Software

An open-source library for building map-based applications with LLM-powered data analysis. GLEN is the shared framework behind several of the conservation decision-support tools below, pairing interactive geospatial visualization with natural-language querying over cloud-optimized spatial data.

Github

Boettiger, C., & Buhler, C. K. (2026). GLEN (Geospatial LLM-Enabled Navigator): A reusable library for map-based applications with LLM-powered data analysis (v3.11.1). Zenodo. https://doi.org/10.5281/zenodo.20693215

An interactive marine conservation decision-support tool for the high seas — the areas of the ocean beyond national jurisdiction. Built on the GLEN framework, it pairs interactive maps with a natural-language agent that runs SQL over H3-indexed ocean data, letting users explore candidate high seas marine protected areas, exclusive economic zone boundaries, seafloor geomorphology, bathymetry, protected-area coverage, and global fishing-effort data.

App Github

Boettiger, C., & Buhler, C. K. (2026). High Seas Explorer (v1.0.3). Zenodo. https://doi.org/10.5281/zenodo.20693689

An interactive decision-support tool for wildlife habitat and public lands in the Greater Yellowstone Ecosystem, centered on Wyoming. Built on the GLEN framework, it pairs interactive maps with a natural-language agent that runs SQL over H3-indexed data, letting users explore Wyoming Game & Fish Department seasonal wildlife range data, public land management boundaries, and related habitat datasets.

App Github

Boettiger, C., & Buhler, C. K. (2026). Public GYE Explorer (v1.0.2). Zenodo. https://doi.org/10.5281/zenodo.20693688

A national conservation decision-support tool integrating protected areas, conservation finance, climate, biodiversity, and social vulnerability data. Built on the GLEN framework, it pairs interactive maps with a natural-language agent that writes SQL over H3-indexed spatial data, enabling real-time joins across conservation investments, ballot measures, carbon, species richness, legislative districts, and environmental justice indicators. It draws on TPL's Conservation Almanac and LandVote, the U.S. Climate and Economic Justice Screening Tool (Justice40), and the CDC Social Vulnerability Index.

App Github

Buhler, C. K., & Boettiger, C. (2026). Trust for Public Land Explorer (v3.0.5). Zenodo. https://doi.org/10.5281/zenodo.20693686

A California-focused conservation decision-support tool for the Trust for Public Land. Built on the GLEN framework, it pairs interactive maps with a natural-language agent that runs SQL over H3-indexed data, letting staff, legislators, and advocates explore land conservation investment recorded in the Conservation Almanac, carbon stocks, legislative and congressional districts, and Indigenous lands to support conservation planning and policy advocacy.

App Github

Boettiger, C., & Buhler, C. K. (2026). Trust for Public Land California Explorer (v1.0.2). Zenodo. https://doi.org/10.5281/zenodo.20693687

A decision-support tool developed in collaboration with the California Biodiversity Network to align with the CA 30x30 Biodiversity Assessment, building off the prototype with more polished interface, additional open weights language models, and a broader set of ecological, socio-environmental, and climate-related data. With this tool, you can explore California’s 30x30 conservation areas, other conservation areas, and non-conserved lands, revealing what is protected, what isn’t, and opportunities to fill the gaps.

Hugging Face Github

Buhler, C. K., & Boettiger, C. (2026). CA 30x30 Planning & Assessment Tool (v1.0.7). Zenodo. https://doi.org/10.5281/zenodo.19561674

An interactive decision-support and exploratory analysis tool for U.S. land conservation ballot measures, integrating the Trust for Public Land LandVote database with election, social vulnerability, and environmental datasets. The application combines a lightweight web interface with open-weights language models to enable natural-language querying of conservation funding, voting outcomes, and jurisdictional trends.

Hugging Face Github

Buhler, C. K., & Boettiger, C. (2026). LandVote LLM Decision-Support Prototype (v1.1). Zenodo. https://doi.org/10.5281/zenodo.18500783

Proof of concept for a decision support tool developed in partnership with California Biodiversity Network participants through a co-design process. The tool can answer complex, real world natural language queries asked by conservation partner organizations, responding with reproducible, verifiable data summaries, charts, maps and text through careful integration of open weights language models and cloud optimized data.

Hugging Face Github

Buhler, C. K., & Boettiger, C. (2025). CA 30x30 Planning & Assessment Prototype (v0.0.1). Zenodo. https://doi.org/10.5281/zenodo.14933817

A spatial planning tool that utilizes mixed-integer nonlinear programming (MINLP), thus can be paired with ecological software (e.g. population viability analysis). Existing spatial tools typically only allow linear inputs and discrete variables. However, our MINLP framework enables the use of linear and nonlinear inputs, as well as both continuous and discrete variables.

Github

Buhler, C. K., & Benson, H. Y. (2024). Conservation-DFO: Initial Release (v1.0.0). Zenodo. https://doi.org/10.5281/zenodo.13742961

A quasi-Newton optimization algorithm that improves the step quality of conjugate gradient methods by selectively using cubic regularization. Compared to its non-regularized counterpart, this method exhibited fewer iteration counts and faster runtime when solving unconstrained optimization problems during numerical testing.

Github

Buhler, C. K., & Benson, H. Y. (2024). Conmin-CG: Initial Release (v1.0). Zenodo. https://doi.org/10.5281/zenodo.13315592