Retired senior lecturer in the Department of Meteorology at Penn State, where he was lead faculty for PSU's online certificate in forecasting.
By: Lee Grenci , 8:08 PM GMT on November 01, 2013
Lee's Opening Remark: I couldn't let the post from a reader pass without a response (copied and pasted below).
I am happy to use the reader's post as a stepping stone to teach (you know me...I'll be a teacher until I die). Many thanks to the reader for this teaching opportunity.
Post from Bruzote: As I see it, you're too willing to point the finger at others on TV while ignoring what's going on in your own back yard. Here in this particular blog entry regarding vertical wind shear, you have failed to address the elephant in the room that has become invisible to those familiar with tropical weather discussions. Wind shear is being referred to with units of speed! Wind shear is not a speed. It is not even a velocity. Wind shear is a velocity divided by a distance. That shear is referred to in knots is "shear" lunacy!
It's time, Lee, for you to crusade against that. If you're against bad and lazy science, what about this abuse of units that has taken place for...I don't even know how long! And that abuse takes place in the Wunderground blogs, even in this one. (Note the units on your wind shear diagram!) So, forget "Kelvins" vs "degrees Kelvin". You've got a bigger issue here.
I'm just starting to follow online hurricane discussions, and I'm genuinely frustrated to realize that having an understanding of science can be useless for understanding a phenomena when others describing the phenomena won't use proper units to describe it. Here in these blogs and elsewhere (that's you, NOAA and NWS!), I seriously cannot make heads or tails of the degree of wind shear being discussed. Not one commenter or blogger seems to use sensible units! I'm not being picky, I'm seriously unable to understand what is meant. When a communication issue reaches that point, perhaps the "experts" need to look in the mirror and ask whose fault that is. Surely, the burden should not be on readers to decipher some lingo that is factually incorrect. The facts should be presented - factually
It is inexcusable for a supposedly expert community to use incorrect units because it is easier than using correct units. Would you have police issuing you a ticket for going 15 miles? How about if a grocer charged you $3.50 for a pint of milk because he was too lazy to charge $3.50 per gallon? If these situations are unacceptable, then so is using a speed unit for wind shear.
I am sure at least one person out there is dying to tell me how to interpret a shear of "knots". The fact is, I don't care two wits about a convention used by lazy people, or what the "kool kids" or the "in-the-know-group" prefer to use. Facts are facts and bad units are bad units. Wind shear is a velocity change that occurs over a distance. Describe it that way. If you tell me wind shear is 20 knots, you've left out two critical pieces of information. Over what distance and in what direction?
PLEASE - address this issue. Lobby people here AND at the NWS (and AMS, WMO, etc.?) to start using factually proper units. Just let people have something correct they can use. I don't care if someone calls a wind shear "20 knots NW per -500 mb" or ".008/s @ 315 degrees". Just please provide something usable. I can't usefully employ incorrect information.
Lee Responds: One of my tenets that I try my best to follow is to not answer posts or e-mails that I deem a bit inflammatory right away. Delaying my response allows me to think about my answer and prevents me from reacting emotionally. It's been several weeks since this reply appeared under my blog, so here goes.
For starters, there are many "misnomers" in the world of science. For example, kinetic energy is often quoted as 0.5 v^2 (v^2 means "velocity squared). Yet this quantity should, technically speaking, be called specific kinetic energy. I admit, however, that pointing the finger at such "misnomers" in science is a lame excuse. As you suggested, the units of vertical wind shear are 10^-4 seconds^-1. It's pretty clear to me that such units aren't very intuitive, so your preference for using these units and admonishing me to "Just please provide something usable" is rather contradictory, in my opinion. I find nothing intuitive about 10^-4 seconds^-1. I'll also wager that weather forecasters, as well as the public, find "knots" to be much more intuitive than 10^-4 seconds^-1. Any bets?
The black arrow represents the magnitude (in knots) and direction (314 degrees) of the bulk vertical wind shear vector in a specified layer of air. The green vector indicates a wind at the bottom of the layer blowing from the southwest (250 degrees) at 10 knots. The blue vector represents the wind at the top of the layer blowing from the northwest (300 degrees) at 40 knots. Courtesy of the Penn State online certificate program in weather forecasting.
"Knots," on the other hand, are much more intuitive, in my opinion. Before I go any further, I point out that you're referring to vertical wind shear as defined in the AMS glossary, which involves partial (local) derivatives. I'm talking about bulk vertical wind shear, which, in my book, is a completely different concept. Indeed, I'm talking about apples, and you're referring to oranges. For the record, bulk vertical wind shear is defined as the vector difference between the winds at the top and bottom of a specified layer of air (see schematic above). It's shortcoming is that it misses any variation in winds between the bottom and top of the layer of air, but, as you will readily see, forecasters at the Storm Prediction Center and the National Hurricane Center use bulk vertical wind shear to formulate their forecasts. Indeed, "knots" have become the standard units for bulk vertical wind shear across our profession...the Storm Prediction Center, the National Hurricane Center (see precursor condition #6), and university researchers (for example, the Cooperative Institute of Satellite Meteorological Studies at the University of Wisconsin).
