Methodology
How Travel Risk Calculator calculates city-level travel and current-location threat.
Assessment modes
- I am here now — current city threat.
- I plan to travel — pre-trip threat.
- I already arrived — on-trip threat.
- Compare two cities — side-by-side.
Scoring weights
Two different views exist intentionally. The interactive calculator (when you submit a trip via /calculator) runs the full 12-category model below. Static city pages at /destinations/… and comparison pages at /compare/… show only the first 9 categories, because the last three (accommodation, arrival-time, personal modifier) only become known once you enter trip-specific details. Both formulas use the same per-category weights, renormalized to the categories actually scored.
| Category | Weight | City page | Calculator |
|---|---|---|---|
| Crime | 18% | ✓ | ✓ |
| Disease & Health | 15% | ✓ | ✓ |
| Government Advisory | 12% | ✓ | ✓ |
| Political & Civil Unrest | 9% | ✓ | ✓ |
| Natural Disaster & Weather | 8% | ✓ | ✓ |
| Transport & Route | 13% | ✓ | ✓ |
| Environmental | 5% | ✓ | ✓ |
| Digital Security | 3% | ✓ | ✓ |
| Money Safety | 3% | ✓ | ✓ |
| Accommodation | 4% | — | ✓ |
| Arrival Time | 4% | — | ✓ |
| Personal Modifier | 6% | — | ✓ |
Risk tiers
- 0–24 Low
- 25–49 Moderate
- 50–69 Elevated
- 70–100 Extreme
- Unknown — no scoring data yet (rendered grey, not green; honest signal of absence)
City vs country fallback
Lookups try city first, then country, then a regional/world default. When a city page is using country-level fallback (no city-specific scoring yet), the page shows a "country-level baseline" label above the fold. The composite number stays comparable across modes because every consumer of the score uses the same CompositeScoreService.
Trip Readiness Score
Computed from the preparation checklist. Readiness measures preparation; it never reduces the threat figure.
Immediate Safety Score
For "I am here now" / "I already arrived" modes: weights short-term factors (crime, political, transport, arrival time, health, disaster).
Predict / Protect / Rescue / Restore
Every briefing includes structured PPRR actions tailored to the assessed location and situation.
AI-assisted assessment
When configured, Travel Risk Calculator queries the web for recent public-source context (Ollama Cloud’s web search by default, with self-hosted SearXNG as fallback) and asks an enabled AI provider for a strict JSON briefing. AI is used as an assessment layer: it explains top drivers, refreshes field recommendations from current open-source signals grounded in cited snippets, and records whether search and AI were actually used. The baseline calculator remains complete when live intelligence is unavailable.
Refresh cadence
New assessments run the live-intelligence pipeline at creation time. Search results are cached for a short window to reduce repeated calls, while saved assessment pages show whether AI/search were used for that briefing. Scheduled refresh jobs can rescan saved assessments as the corpus grows.
Data limitations
Baseline scores are demo estimates. Always verify with official advisories before travel or movement.