The hook echo

A hook echo is a hook-shaped extension of a thunderstorm’s radar reflectivity, usually on the right-rear part of the storm relative to its motion. That is close to word for word the National Weather Service glossary definition, which adds that a hook is often associated with a mesocyclone and indicates conditions favorable for tornado development.

Note the verbs: associated with, favorable for. The glossary does not say a hook means a tornado, and neither should you. The tornadoes on radar guide ranks the hook as the weakest of the three tornado signatures. This page explains why, and what the shape is actually telling you.

Where the hook comes from

A supercell is organized around a rotating updraft, the mesocyclone. Precipitation that forms in the updraft does not simply fall where it formed. Some of it is carried downwind into the forward flank, and some descends on the back side of the updraft, where the rear-flank downdraft (RFD) wraps around the rotation. Drawn around the circulation, that rear-side precipitation curls into an appendage. On a low-elevation reflectivity scan, the appendage is the hook, and the comparatively echo-free notch it partly encloses marks the inflow and updraft region.

The short version, the one in the pillar guide, holds: the hook is precipitation being wrapped, not the tornado itself. The research picture is more detailed than “rain blown in a circle.” Fujita (1958), re-analysing the 1953 Illinois radar film in the Journal of Meteorology, framed the hook as precipitation advected around a rotating storm. Later work looked at the vertical dimension. Rasmussen, Straka, Gilmore and Davies-Jones (2006, Weather and Forecasting) defined the descending reflectivity core: a reflectivity maximum hanging from the rear side of the echo overhang above the weak-echo region, which descends toward the ground. In every tornadic case in their preliminary sample, a descending core preceded tornado formation, though the authors state the sample was far too small to say how often that happens in general.

The practical consequence: a hook is a low-level feature built by processes happening aloft. What you see at the lowest tilt is the bottom of a three-dimensional structure, and the tilts above it are where it begins.

The first one on film: Champaign, 9 April 1953

The hook echo has a precise origin story, documented by the Illinois State Water Survey. Late on 9 April 1953, Donald Staggs of the Survey staff was testing repairs to a radar the Survey ran near Champaign, Illinois, for rainfall research, not tornado detection. He noticed an unusual appendage on a nearby storm, kept the radar running and recorded the scope on 35 mm film. The storm produced a tornado north of Champaign, and damage reports along its path later matched the echo.

Glenn Stout and Floyd Huff published the case that year as “Radar records Illinois tornadogenesis” in the Bulletin of the American Meteorological Society (vol. 34, pp. 281–284). A fuller study by Huff, Hiser and Bigler followed in 1954 as Survey Report of Investigation 22. The Survey describes it as the first documented case of a tornado detected by radar. That was a reflectivity-only radar, and for years afterward reflectivity shape was the main tornado clue radar offered, which is part of why the hook still carries more weight in popular imagination than it deserves.

How reliable is a hook?

The most-cited attempt to measure this is Forbes (1981), “On the reliability of hook echoes as tornado indicators” in Monthly Weather Review. Forbes examined radar echoes during the 3 April 1974 outbreak. Two findings pull in opposite directions:

• When a true hook appeared, it was usually right. Hook-shaped echoes carried a false alarm rate of 16% as indicators of a tornadic storm.
• But hooks were relatively rare. A looser category Forbes called “distinctive”, meaning a marked appendage on the right-rear flank, or a spiral, comma or line-echo-wave-pattern shape, caught 65% of the tornadic echoes.

Read those together and the hook is a signature with decent precision and poor coverage: good when it is there, absent far too often to rely on. That study is one outbreak, scanned by pre-Doppler radars, so treat the figures as a historical benchmark rather than a rate for modern data. Training material from the COMET program makes the same point from the other direction. It calls the hook one of the most commonly misdiagnosed severe-storm signatures, often confused with appendages on multicell storms and flanking lines. It also notes evidence that a hook is no better at identifying a tornadic storm than the presence of a mesocyclone.

Why many tornadoes show no hook

• High-precipitation supercells. The NWS Spotter’s Field Guide notes that in HP storms, precipitation often surrounds the updraft and may hide it completely. On radar the notch fills in and the hook blurs into a broad, bright mass, often described as kidney-bean shaped. The rotation is still there; the reflectivity shape is gone. The supercell guide covers the classic, HP and LP spectrum.
• Linear systems. Tornadoes in squall lines and bow echoes develop along the leading edge of a line, usually without the isolated rear-flank appendage that makes a hook. See squall lines, bow echoes and derechos.
• Geometry. The beam rises with distance. Far from the radar, the lowest scan can sit above the level where the hook is best defined. A storm can also be too small, relative to the beam’s widening footprint at range, for its appendage to resolve.
• Timing. Hooks evolve. One can wrap up, fill in or wash out between volume scans, and a tornado can occur during a phase when the reflectivity shape is ambiguous.

Hook, couplet, debris: how they relate

Think of the three signatures as answers to different questions about the same place in the storm.

The hook tells you where to look. The tightest part of the circulation usually sits near the tip of the hook, at the junction of the rear-flank and inflow regions, so that is where you check velocity for strong inbound beside strong outbound. If a couplet is there, is tight, low and persists across tilts and scans, you have a real detection. If dual-polarization data then show a tornado debris signature at the same spot, you have evidence of a tornado lofting debris, which also means damage has already occurred. A reflectivity blob at the hook tip, the “debris ball,” can suggest that; correlation coefficient is what confirms it.

The failure mode runs in both directions. Watching reflectivity alone, you will over-warn on hooks that never produce and miss rain-wrapped tornadoes that never show one. That is why reading velocity alongside reflectivity matters, a point the how to read weather radar guide makes in general terms.

WeatherStory Radar reads full-resolution Level II data from NOAA’s public archive, so the hook, the velocity field and correlation coefficient come from the same volume scan; see what it does.