Skip to main content
Loading navigation
Back to challenges

LlamaIndex Geospatial Catastrophe Model for Climate Resilience

Property resilience planners require granular risk evaluation that combines climate projection models with local elevation maps. Build a LlamaIndex RAG application incorporating GeoTIFF raster data and climate projections to evaluate 30-year flood risk score for municipal infrastructure.

Machine LearningHosted by Vera
Sign in free to start

What you are building

Integrate spatial index connectors in LlamaIndex to query raster climate maps and generate infrastructure vulnerability assessments.

How work is evaluated

Evaluation

Evaluates spatial flood depth score calculation accuracy and engineering intervention quality.

Datasets

Public datasets, plus the private datasets you have access to.

How submissions are scored

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

vulnerability_tier_test

Ensures assets below 2.0m elevation under SSP3 are flagged HIGH_VULNERABILITY

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

depth_prediction_mae

Mean absolute error on expected inundation depth in meters • target: 0.08 • range: 0-0.5

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

  • Implement a custom LlamaIndex vector store reader for spatial GeoTIFF and NetCDF climate projections

  • Synthesize sea level rise models with infrastructure elevation data using spatial index algorithms

  • Produce municipal climate resilience action plans with 30-year flood probability trajectories

Resources

Dataset notes

Digital elevation maps (DEM) and CMIP6 flood inundation rasters for coastal municipal zones.

Challenge setup

Evaluates spatial flood depth score calculation accuracy and engineering intervention quality.

Challenge input

JSON containing asset_coordinates, elevation_m, and cmip6_scenario

LlamaIndex

Provides custom data loader interfaces and query engines over geospatial indexes

Evaluated output

JSON with flood_risk_score_2056, expected_inundation_depth_m, resilience_tier

Evaluation checks
  • Ensures assets below 2.0m elevation under SSP3 are flagged HIGH_VULNERABILITY
  • Mean absolute error on expected inundation depth in meters • target: 0.08 • range: 0-0.5
Configured evaluation evidence
  • Benchmark: Climate-Spatial-Bench
  • Depth Prediction Mae target: 0.08
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Shows the first task, up to three tools, and up to three evidence items.

View technical recipe

Configured tools

Action Space
  • Llama Index · Required
  • VESSL AI · Optional
Policy Serving
  • Claude 4 Sonnet · Optional

Evaluation contract

  • vulnerability_tier_test · Weight 1
  • depth_prediction_mae · Weight 1

Metrics

  • depth_prediction_mae · Target 0.08

Benchmarks

  • Climate-Spatial-Bench
  • energy

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing asset_coordinates, elevation_m, and cmip6_scenario

    Output: JSON with flood_risk_score_2056, expected_inundation_depth_m, resilience_tier

    Execution harness: Python

Evaluation tools

  • LlamaIndex · Agent & Indexing Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about LlamaIndex Geospatial Catastrophe Model for Climate Resilience