Screen 11 · Section 5.4
Risk Matrix
Nine risk statements from the proposal, each paired with its mitigation. No severity scores are added — the source doesn't provide them.
Category
Risk
Mitigation
Field access
Sites inaccessible during monsoon or restricted by local authority.
Sequence fieldwork by state; keep contingency clusters; coordinate access letters via PMU.
Data availability
Avni / Dalgo / Deloitte inputs delayed or partial.
Confirm data-sharing at inception; adjust confidence tier if a stream is missing.
Comparison-site
Comparison sites contaminated by concurrent schemes.
Screen for concurrent PMKSY / MGNREGA works; document exposure explicitly.
Seasonality / rainfall
Anomalous rainfall confounds attribution of water-availability changes.
Anchor readings to IMD normals; report signals with rainfall caveats.
Scientific testing
Soil / water lab turnaround exceeds fieldwork window.
Pre-book labs at inception; stagger sample dispatch.
Satellite / geospatial
Cloud cover degrades water-spread readings for target season.
Use multi-sensor composite; fall back to nearest clear window with caveat.
Recall bias
Household recall of pre-intervention conditions weakens Phase 1 estimates.
Triangulate with Deloitte baseline where available; use bounded recall periods.
Attribution
Confounding schemes limit causal inference.
Explicit contribution analysis; DiD where baseline permits; qualified attribution language throughout.
Coordination / Data privacy
Multi-state coordination delays; PII in household data.
PMU liaison per state; anonymise PII on ingestion; role-based access to raw data.