The word “derecho” has been in the meteorology lexicon for over a century, but it has recently garnered broader attention through a notable derecho event on June 29, 2012. This marked increase in interest of the term can actually be seen in Google Trends data, of which there is a screen shot below. I used the phrases “derecho” and “storm” in the Google Trends search in a basic attempt to filter out more common searches for the word derecho in its language of origin – Spanish.

Google Trends data on Google searches for the phrases "derecho" and "storm" from 2004 to 2013.

Google Trends data on Google searches for the phrases “derecho” and “storm” from 2004 to 2013.

The first small peak in 2009 is related to the May 8, 2009 derecho that swept from Kansas and Missouri eastward into portions of Kentucky and Tennessee. However, it is clear that the term was not receiving many Google searches until late June and July 2012 when there is a significant peak.

Since that time, there have been a number of convectively-induced wind events which have generated discussion amongst meteorologists about whether they could legitimately be classified as a derecho. Comparisons to the June 29, 2012 derecho, however, create an extremely high standard to meet for subsequent events. The June 29th derecho traveled for approximately 700 miles and produced over 1000 wind damage or severe wind gust reports (SPC Page, link here) – including 26 measured wind gusts of at least 33 m/s (65 kt, 74 mph). 4 of those measured wind gusts were at least 38 m/s (74 kt, 85 mph). Millions of people were without power for up to a week.

A radar mosaic composite showing the evolution and progression of the storms that produced the June 29, 2012 derecho that affected portions of the Great Lakes, Ohio River Valley, and Mid Atlantic. Graphic from the Storm Prediction Center web page.

A radar mosaic composite showing the evolution and progression of the storms that produced the June 29, 2012 derecho that affected portions of the Ohio River Valley and Mid Atlantic. Graphic from the Storm Prediction Center web page.

I won’t be discussing classification of specific events in recent years in this blog post. What I will attempt to do is to provide an overview of the derecho “criteria” that exist in the scientific literature, and then let others sort out how exactly they wish to apply that criteria. Of course, this is mainly of interest to meteorologists; the average person probably doesn’t mind what sort of label one applies to a storm, but rather is most likely just interested in how the storm impacts him or her.

So what exactly is a derecho? The definition was first explored in depth by Johns and Hirt (1987), hereafter JH87. The full text of the paper is available on the SPC website by clicking here, and the basic answer to that question is answered in the title, “Derechos: Widespread Convectively Induced Windstorms”. The literature review in the introduction provides some historical context:

This windstorm phenomenon was first recognized in the scientific literature in the latter part of the 19th century and was called a derecho (Hinrichs, 1888). Derecho (pronounced day-ray’-cho, for plural add “s”) is a Spanish word which can be interpreted as “straight ahead” or “direct”. […] From drawings produced by Hinrichs it appears that derechos are similar in scale to families of downburst clusters and to some of the larger downburst clusters (Fujita and Wakimoto, 1981). To avoid any ambiguity, we are specifically defining the term derecho to include any family of downburst clusters produced by an extratropical mesoscale weather convective system.
The term "derecho" was proposed by Dr. Gustavus Hinrichs in a paper published in the American Meteorological Journal in 1888. This figure is from that paper, and was posted on the Storm Prediction Center derecho website.

The term “derecho” was proposed by Dr. Gustavus Hinrichs in a paper published in the American Meteorological Journal in 1888. This figure is from that paper, and was posted on the Storm Prediction Center derecho website.

The remainder of the JH87 paper goes on to propose a specific criteria for derechos. Since that time, there have been a variety of papers that have discussed derechos and derecho climatology, but not all have used the same definition that was originally published in 1987. The reasons for that are described well in the first few pages of Coniglio and Stensrud (2004), hereafter CS04 – “Interpreting the Climatology of Derechos” – that you can find on the SPC website by clicking here. This probably contributes to the disagreement on whether or not certain damaging wind events can be classified as a derecho.

Areas of Agreement in Criteria

The following two JH87 criteria have remained basically unchanged in most publications since that paper was written, as described in CS04. These are quoted directly from JH87.

There must be a concentrated area of reports consisting of convectively induced wind damage and/or convective gusts > 26 m/s (50 kt, 58 mph). This area must have a major axis length of at least 400 km (250 miles).

Although there is no specific requirement for number of reports within a given area, one should look for a “concentrated area of reports”. JH87 considered both measured wind gusts of at least 50 knots and reports of wind damage. Measured wind gusts are only possible at discrete points, and therefore do not necessarily reflect conditions across the entire parent mesoscale convective system (MCS). Small-scale features within the parent MCS like microbursts and mesovortices can produce concentrated areas of much stronger winds, and these may not always be sampled directly by an anemometer. As a result, equivalent wind damage reports are considered as well when identifying windstorms that may meet derecho criteria. JH87 specifies that the reports should progress along an axis that stretches at least 400 km, which converts to 250 miles. The next criterion is straightforward.

