Authors

Yin Su, Yao Zhao, Xiaokan Tian

Research Question

1. How do climate adaptation and renewable energy policies across federal, state, and Tribal levels in Oklahoma, New Mexico, and Arizona align with Tribal sovereignty, equity, and environmental justice, and what themes emerge from these policies through large language model analysis?
2. What spatial and demographic patterns can be identified through GIS and GeoAI analysis regarding the implementation and outcomes of these policies on Tribal lands, and how do these patterns reveal disparities or systemic barriers?
3. How can data-driven, evidence-based insights inform more coherent, equitable, and place-based climate adaptation strategies that better support the unique needs and governance structures of Tribal Nations across these three states?

Background

Tribal Nations across Oklahoma, New Mexico, and Arizona face intensifying climate-related challenges, including prolonged droughts, extreme weather events, and disruptions to traditional ecological systems. These threats are compounded by legacies of marginalization and limited access to policy mechanisms that support environmental resilience. At the same time, these three states are home to a high concentration and diversity of federally recognized Tribal Nations: 39 in Oklahoma, 23 in New Mexico, and 22 in Arizona, which makes them critical geographies for understanding the relationship between climate adaptation policy and Indigenous well-being. Each state presents a unique combination of legal frameworks, governance structures, and natural resource conditions, offering an opportunity for comparative analysis of how climate adaptation strategies play out in Tribal contexts.

This project uses Artificial Intelligence (AI), specifically Large Language Models (LLMs), to analyze climate-related policies and strategies across federal, state, and Tribal levels. These tools will help process and interpret large volumes of policy documents of various quality and completeness, extracting patterns, themes, and insights related to governance, equity, and implementation. In addition, spatial analysis will be conducted using GIS software (ArcGIS Pro) and GeoAI (Geospatial Artificial Intelligence) to examine how climate risks, energy infrastructure, and policy investments are distributed across Tribal lands. Combined with demographic and socioeconomic data, this analysis will provide a comprehensive, data-driven understanding of how current policies impact Tribal Nations and what could lead to future improvements. The goal is to generate evidence-based insights and spatially grounded solutions that can better align climate adaptation policies with the sovereignty, needs, and sustainability priorities of Tribal communities.

Methodology

This project employs a mixed-methods approach combining AI-powered analysis with geospatial and quantitative techniques to evaluate the impact of climate adaptation and renewable energy policies on Tribal Nations in Oklahoma, New Mexico, and Arizona.

• Policy Analysis with Large Language Models (LLMs): Climate-related policy documents from federal, state, and Tribal sources will be analyzed using LLMs to identify themes related to equity, sovereignty, and policy structure.
• Spatial Analysis with GIS and GeoAI: ArcGIS and GeoAI will be used to map climate risks, infrastructure, and policy implementation across Tribal lands, revealing spatial patterns and disparities.
• Quantitative Analysis: Socio-demographic and environmental datasets from sources like the U.S. Census, NOAA, and EPA will be analyzed using statistical methods to assess policy outcomes and equity impacts.
• Informational Stakeholder Input: When feasible, the team may conduct non-extractive, informational discussions with Tribal representatives to contextualize findings and enhance cultural relevance.

Deliverables

• Policy Analysis Report (in a format of an academic journal manuscript): A comprehensive, data-driven report summarizing key findings from the analysis of federal, state, and Tribal climate adaptation policies using large language models. The report will highlight themes, equity implications, and policy gaps impacting Tribal Nations in the three-state region.
• Geospatial Visualization: An interactive ArcGIS-based StoryMap or dashboard displaying spatial data on climate risks, policy implementation, infrastructure, and demographic indicators across Tribal lands, enabling stakeholders to explore patterns and disparities.
• Evidence-Based Policy Recommendations: A set of actionable, equity-centered recommendations informed by AI and spatial analysis, designed to guide federal, state, and Tribal stakeholders in improving climate adaptation strategies for Indigenous communities.
• Peer-reviewed Journal Publication: The result of this work is expected to be published in a peer-reviewed academic journal.

Data Sources

The project will primarily rely on publicly available datasets and documents. These include:

• Policy and Planning Documents: Federal, state, and Tribal climate adaptation and energy policy reports, legislation, and strategy documents - sourced from government websites, Tribal archives, and research repositories. These will be compiled and pre-organized for the Capstone team.
• Socio-Demographic Data: U.S. Census Bureau and American Community Survey (ACS) datasets, along with Tribal population profiles and socio-economic indicators - provided through public APIs or downloadable formats.
• Climate and Environmental Data: Open-source datasets from sources such as NOAA, U.S. Large-Scale Solar Photovoltaic Database (USPVDB); U.S. Wind Turbine Database (USWTDB), EPA, and state-level climate resilience portals - covering climate risk, land use, and environmental vulnerability.
• Spatial Data: Shapefiles and geospatial layers for Tribal lands, infrastructure, and hazard zones - available through ArcGIS Online, federal databases (such as TIGER), and academic GIS libraries.