[Good First Issue] Wire carbon analytics into API response and add impact report endpoint

[Oshgig]

Overview

When a deforestation run completes, the API should include carbon loss estimates in the response. The src/climatevision/analytics/carbon.py module exists but is not wired into the API response.

Scope

• Review src/climatevision/analytics/carbon.py for production readiness
• If not ready, wrap carbon estimation behind a feature flag (e.g., enable_carbon: true in request)
• Integrate carbon computation into run_inference() or post-processing step for analysis_type=deforestation
• Add carbon_tonnes and hectares_lost to the API response schema
• Add GET /api/reports/{run_id} endpoint returning structured impact report:
  • hectares_lost
  • carbon_tonnes
  • confidence_interval
  • region_bbox
• Wire the report endpoint to the frontend dashboard

Acceptance Criteria

• Deforestation API response includes carbon_estimation field
• /api/reports/{run_id} returns a complete impact report
• Feature flag gracefully degrades if carbon module is unavailable
• At least one test validates carbon math against known values

Resources

• src/climatevision/analytics/carbon.py — existing carbon module
• src/climatevision/api/main.py — API endpoints
• src/climatevision/inference/pipeline.py — inference flow
• team_docs/PM_ROLE_UPDATES_FOR_GAPS.md — Francis' tasks

Difficulty: Intermediate
Labels: help wanted, backend, analytics, api