# TremorWatch Analyst Review Pack

Snapshot: 2026-05-12

This pack is for analysts and consultants evaluating whether TremorWatch public risk exposure data can fit client memos, market notes, exposure monitoring, or internal research workflows.

## Files

- `50_recent_mapped_events.csv` — 50-row review sample prioritizing FIRMS/GDACS rows, APAC rows, mapped exposure rows, and non-noise high-severity events.
- `schema_dictionary.md` — field-level dictionary.
- `source_counts.md` — sample row counts by source, severity, country, event type.
- `null_rates.md` — missing/present rate by column.
- `mapping_coverage.md` — country, lat/lng, commodity, chokepoint, KO/JA coverage.
- `sample_queries.sql` — starter analyst queries.
- `reproducible_findings.md` — 3 generated findings to validate.
- `limitations.md` — known gaps and non-claims.
- `events_sample.csv` — original marketplace sample copy.

## Review Questions

1. Would this fit a client memo, market note, dashboard, or internal briefing?
2. Which field is missing before this becomes usable?
3. Is country / commodity / chokepoint mapping valuable, or too noisy?
4. Do you need confidence scores or mapping rationale?
5. Is daily refresh useful, or would weekly be enough?
6. Would CSV be enough, or do you need API / RSS / BigQuery?
7. What would make this worth paying for?
8. Which existing tool/data source would this replace or complement?

## Selection Note

The 50-row review sample intentionally excludes obvious low-confidence / no-bearing GDELT rows from the outreach sample where detected. The full `events_sample.csv` is still included so reviewers can inspect raw sample quality and false positives.