Skip to content

This project utilizes open science, optimization, and AI to support planning and decision-making for protected areas.

License

Notifications You must be signed in to change notification settings

CU-ESIIL/AI-opt-decision-making

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Optimization, Open Science, and AI for Biodiversity Conservation Decision-Making

Welcome to the Optimization, Open Science, and AI for Biodiversity Conservation Decision-Making repository, part of the Environmental Data Science Innovation and Inclusion Lab (ESIIL).

The Project

This project utilizes open science, optimization, and AI to support spatial planning and decision-making for protected areas. In particular, we focus on building a decision-support tool for the 30x30 initiative (preserving 30% of the Earth’s lands and waters by 2030).

Project Proposal

The ongoing biodiversity crisis is rapidly escalating with alarming rates of ecosystem deterioration and species extinction. In the face of biodiversity loss, decision-support software allows scientists and policymakers to model, monitor, and analyze ecological data, facilitating more informed and timely decisions for conservation and restoration efforts. However, software that relies on closed-platform data restricts the user's ability to dissect and extract specific information, ultimately reducing accessibility and limiting innovation.

Our goal is to deliver customizable and interactive open science decision tools that 1) are better designed to yield more equitable solutions with state-of-the-art optimization methods and 2) effectively communicate the framework and results to communities impacted by these decisions. From exploring local biodiversity to making informed land investment decisions, we aim to promote public engagement and scientific research by enabling users to explore data tailored to their needs.

To date, we have developed a cloud-native geospatial visualization tool with chat-driven interfaces for users to map and query California protected land data, enable by open data layers and LLMs. Our decision-support tool will leverage more flexible optimization frameworks that address environmental goals while also prioritizing the needs and rights of those affected by these decisions.

Preliminary results were presented at AGU 2024: AGU Poster.

Our working prototype is hosted on Hugging Face: California 30x30 Planning & Assessment Prototype.


A screenshot of the California 30 by 30 Planning and Assessement Prototype with a map of protected areas, color-coded by year (green for pre-2024, yellow for 2024), toggles for grouping and filtering by attributes, and a donut chart showing 25.2% land protected.
The California 30 by 30 Planning and Assessement Prototype

Collaborators and Co-Authors

  • Dr. Cassie Buhler | Website | ESIIL Postdoc
  • Dr. Carl Boettiger | Website | ESIIL Mentor at UC Berkeley

Code Repository

The code is available at this repository: California 30x30 GitHub Repository.

Meeting Notes and Agendas

We managed and organize our project in the CA-30x30 repository, where ongoing work and project milestones are tracked in the issues.

About

This project utilizes open science, optimization, and AI to support planning and decision-making for protected areas.

Resources

License

Stars

Watchers

Forks

Packages

No packages published