COA Connectivity Lab — data-management (active here)

A modular, reproducible workflow for modelling ecological connectivity, integrating database management, machine learning, GIS visualisation, and biodiversity data ingestion.

Overview

The Eco Connectivity Workflow is a modular GitHub-based research stack designed to support scalable, reproducible ecological connectivity modelling.

It enables multiple teams to work in parallel while maintaining full provenance tracking, reproducibility, and interoperability across all stages of the pipeline.

The system transforms raw biodiversity and environmental data into:

• connectivity maps
• resistance surfaces
• spatial conservation products

---

The Modular Stack

Module	Purpose	Repository
Database	PostgreSQL/PostGIS schema and workflow views	https://github.com/coa-connectivity-lab/db-schema
Machine Learning	Resistance modelling and connectivity analysis	https://github.com/coa-connectivity-lab/ml-models
GIS Projects	QGIS projects, styles, and cartographic outputs	https://github.com/coa-connectivity-lab/qgis-projects
Data Management	GBIF ingestion, raster preprocessing, and species harmonisation	https://github.com/coa-connectivity-lab/data-management
Reproducibility Logs	Scenario tracking, outputs, and execution provenance	https://github.com/coa-connectivity-lab/reproducibility-logs

---

System Architecture

flowchart TD
    A[Data Team] -->|Ingest GBIF, rasters, field data| B[Database]
    B -->|PostGIS schema + spatial views| C[ML Models]
    C -->|Resistance & connectivity outputs| B
    C --> D[GIS/QGIS]
    D -->|Maps & spatial products| E[Figures & Publications]
    A --> F[Reproducibility Logs]
    B --> F
    C --> F
    D --> F

Loading

---

Reproducibility

This project is now fully reproducible using:

• Julia 1.10+
• Pkg environment manager
• System-installed GDAL/PROJ libraries
• Ubuntu CA certificate store

No external environment managers (such as Guix) are required.

---

Setup (Julia environment)

cd omniscape-stack
julia setup.jl

Then run:

julia run_omniscape.jl

Expected output:

✔ Omniscape environment loaded successfully

---

Notes

• Ensure system packages are installed:

sudo apt install gdal-bin libgdal-dev proj-bin libproj-dev

• Ensure SSL certificates are available:

/etc/ssl/certs/ca-certificates.crt

---

Summary

This repository now follows a standard, system-native Julia workflow with no dependency on Guix or external environment managers, improving reproducibility, portability, and long-term maintainability across research systems.