Skip to content

Add AI Assistant to MapViewer for Multi-Sectoral Geospatial Risk Analysis #511

@marta-baraibar

Description

@marta-baraibar

Summary
Add a conversational AI assistant directly within the MapViewer interface, enabling users to query geospatial data in natural language — for example, "analyse risk of flooding in the northern region" — and receive contextual, actionable insights without needing deep GIS expertise.

Motivation
ClimWeb serves a wide range of users — from meteorological professionals to policymakers and field officers — who need to interpret complex spatial datasets to make real-world decisions. Currently, extracting meaningful insight from the MapViewer requires technical knowledge of the data layers and their relationships. This creates a barrier to effective use, particularly in multi-sectoral scenarios involving overlapping climate, land use, population, and hazard data.
Projects like Chat2Geo demonstrate that LLM-powered geospatial assistants can meaningfully lower this barrier — turning complex layer combinations into structured, human-readable analysis outputs.

Proposed Feature
Integrate an AI assistant panel or prompt interface into the MapViewer that:

Accepts natural language queries tied to visible or selectable map layers (e.g. "analyse risk of drought for agricultural zones")
Interprets multi-sectoral context — combining climate, hazard, land use, and demographic layers
Returns structured outputs: summaries, risk ratings, and recommended actions
Keeps the experience focused and in-context (no need to leave the map view)
Supports decisions at national and sub-national levels for NMHSs and partner agencies

Expected Outcomes

Users can derive actionable intelligence from geospatial data without GIS expertise
Multi-sectoral risk analysis becomes faster and more accessible
Outputs are easier to communicate to non-technical stakeholders
Geospatial intelligence becomes more practical, precise, and impactful across sectors

References

Chat2Geo — example of LLM-assisted geospatial analysis
Related issue: MapViewer current feature set

Metadata

Metadata

Assignees

No one assigned

    Labels

    climwebclimtech specific issues

    Projects

    Status

    Mapviewer Backlog

    Milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions