-flow susceptibility, and quick-clay hazard zones. Returned landslide-type codes are translated for both short NVE/NGU code ranges and the longer code ranges used by those data domains.

A key interpretation rule is that susceptibility is not the same as a formally defined hazard zone. The nve-mcp MCP server reports coverage layers so that an empty result can be interpreted correctly. A point with no matching feature outside an analyzed area should be treated as not mapped, rather than as evidence that the location is safe. Each response also includes a service URL, a thematic map link, and licensing information for the source data.

## Setup and configuration

The repository provides a local stdio MCP server. Install its Node.js dependencies, compile the project, and start the generated entry point:

```bash
npm install
npm run build
node dist/index.js
```

The README also shows an end-to-end test command:

```bash
node test-e2e.mjs
```

A Claude configuration example registers the server under the name `nve` in `~/.claude.json`. No environment variables or credential settings are documented. The project is licensed under MIT, while the underlying NVE and NGU data is identified as open data under NLOD or CC BY.

## Tools and capabilities

The documented tools are:

- `hent_flomaktsomhet(lat, lon)`: Queries flood susceptibility, analysis coverage, and available detailed flood-zone information.
- `hent_skredaktsomhet(lat, lon)`: Queries multiple landslide susceptibility and quick-clay hazard layers, including coverage checks.

The nve-mcp MCP server is useful for enriching a location record with separate flood and landslide fields. It is less suitable as a standalone safety determination because the data distinguishes susceptibility, mapped coverage, and detailed hazard-zone information rather than providing a single universal risk verdict.

## Limitations and notes

The server is limited to the NVE and NGU layers exposed by the documented ArcGIS services. Results depend on whether the supplied point falls within a layer’s analyzed or mapped coverage. An absence of a matching feature does not necessarily indicate an absence of natural hazard. Consumers should preserve the returned coverage information and present the source links when displaying results.

The README documents verification points in Lillestrøm, Oslo, Sørreisa, and Gjerdrum, but those examples do not establish coverage for every Norwegian location or guarantee that the upstream datasets remain unchanged.

_Full upstream README: https://allmcps.com/mcp/nve-mcp/readme_