Skip to content

Ecological Monitoring Demo

Prompt

User asked for the repository to be repositioned as a hands-on training sandbox where environmental data science researchers can practice quantum-ready workflows using familiar biological and geospatial decision problems.

Datasets

This workflow uses a synthetic ecological monitoring candidate-site table by default. The generated table includes:

  • site_id
  • lat
  • lon
  • species_richness
  • climate_refugia_score
  • habitat_connectivity
  • cost
  • region
  • mean_temp
  • annual_precip
  • elevation

The synthetic table is a teaching stand-in for a real table produced from harmonized environmental layers, geospatial data cubes, field observations, or working-group datasets.

Method

The workflow selects 12 monitoring sites from 100 candidates. It builds a QUBO-style binary optimization model where each variable represents whether a site is selected. The model rewards biological value and environmental coverage, then penalizes cost, redundancy, and missing the requested number of selected sites.

The default solver runs locally. It uses dwave-neal simulated annealing when available and falls back to a small randomized local-search emulator otherwise. No D-Wave cloud account is required.

Result

The workflow writes selected sites from the quantum-inspired emulator and a greedy classical baseline, plus a comparison table and map:

  • selected_sites_quantum_emulator.csv
  • selected_sites_quantum_emulator.geojson
  • selected_sites_greedy_baseline.csv
  • selected_sites_greedy_baseline.geojson
  • site_selection_comparison.csv
  • harmonized_visualization.png

This result is for practice and interpretation. It does not claim quantum advantage.