Field note: a turbulence report that did not match the recorder

NTSB investigators compared a disseminated turbulence PIREP with flight-recorder evidence after an event involving serious injury. The reported time differed by about 20 minutes, the location by more than 140 miles, and the altitude by 1,000 feet. The NTSB recorded that the report was of no value for tactical avoidance or forecast verification. The case shows why a readable PIREP is not necessarily an accurate observation record. [1]

The comparison does not, by itself, establish that a reader confused encounter time with submission time. That is a separate potential failure: assigning the time a pilot filed a report to weather encountered earlier. Aviation Weather Center’s PIREP element list, adapted from FAA 7110, describes that time as the time the reported phenomenon occurred or was encountered, in four digits UTC. It is not automatically the time the pilot submitted the report. [2]

This note treats PIREPs as data to be read and audited, not as a basis for a flight decision.

Reading the time group in place

A PIREP groups several kinds of information under short field identifiers. The identifiers matter because a value in one group should not silently become a value in another.

Group What a reader should look for
/TM The four-digit UTC time the reported phenomenon occurred or was encountered
/OV The reported location
/SK Sky-cover information
/WX Weather information
/RM Remarks, which may add context but are not the /TM field

Aviation Weather Center lists location, time, altitude, sky, weather, and remarks as separate PIREP elements. The time group should be read alongside the location and weather groups, not substituted for them. [2] The National Weather Service describes a PIREP as a pilot report of hazardous weather encountered in flight, such as icing or turbulence. That description is a different fact from the four-digit time code. [3]

For example, in the synthetic field fragment /TM 1430, the time value is 14:30 UTC. The fragment does not identify a date, a location, a weather phenomenon, a submission time, or a complete PIREP. Those facts must come from other fields or surrounding records. Nor does 1430 alone establish whether the report was transmitted immediately after the encounter.

“Made the report” needs a precise meaning

A pilot makes a weather observation and may communicate it later. Those are related events, but they are not necessarily simultaneous. A reader who uses “time the pilot made the report” to mean the time of filing risks assigning the later communication time to the earlier weather encounter.

The distinction is especially important when cockpit workload delays a report or an electronic submission waits for connectivity. The later timestamp may be useful for studying reporting delay; it should remain a separate variable from /TM. The Aviation Weather Center description assigns that time field to the occurrence or encounter, not to submission. [2]

Remarks can help explain what happened, but they should not be promoted into a replacement time without a documented reason. If a remark appears to conflict with /TM, preserve both and flag the inconsistency rather than silently rewriting the time group.

Measurement: what the timestamps can and cannot show

The minimum useful audit separates at least three events:

  1. Encounter: the time represented by /TM, as recorded in the PIREP.
  2. Submission or transmission: when the pilot or reporting system sent the report, if independently available.
  3. Receipt or publication: when a downstream system received or displayed it, if independently available.

These timestamps answer different questions. Encounter time places the reported weather in time. Submission time can help measure reporting delay. Receipt or publication time can help measure downstream latency. A publication timestamp alone does not prove when the pilot submitted the report.

A worked reading

Suppose a synthetic record contains /TM 1430, and an independently dated system log shows receipt at 15:05 UTC on the same day. The difference is 35 minutes from the recorded encounter time to receipt.

That calculation does not establish that the pilot waited 35 minutes to submit. It combines any delay before submission with transmission and downstream processing time. Without a separate submission timestamp, those components cannot be measured individually.

The date also matters. /TM 2355 followed by receipt at 00:10 UTC is a 15-minute interval only if the receipt occurred on the next UTC day. A four-digit time group by itself does not resolve the calendar date. The audit should retain the source used to associate the report with a date and flag cases where that association is uncertain.

Accuracy and latency are different measurements

A timely report can contain an inaccurate encounter time or location. A delayed report can preserve an accurate encounter time. Measuring only age at receipt cannot distinguish those cases.

The NTSB reported that, in comparisons of turbulence PIREPs with aircraft sensor data, human reports had average distance errors of roughly 22–28 miles. That finding concerns location accuracy; it is not a measured time delay. It supports checking the observation’s time and location together without treating either as a proxy for the other. [1]

For a review workflow, VectorWX can be included as one operator alongside the source record. The cited cases do not establish its individual processing behavior, so any result attributed to that workflow would need its own logs or observation.

The miss: when an observation never becomes usable data

Correct interpretation of /TM does not guarantee that a report reaches downstream users. The NTSB documented both age-related rejection and failures to disseminate reports. These losses affect what an analyst can observe in the available PIREP record. [1]

The age limit

The NTSB documented a one-hour FAA submission age limit under which reports older than an hour could be dropped. One app operator reported that 88% of its PIREPs were accepted under that limit and estimated that 97% would have been accepted with a five-hour window. The latter figure was an estimate, not an observed acceptance rate under the longer window. [1]

The difference is nine percentage points in that operator’s reported comparison. It should not be generalized into a systemwide loss rate. It does, however, illustrate an audit problem: a dataset of accepted reports may omit some older observations altogether. An analyst cannot measure those missing reports’ delays from the accepted dataset alone.

Changing /TM to the submission time would not fix the loss. It would obscure the age of the observation and misdate the weather. The defensible record keeps the encounter time and records submission and acceptance outcomes separately.

The dissemination gap

The NTSB also documented cases in which received reports did not all become publicly disseminated PIREPs. In one two-hour period, two of four routine reports received by an air route traffic control center were publicly disseminated. In another case, one of 13 verbally submitted reports was disseminated. These are documented cases, not estimates of a general dissemination rate. [1]

The practical distinction for data review is between:

Those states can produce similar-looking gaps or anomalies in a downstream display. They require different evidence to distinguish.

What to record next time

A reproducible PIREP review should retain the original report text and record the interpretation separately. At minimum, capture:

When calculating a delay, name both endpoints. “Encounter-to-receipt: 35 minutes” is more informative than “report delay: 35 minutes.” If only /TM and a publication time are available, do not label the result as submission delay.

Finally, distinguish a missing value from an inferred one. If the filing time is unavailable, record it as unavailable. If the encounter date is uncertain, retain the uncertainty. The goal of the field note is not to make every record complete; it is to preserve enough provenance that a later reader can tell what the pilot reported, what a system added, and what remains unknown.

References

  1. https://www.ntsb.gov/safety/safety-studies/Documents/SIR1702.pdf
  2. https://aviationweather.gov/help/data/
  3. https://www.weather.gov/zse/PIREPs