Methodology / data freshness
How HazardNow works
HazardNow combines public hazard, infrastructure, travel, environmental, health, and stability signals into a single situational-awareness dashboard.
What it does
Aggregates public signals, labels them for fast scanning, and links users back to source context and the live dashboard.
What it does not do
HazardNow is not an official alerting system, a replacement for local emergency management, NWS, FEMA, public-health agencies, utilities, airlines, or other official sources, and it does not provide personal emergency instructions.
How to use it
Use the dashboard to notice context and decide what to verify next. Use official alerts, local emergency management, utility, airline, and agency sources for final decisions.
Status-label methodology
Dashboard cards convert source-specific data into simple labels such as normal, elevated, delayed, tight, stale, degraded, or unavailable. The label is a scan aid, not a final decision authority.
Each card favors plain-English summaries, source freshness, known limitations, and links to related source pages. When source data is missing or too old for the expected cadence, HazardNow should show degraded, stale, unavailable, or similar language instead of silently implying normal conditions.
Weekly or slower public sources should not be treated as stale simply because they do not update hourly. Freshness is interpreted relative to each provider's normal publication cycle.
Source transparency table
Examples of source families, cadence expectations, and where they appear in the dashboard.
| Category | Example sources | Typical cadence | Dashboard area |
|---|---|---|---|
| Historical disaster records | OpenFEMA Disaster Declarations Summaries v2, NOAA NCEI Storm Events Database | The FEMA heatmap refreshes weekly. Atlas profiles use validated source generations and retain the last-known-good generation when a refresh fails. | Disaster History Atlas |
| Weather alerts and forecasts | National Weather Service, NWS API | Refreshed frequently by the dashboard and HazardNow cache process; exact timing can vary by NWS endpoint and region. | Live dashboard |
| Weather map layers | NOAA/NESDIS/STAR GOES imagery, RainViewer, Storm Prediction Center | Varies by source. The GOES loop route uses NOAA/NESDIS/STAR imagery with a 300-second route revalidation/cache TTL; GOES CONUS frames are labeled about 5 minutes, full-disk about 10 minutes, and mesoscale sectors about 1 minute. Radar and outlook products follow their own source cadence. | Live dashboard |
| Wildfire, smoke, and air quality | NASA FIRMS, AirNow, National Interagency Fire Center | FIRMS observations depend on polar satellite overpasses; regional RT/URT delivery can be fast after a pass but does not provide continuous observation between passes. GOES rapid fire detections are documented separately. Air quality and incident context follow their own publication and cache timing. | Live dashboard |
| Rapid fire observations | NOAA Wildfire Data Portal, NGFS scene detection definitions, NASA FIRMS delivery and overpass context | GOES CONUS scenes are about every five minutes; selected mesoscale scenes about every minute. HazardNow checks the source through its shared two-minute writer and polls the cached dashboard every 60 seconds. Displayed observation age includes publication and collection delay. | Live dashboard |
| Evacuation reports | Cal OES evacuation aggregation dataset and public-use terms, Cal OES county coverage inventory | Cal OES advertises five-minute aggregation. HazardNow uses a five-minute minimum ingest interval on the shared two-minute scheduler, normally yielding six-minute checks. Zone EDIT_DATE is a status-change time, not a heartbeat; the browser polls cache every 60 seconds. | Live dashboard |
| Independent connectivity observations | RIPE Atlas, RIPE Atlas commercial-use conditions, IODA | Provider status/event publication latency is not guaranteed. HazardNow uses a five-minute minimum ingest interval on shared two-minute scheduler ticks, normally yielding six-minute checks; browser reads are cache-only every 60 seconds. Probe timestamps remain separate from fetch time. | Live dashboard |
| Rapid rainfall and flood potential | NOAA FLASH, MRMS product cadence table, MRMS data access | MRMS produces this product about every two minutes. HazardNow ingests through the shared two-minute writer and polls cache every 60 seconds. Dissemination and collection delay can make the analysis older than the nominal product cadence. | Live dashboard |
| Observed road hazards | Caltrans RWIS feed documentation, Caltrans conditions of use | Caltrans files refresh about once per minute. HazardNow checks through the shared two-minute writer and polls cache every 60 seconds; each station's recordEpoch determines observation age. | Live dashboard |
| Coastal water-level observations | NOAA Coastal Inundation Dashboard, NOAA CO-OPS Data API, NOAA CO-OPS disclaimers and public-use policy | This feed uses preliminary six-minute measurements, checked through the shared two-minute writer and polled from cache every 60 seconds. Station publication delay varies; one-minute NOAA products are not ingested by this adapter. | Live dashboard |
| Developing weather assessments | SPC mesoscale discussions, SPC RSS documentation, NOAA discussion GIS service and update timing | SPC issues discussions as needed. HazardNow checks RSS and direct text through the shared two-minute writer and polls cache every 60 seconds. NOAA GIS publication can lag by up to 15 minutes; source validity determines expiry. | Live dashboard |
| Official service incidents | Cloudflare status, Google Cloud Service Health, AWS Health status | Providers publish incident updates as events develop, with no guaranteed detection latency. HazardNow checks through the shared two-minute writer and polls cache every 60 seconds; publication time remains separate from the source-check time. | Live dashboard |