EFB Model Hazard Product Versus Official Warnings
Your Part 121 operator is evaluating a new electronic flight bag (EFB) product that ingests a global numerical model and displays colour-coded turbulence and icing risk along the route. The vendor says it is ICAO-aligned but provides no model run time, no verification statistics and no approval statement. On a trial flight from Christchurch to Auckland, the app shows severe turbulence over the Southern Alps at FL180, while the official SIGMET is only moderate turbulence and the AIRMET is moderate icing below FL160. The CEO wants to use the app to cancel flights and to brief crews. The safety manager asks whether it can be used operationally and what training is needed. The app is not an official meteorological product and has no CAA acceptance.
1. Classify the EFB product and explain how it differs from a SIGMET/AIRMET for operational decision-making.
Read the worked answer
The EFB product is a model-derived hazard display. It is not an observation, not a forecast issued by the meteorological authority, and not an official warning. A SIGMET is issued under ICAO Annex 3 criteria for severe or potentially hazardous weather and has a defined phenomenon, vertical and horizontal extent, validity and movement. An AIRMET covers moderate hazards using similar formal criteria. Both are authoritative warnings that crews are expected to use. The app instead interpolates a numerical model and applies a proprietary colour scale. Its severe turbulence over the Southern Alps may be a useful heads-up, but it is not evidence that the official SIGMET is wrong. Without model run time, resolution, latency and verification, the app cannot be treated as equivalent to an official product.
2. What criteria would you require before allowing operational use?
Read the worked answer
Before operational use I would require documented provenance: model name, run time, resolution, update latency, domain and known limitations. I would require independent verification against PIREPs, radar, SIGMET/AIRMET and accident/incident data, including false-alarm and miss rates in New Zealand terrain. The display must clearly distinguish model output from official warnings and must not hide the official product. There must be human-factors testing, crew training on interpretation and failure modes, a documented fallback when the app is unavailable, and a safety-case or change-management record approved by the operator. The app must not be the sole basis for cancelling a flight or for penetrating an area warned by SIGMET. Until those criteria are met, I would restrict it to trial and situational awareness only.
3. How should a model-derived hazard product be used in safety management and crew briefing?
Read the worked answer
It should be used as supplementary information, not as a replacement for official MET products. In briefing, the crew should first read the TAF, METAR/SPECI, SIGMET, AIRMET and GAFS, then use the app to visualize possible areas of turbulence or icing and to prompt discussion of escape routes. If the app and the official warning conflict, the official warning prevails and the crew should seek clarification from ATS or the meteorological provider. The operator should record how the product is used, monitor occurrences where it influenced a decision, and review its performance in the safety management system. The CEO should not use a single unverified model display to cancel flights; cancellation decisions should rest on official warnings, forecast trends and documented risk assessment.