The reports within this area must also exhibit a nonrandom pattern of occurrence. That is, the reports must show a pattern of chronological progression, either as a singular swath (progressive) or as a series of swaths (serial).

Nuances in Criteria

Although there are slight differences in how these are specifically defined, there is also agreement that the associated mesoscale convective system (MCS) must have spatial and temporal continuity, multiple swaths of damage must be part of the same MCS, and that no more than 2-3 hours (varies slightly) can elapse between successive wind damage reports. These minor distinctions are described in a table at the top of the second page in CS04 (link).

The main difference, therefore, appears to arise from the inclusion or exclusion of the JH87 requirement for “at least three reports of either F1 damage or wind gusts greater than 33 m/s (65 kt, 74 mph) separated by at least 64 km (40 miles) during the MCS stage of the event”.

Bentley and Mote (1998) did not include that criterion for 33 m/s wind gusts or equivalent damage (link), and therefore considered all wind damage swaths that extended at least 400 km, as long as they met the other criteria as described above. This is probably the broadest definition of derechos that exists in the literature, with JH87 applying a slightly more stringent requirement for significant wind gusts.

Recognizing the differences in how various derecho climatologies were constructed, CS04 parsed out derechos into three categories:

  • Low-end: meeting only the Bentley and Mote (1998) criteria.
  • Moderate: meets the JH87 requirement of 33 m/s wind gusts
  • High-end: at least three reports of wind gusts greater than 38 m/s (75 kt, 85 mph) separated by at least 64 km (40 miles), with at least two reports occurring during the MCS stage of the event.

Applying the requirement for “high-end derechos” creates a derecho subset with only the most intense convectively induced windstorms. This subset includes the likes of the June 29, 2012 derecho, the May 8, 2009 “Super-Derecho”, and the May 30-31, 1998 derecho that swept across the southern Great Lakes and produced multiple instances of wind gusts in excess of 100 mph. The authors do suggest that in the absence of measured wind gusts in excess of 38 m/s, equivalent damage may be used based on damage descriptions in Storm Data:

To determine the types of damage that should qualify for high-end classification, the damage descriptions in Storm Data that are associated with wind gusts of > 75 kt for a wide variety of reports are examined. The complete destruction of mobile homes (usually tossed far from their blocks), significant damage to well-constructed homes and businesses (often including several roofs blown off), and large swaths of trees being flattened within forested areas are examples of this type of damage. Miller and Johns (2000) find this type of damage in several intense derecho events.
A composite of hourly radar reflectivity showing the development and progression of the May 8, 2009 derecho.

A composite of hourly radar reflectivity showing the development and progression of the storms that produced the May 8, 2009 derecho. This derecho would qualify as a high-end derecho under the Coniglio and Stensrud (2004) criteria.

Coniglio and Stensrud examined derecho cases for the years from 1986 to 2001, a total of 16 years. They separated the cases into the low-end, moderate, and high-end categories, and each derecho case was classified in only one of those categories. 244 cases in total were identified, for an average of approximately 15.3 per year. A total of 171 would have met the original JH87 criteria (classified as moderate or high-end), for an average of 10.7 per year.

Even applying the most stringent criteria as outlined in CS04, the authors identified 55 high-end derecho events, which would work out to an average of 3.4 events per year based on the 1986-2001 period of record. These statistics suggest that applying more stringent criteria to classify a damaging wind event as a derecho would still yield at least a few cases per year, on average.

Conclusions

Even though we are hearing the word “derecho” more than ever, the statistics suggest that one would expect to see it applied close to a dozen times per year on average using the original JH87 criteria. There will obviously be differences in how people wish to parse derechos. The simplest solution to confusion about what constitutes a derecho is to be transparent about what sort of criteria one is applying. No matter which set of criteria that someone chooses to apply, they will probably have a basis in the scientific literature to support that choice. I’m not going to advocate for any particular set of criteria; I am merely pointing out that there are different criteria that exist and that no matter which criteria one chooses, derechos are more common on average than a once-per-year type of event.

Perhaps more importantly, impacts of a convective system producing widespread wind damage should be emphasized first and foremost over the particular classification of the event. This is far easier for people to understand and process, and gets the key message across in a less confusing fashion. The term derecho is obviously nuanced and has been applied differently over the past few decades. Therefore, simply saying “this is a derecho” or “this is not a derecho” may not tell the entire story, and will at the very least require more explanation for a general audience.


If you have a comment on this post, contact me through the Contact page on this website. This post represents my opinion or work only.