Augur
Sign inStart free
Live demoPlatformHow it worksPricingStatusFAQSupply chainEnergy + infrastructureNewsroom OSINTAll integrationsSign inStart free

Operations · 4 min

Building a Disaster Awareness Dashboard with GDACS and ReliefWeb

GDACS bulletins + ReliefWeb humanitarian briefs combined give NGOs, insurers and supply-chain teams 24/7 disaster awareness with population-exposure scoring. Practical wire-up + filter patterns.

2026-05-26

GDACS — the Global Disaster Alert and Coordination System — is the EU's clearinghouse for major disaster bulletins. ReliefWeb is OCHA's humanitarian situation report archive. Wire both and you have the most reliable free disaster monitoring stack outside national agencies.

This post covers the practical pipeline: which feeds to consume, what severity scoring works, and how insurers, NGOs and logistics teams keep on top of disasters without staffing a 24/7 watch desk.

GDACS — the headline feed

GDACS publishes coordinated bulletins on:

  • Earthquakes (cross-referenced with USGS + EMSC)
  • Tropical cyclones (cross-referenced with NHC + JMA + JTWC)
  • Floods (modelled from GloFAS satellite hydrology)
  • Volcanic eruptions
  • Droughts (modelled)
  • Forest fires (cross-referenced with NASA FIRMS)

Real-time feeds via RSS/CAP:

https://www.gdacs.org/xml/rss.xml
https://www.gdacs.org/xml/rss_alerts_3w.xml  (last 3 weeks)

Each bulletin carries a population-exposure estimate, severity colour (green / orange / red), and links to the source agency's detailed report. The severity colour is the single most useful normalised signal — GDACS does the cross-source aggregation work for you.

ReliefWeb — humanitarian context

ReliefWeb publishes situation reports from UN agencies, ICRC, MSF, World Bank, government emergency-management offices. Real-time API:

https://api.reliefweb.int/v1/reports?appname=YOUR_APP&filter[field]=country&filter[value]=SY

No key required. Slow-changing feed compared to GDACS — most reports drop hours-to-days after the event peaks. The value is in the humanitarian context layer: not just "earthquake happened" but "70,000 people displaced, 12,000 buildings damaged, three NGOs deploying."

The pattern: GDACS for trigger, ReliefWeb for context

Combine them like this:

  1. GDACS fires the initial alert with population-exposure estimate
  2. Auto-query ReliefWeb for the affected country in the following 6-72 hours
  3. Attach the latest ReliefWeb sitrep to the GDACS alert thread

For insurers and humanitarian logistics teams this combination drives the entire decision loop: deploy or stand down, escalate or wait.

Severity scoring

GDACS already does most of the work. Their severity colour maps to:

Green  = sev 30  (minor, local impact only)
Orange = sev 60  (regional impact, mobilise national response)
Red    = sev 85  (international impact, mobilise UN response)

Bump severity by event type if your operation has a specific risk profile. Insurance underwriters care more about earthquakes (asset damage); logistics teams care more about cyclones + floods (route disruption); NGOs care more about droughts + displaced-persons reports.

Geofencing for disaster awareness

Three patterns:

1. Country-level polygons

Most insurance and NGO operations care at the country level. Download Natural Earth or OSM country boundaries. Set severity threshold to ≥60 (orange + above).

2. Asset-radius circles

For supply-chain teams with specific port / facility exposure, define 100km circles around top assets and set severity threshold to ≥50 (orange + above + green-but-close).

3. Disaster-corridor polygons

For ongoing crises (Ukraine, Gaza, Sudan, Yemen), define wide polygons around the active conflict zone and lower the severity threshold to ≥40. You'll catch every relevant ReliefWeb sitrep + GDACS bulletin.

Practical false-positive reductions

GDACS is curated — almost zero noise. ReliefWeb is high-volume but high-quality. Two tunings worth applying:

  1. Filter ReliefWeb by format=Situation Report to drop the daily news roundups. SITREPs are the actionable subset.
  2. Dedupe GDACS by event_id to avoid double-firing on initial bulletin + first revision.

Free starter stack

Build the minimum viable disaster awareness dashboard this week:

  1. Poll GDACS RSS every 15 minutes
  2. Poll ReliefWeb API every 60 minutes per country
  3. Define country polygons or asset-radius circles for your top markets
  4. Normalise severity to 0-100 using GDACS colour + ReliefWeb sitrep count
  5. Slack channel #ops-disasters, route via incoming webhook

Augur's GDACS + ReliefWeb ingest wraps the dedupe + cross-reference + geofencing pipeline. Free tier covers most NGO and insurance use cases.

When you need more

For dedicated disaster-response operations:

  • PDC DisasterAWARE — Pacific Disaster Center's paid platform, sub-minute latency
  • Aon Impact Forecasting — actuarial-quality damage estimates within hours
  • Munich Re NatCatSERVICE — insurance-industry standard

For the first 90% of operational awareness use cases, GDACS + ReliefWeb + a free OSINT geofencing layer is more than enough.

What this looks like in production

Three real outcomes from teams running disaster monitoring this way:

Microfinance NGO: wired GDACS + ReliefWeb against 12 operating-country polygons. Deployment decisions now driven by orange-bulletin + sitrep arrival, not weekly all-hands.

Mid-market reinsurer: GDACS orange/red alerts + 100km asset circles drove a 4-hour reduction in first-notification-of-loss reserve adjustments.

Supply-chain ops at a global brand: GDACS bulletins automatically push polygon zones to dispatcher; the team gets a "your suppliers in [country] are now in an orange-bulletin zone" message instead of reading the bulletin themselves.

The signal is public. The wire-up is the moat.

← Back to blog · Start free