Results: Community-Informed Climate Impact Assessment of Nunapitchuk, Alaska
This is a preliminary analysis and interactive site combining results of Arctic Infrastructure Analyses that link narrative interview information with on the ground environmental information collected using drone sensors. The study presents a mixed-methods analysis integrating stakeholder interviews (n=7, ages 43-80) with geospatial data to assess climate-induced infrastructure vulnerabilities in Nunapitchuk, Alaska. Through systematic coding of interview transcripts, we identified 47 references to permafrost degradation, representing the dominant environmental concern across all stakeholder groups. The analysis revealed critical infrastructure failures affecting 38 documented structures, with differential subsidence rates showing south-facing foundations degrading 40% faster than north-facing structures.
Our Scientific Variable Object (SVO) framework mapped nine quantifiable metrics to stakeholder priorities, revealing high concordance (>80%) between elder traditional knowledge and scientific measurements for permafrost thaw rates and ground subsidence indicators. Water quality assessments detected arsenic concentrations downstream from waste facilities, corroborating community-reported health concerns including unprecedented cancer clusters and first-time waterborne illnesses among lifelong residents.
The perceived environmental change trajectory, validated through temporal analysis spanning 1920-2024, demonstrates accelerating degradation with ground stability declining from 95% (1970s) to 30% (2020s). Multiple stakeholders independently estimated a 3-5 year window before complete infrastructure failure, aligning with geotechnical projections showing active layer depths increasing at 12cm/year.
Economic impact analysis revealed a 900% increase in energy costs over one generation ($20/barrel to $190/month), threatening community sustainability. The 3D point cloud visualization captured 2.3km² of riverbank terrain at 15-point/m² density, documenting erosion rates of 2.5m/year and permafrost-related surface deformation patterns.
Critical findings indicate immediate federal intervention is required, with $11 million already allocated for waste management upgrades. The convergence of traditional knowledge with quantitative measurements strengthens the validity of relocation timelines. This integrated assessment framework provides a replicable methodology for climate impact evaluation in Arctic communities, demonstrating the value of incorporating indigenous knowledge systems into engineering assessments for infrastructure resilience planning.
Comparison: AI vs Human Coding AnalysisKey Differences:
The AI analysis identified significantly more thematic references (215 total) compared to human coding (19 entries), with AI emphasizing permafrost degradation as the primary concern while human coders focused more granularly on specific infrastructure issues like boardwalks and dump management. Human coding showed the City Manager contributing 58% of coded entries, suggesting more detailed attention to administrative perspectives, whereas AI analysis distributed concerns more evenly across stakeholder groups. The AI approach captured broader thematic patterns and frequencies, while human coding provided more precise, quote-specific categorizations with detailed contextual understanding. Both approaches identified waste management and infrastructure failure as critical issues, but differed in their emphasis on cultural preservation (AI) versus practical infrastructure solutions (human coders).Similarities: Both identified dump/waste management and infrastructure problems related to ground instability as critical issues, recognized water quality concerns, and captured the urgency of the situation requiring immediate intervention.
