National Biodiversity Assessment System (NBAS): Local Standalone Version

This repository provides a local standalone version of the National Biodiversity Assessment System (NBAS) used in PLANR. It enables users to run NBAS assessments independently of the PLANR website and web service.

The NBAS was developed as part of the Ecological Knowledge System (EKS). The EKS is a partnership between the Commonwealth Scientific and Industrial Research Organisation (CSIRO) and the Department of Climate Change, Energy, the Environment and Water (DCCEEW) to establish a transparent and authoritative source of information, biodiversity assessment and forecast capability for the Nature Repair Market.

The NBAS is a software and data package that provides a nationally consistent approach to assessing an area’s current contribution to biodiversity and forecasting the expected contribution to biodiversity following successful implementation of a given project within the Nature Repair Market.

The standalone NBAS (version 1) produces the following metrics:

• Connectivity scores
• Conservation significance scores
• Contribution to biodiversity persistence scores

Data inputs required to run the standalone version include:

• Ecosystem type identification number (The ecosystem identification key for each ecosystem type is available from the NBAS data link below)
• Starting and target ecosystem condition state reference number; and starting and forecast ecosystem condition scores (Please refer to the [NBAS settings] (https://www.dcceew.gov.au/environment/environmental-markets/nature-repair-market/incorporated-documents-and-resources#toc_7) for use with the Replanting native forest and woodland ecosystems method
• An example payload is provided here.*

For detailed documentation of NBAS methods, data sources, and outputs, please refer to:

• The Ecological Knowledge System for the Nature Repair Market. Technical report
• Ecological Knowledge System: National Biodiversity Assessment System (NBAS) Data
• Ecological Knowledge System: State and Transition Models

When to Use the Standalone NBAS Version

This standalone version of NBAS is intended for advanced users who need to run assessments on a local computer, at scale, or as part of automated workflows. It assumes familiarity with command-line tools, Docker, and basic data handling. If you're only assessing a small number of locations, or prefer a guided interface, we recommend using the PLANR website, which provides access to NBAS with examples and visual outputs.

This repository does not include detailed examples or tutorials. It is designed for users who:

• Need to run NBAS offline, at over large areas, or in secure environments
• Want to integrate NBAS into larger data pipelines
• Are comfortable setting up and running Docker containers

Computational Requirements and Software

Minimum recommended specifications:

• OS: Linux, macOS, or Windows 11 (with WSL(Windows Subsystem for Linux))
• CUDA enabled GPU with compute capability > 8
• NVIDIA container toolkit
• Docker

Setup Instructions for Linux and Windows

Linux/macOS

1. Install Docker

2. Download (pull) docker image

`sudo docker image pull haizeaanalytics/nbas:latest`

3. Run docker image and mount the home directory

`sudo docker run -ti --gpus all -v $HOME:/data haizeaanalytics/nbas:latest /bin/bash`

Note: This opens an interactive session inside the container with your home directory mounted in /data. Please ensure your payload is available in this directory. An example payload can be found here.

4. Run NBAS in the docker container

python3.11 main.py /data/path/to/payload.json

5. Copy result to home directory

cp result/xxxx-result.json /data/path/to/result.json

Windows 11

Tested successfully with:

• Intel i7-13700H, 64 GB RAM
• Nvidia RTX 4060 GPU
• CUDA Toolkit + Visual Studio (with Python, Node.js workloads)
• Docker Desktop for Windows
• Git Bash or PowerShell

1. Install Docker Desktop

2. Enable WSL and sudo: Run 'wsl --update' in Git Bash

3. Enable "sudo" in Settings > System > For Developers

4. Download (pull) Docker image

'sudo docker image pull haizeaanalytics/nbas:latest /bin/bash'

5. Create a folder on your local machine (e.g. C:\Users\your_username\nbas_payloads) and save your prepared input file (i.e. payload) in .json format to this folder (e.g. test_payload.json)

* An example payload can be found [here](https://raw.githubusercontent.com/chamith-ed/nbas-standalone/refs/heads/main/tests/test_payload.json).*  

6. Open Windows PowerShell and run container. For example:

'docker run -ti --gpus all -v C:\Users\your_username:/data haizeaanalytics/nbas:latest /bin/bash'

This command:
	- Starts the NBAS container
	- Mounts your local folder (C:\Users\your_username) into the container as /data
	- Opens a terminal inside the container
	
You will receive a prompt like:root@5adb621718c3:/app#
This means you're now inside the container and ready to run NBAS

7. Inside the container, run:

' root@5adb621718c3:/app#python3.11 main.py /data/nbas_payloads/test_payload.json'

8. NBAS will generate a results file. For example: Results written to results/2025-06-17T10:06:09-results.json

9. To move the results file from the container to your local folder, run:

'cp results/2025-06-17T10:06:09-results.json /data/nbas_payloads/result.json

Disclaimer

Results from this standalone desktop version of NBAS differ slightly from the NBAS service in PLANR due to slight differences in how areas are clipped from the full datasets.

Biodiversity benefit scores will be reviewed by the Clean Energy Regulator as part of the project registration process.

The NBAS code provides an estimate of biodiversity benefit only. We do not warrant the accuracy or completeness of any information contained in or calculated by the NBAS code, and users should use this information for estimation and guidance purposes only.

The NBAS code should not be relied on to confirm eligibility for any particular project, program, scheme or entitlement, and users should undertake their own independent assessment and seek appropriate expert advice to confirm such eligibility.

Do not rely solely on the NBAS code to make decisions about your project, as there may be a number of other factors to take into account.

The NBAS will be updated as the market evolves, knowledge improves, and technology changes over time. As this happens, the scores that NBAS calculates may change.

Attribution

This repository is developed and maintained by Haizea Analytics Pty Ltd. The NBAS code and NBAS data is based on research created under the Project "An Ecological Knowledge System for the Nature Repair Market", which was funded by DCCEEW. CSIRO is leading development of the EKS in partnership with DCCEEW.

Ownership of intellectual property rights

This publication is based (in part) on research and data created as part of the Project. The Commonwealth owns the intellectual property rights in any such new material developed while carrying out the Project.

Licensing

The code in this repository is licensed under an MIT Licence (See MIT license for more details). Additional material published as part of the Project is licensed under a Creative Commons Attribution 4.0 International licence, except for the Commonwealth Coat of Arms, any logos, any material supplied by third parties, any material protected by a trademark and any images and/or photographs. More information on this CC BY license is set out at the Creative Commons website.

Questions

For questions, please contact eks@dcceew.gov.au.