How to read a hail report: what each figure means

The date and the window

A hail report is read from the top: first the window it searched, then the days inside the window that carry hail, then the records on each day. The window is the pair of dates the document was compiled for, and it is the first thing to check, because a report is only evidence about the days it looked at. A claim names one date of loss; a report compiled for that day alone would be worth exactly what the date was worth. A report compiled for a window shows the named day beside its neighbours, so the reader can see whether it stands alone in the record or is one of several.

Inside the window, the document lists storm days: each calendar date on which hail was reported near the address, in date order. A day is a date as the source filed it, in the local time of the office that filed it, and a storm that crossed midnight is filed on two dates. The report does not merge them into one storm, because the record did not; it lists both days and lets the reader see that they are a pair.

Each day carries a handful of figures: the largest size measured on the ground, the largest size detected and not measured, the number of records, and the number of independent sources that carried the day. The rest of this guide is those figures, one at a time. The day's page on the storm pages shows the same figures for the whole country; the report shows them for the roof.

Measured against detected

The size column is the one a reader goes to first, and it is two columns. A hail size in the first is a stone somebody recorded on the ground: a spotter who compared it with a coin, an observer with a gauge, an office that accepted a report from the public. The record files it as a ground observation, at a point, at a time. A size in the second column was computed from radar for a grid cell, whether or not anyone stood in the cell. The record files it as a gridded analysis and the document prints its qualifier, Analyzed, beside every value.

The two are different objects, and the document never lets them share a column, a mark or a sum. A measured size is drawn as a filled mark in the observed tone; a detected size is drawn as a hollow ring in the estimated tone. The shapes differ as well as the colours, so the difference survives a photocopy and a reader who cannot tell the tones apart.

A measured size

The diameter of the largest stone found at a place, as the observer assigned it. The stronger kind of evidence, and not a laboratory figure: a stone picked up after some melting and compared by eye.

A detected size

The largest stone the storm could have produced over a grid cell, as the radar algorithm computed it. A ceiling by construction; the algorithm was built so that most measured stones fall below it.

Both on one day

Two independent kinds of evidence that hail fell, not one figure attested twice. The report prints each with its source and adjusts neither toward the other.

A day counts as severe on these pages when its largest report reached 1.00″, the National Weather Service criterion, whichever kind of report it was, and the page says which. A reader who wants to know what fell reads the measured column. A reader who wants to know whether the storm's hail signature passed over a roof nobody reported from reads the detected column. Neither column answers the other's question.

Distance and bearing

Every record on the report carries two more figures: how far it sits from the address, and in which direction. The distance is the first thing a reader checks and the first thing a reader misreads, so the document does not leave it to judgment. It bands every record the same way for every address.

The bearing is the compass direction from the roof to the record, and it matters as much as the distance. A storm moves, and a report upstream of a roof on the day the cells tracked toward it is a different piece of evidence from one the same distance downstream. The document prints the bearing so the reader can lay the records against the direction the storm travelled, which the map shows as a swath drawn from radar.

Some records carry no distance at all. The reviewed federal record files many events to a county zone rather than a point, and a county has no location to measure from. The document names the county and prints no distance, because there is none to print. A county record given a precise distance elsewhere is carrying a figure its source never stated, and how far away a report can be and still count is the subject of its own guide.

Independent sources

The last figure on each day is the count of independent sources that carried it. It is the figure a court will ask about, and it is a count of sources rather than of documents. A spotter's measurement and a radar signature over the same place on the same afternoon are two. A feed that republishes the spotter's line is still one. A second size computed from the same radar volume by a second vendor is one source with two names.

The count is what makes a day hard to argue with. A day carried by one source is one office's record that it happened; a day carried by three can be checked against itself, and when one of the three is challenged the other two are still standing. The report names each source on each record, with the public address of the line it came from, so the reader can pull the same entry from the federal archive and, where the archive offers it, order a certified copy.

The document also lists the sources that found nothing. A source searched and empty is part of the method: a report that found no record is only as strong as the list of places it looked, and without that list a clean report cannot be told from one that never searched.

What a report cannot say

A report ends with a statement of what it does not claim, and the reader should hold the document to it. The record attests four things: that hail occurred, on a date, of a size, at a stated distance from the address. Each of those is a fact about the public record, traceable to a federal line, and the document reproduces it as the source published it.

• It does not say the hail reached the roof. A record is a point; the inference from a point to a roof is the reader's.
• It does not say the hail damaged anything. Cause, extent and cost of damage are the inspection's, and a report that states no opinion on them has nothing there to be struck.
• It does not choose the date. It shows every day in the window and how each is attested, so a date can be chosen from the record and defended against it.
• It does not average, add or adjust. A measured size and a detected size are printed apart, and a day carried by both is attested twice, not attested at the larger figure.

Before any address is run, the free record pages answer the narrower question: which days within 10 miles of a city's centre carry hail on the record, and how many sources carried each. They measure from nothing and name no source for a property. The report is the document that does both, for the address, with its method stated on the page and a fingerprint a reader can recompute.

Questions this guide answers

How do I read a hail report

Start with the window searched and the date each record was filed on, then read the measured sizes apart from the detected ones, then the distance and bearing of each record from the address, then how many independent sources carried the day. Finish with the statement of what the document does not claim.

What does detected, not measured mean on a hail report

That the size was computed from radar for a grid cell rather than recorded on the ground. It is a ceiling on what the storm could have produced over that cell, printed in its own column with a qualifier and never added to a measured size.

What do distance and bearing mean on a hail report

How far the record sits from the address and in which compass direction. A report banded as primary is close enough to be evidence about the storm at the property; a contributing report says the storm was in the area that day. A county record carries no distance because it was filed to a zone, not a point.

Does a hail report prove the roof was damaged

No. A report establishes occurrence, date, size and distance from the public record. Whether a mark on the roof belongs to one of those days is the inspection's question, and a report that offers no opinion on damage has nothing there to be struck.