I've written several blogs using the vector difference between the winds at the top and bottom of a layer of air to represent the bulk vertical wind shear. As it turns out, this approach is exactly the method used to calculate the thermal wind, which is the vertical shear of the geostrophic wind (the idealized wind resulting from the balance of the Coriolis and pressure-gradient forces). On Page 69 of the second edition of An Introduction to Dynamic Meteorology (which I consider to be one of the best textbooks for dynamic meteorology), author James Holton expresses the thermal wind as the vector difference between the geostrophic wind at the top of a specified layer of air and the geostrophic wind at the bottom of the layer. So, my friend, there is a scientific precedent for using a vector difference to compute a bulk vertical wind shear and "knots" to describe the magnitude of this bulk vertical wind shear.
The HWRF (Hurricane Weather Research and Forecast System) 12Z model analysis of bulk vertical wind shear between 850 mb and 200 mb on November 1, 2013. The magnitude of bulk shear is color-coded in meters per second (convert to knots) and wind barbs designate the direction of the bulk-shear vector. Courtesy of Penn State.
On a more fundamental level, the magnitude of vertical wind shear can be expressed as delta v (the difference in the magnitudes of the wind vectors at the top and bottom of a specified layer of air) divided by delta p, the difference between the pressure at the top and bottom of the layer. Given that the difference in pressures is always the same for a specified layer of air (between 850 mb and 200 mb, for example...see today's chart for Tropical Depression Eighteen-E above), delta p is constant. Thus, you can get a reasonable proxy for the magnitude of the vertical wind shear in the specified layer of air simply from the magnitude of the numerator (the difference in the magnitude of the wind vectors at the top and bottom of the layer).
The special 17Z skew-T at Norman, OK, on May 20, 2013. I circled the hodograph. The length and curvature of the hodograph indicated strong vertical wind shear capable of supporting tornadic supercells. This was the day that the violent twister tore through Moore, OK. Courtesy of SPC.
As a professional weather forecaster, I frequently used bulk vertical wind shear, primarily in the context of predicting severe thunderstorms, and, secondarily, tropical cyclogenesis. With regard to forecasting severe weather, my recollection is that the depth of the layer over which vertical wind shear was considered in an operational setting has more or less always been ignored. Indeed, the widespread adoption of hodographs for assessing the variation of winds with altitude essentially put the depth of the layer of air "out of business," at least from the standpoint of operational forecasters. For starters, a hodograph is a plot of the curve connecting the tips of the wind vectors (speed and direction) determined by a radiosonde at mandatory and significant levels. For example, check out the hodograph above based on the radiosonde data from Norman, Oklahoma, on May 20, 2013 (the day of the Moore tornado). Focusing your attention on the hodograph, it's pretty clear that this plot provides forecasters with clues about how the wind changes with altitude. In other words, hodographs give forecasters information about vertical wind shear. To gain further insight, I note that the vector difference between any two points on the hodograph (delta v) is simply a vector connecting the two points. When the magnitude of this vector is long, forecasters know that there's large vertical wind shear between the top and bottom of the specified layer. In other words, a relatively long hodograph (like the one at Norman, Oklahoma, on May 20, 2013) implies large vertical wind shear.
SPC forecasters looking at the length and curvature of this hodograph no doubt knew that supercells were likely to be initiated later that tragic afternoon. Are you really advocating that the use such tools is wrong because SPC forecasters didn't take into account the depth of the layers where "vertical wind shear" was large? I asked SPC forecasters about this issue and one forecaster told me that "the vector difference was adopted as a short hand so as not to be pedantic."
The 12Z Rapid Refresh model analysis of bulk vertical wind shear between the surface and an altitude of six kilometers. Isotachs are expressed in knots, and barbs indicate both the magnitude (in knots) and direction of the shear vector. Courtesy of SPC.
Don't get me wrong...I am quite aware that, if pressure is used in the finite-differencing scheme to calculate the denominator in the expression for vertical wind shear, that the depth of the layer can change, depending on mean temperature in the layer. So SPC uses layers such as the surface to six kilometers to calculate the bulk vertical wind shear (see this morning's 12Z Rapid Refresh model analysis of bulk shear above). Indeed, research has shown that the mode of relatively discrete thunderstorms is tied to the magnitude of the bulk vertical wind shear in this layer. Ultimately, I believe SPC forecasters view bulk vertical wind shear (the vector difference between winds at the top and bottom of a specified layer of air) a proxy for vertical wind shear. In short, bulk vertical wind shear is akin to the earlier use of "delta T" (change in temperature from the bottom to the top of a specified layer of air) to approximate lapse rates. More specifically, SPC forecasters recognize that a 700-500 mb delta T of 20 degrees Celsius is pretty unstable, even though they sometimes don't know how this delta T compares to the actual layer lapse rate. See what I mean?
There's some history to this practice by forecasters at the Storm Prediction Center. Calculating the "true" vertical wind shear "on the fly" required an extra mathematical step (something not easily done before the days of widespread, vertically-gridded model data). Moreover, given the relatively coarse network of radiosonde observations, I believe that old timers like me would agree that variations in the depth of the layer over which vertical wind shear was calculated wasn't worth the effort, especially during critical times when severe weather was looming.
Tropical meteorologists also typically refer to the vector difference between winds at the top and bottom of a specified layer of air, rather than using the gradient of the vertical wind. Like the prediction of severe thunderstorms, there's also some history behind the use of bulk vertical wind shear in matters relating to tropical cyclogenesis.
Given the vast horizontal distances between rawinsonde stations in the Tropics, winds were typically analyzed at only a few levels. In the 1970s, for example, 200 mb was a popular pressure altitude because of the advent of cloud-tracked winds from satellites and observations from jet aircraft. At the lower altitudes, the network of surface observations allowed tropical analysts to roughly estimate the winds at the top of the boundary layer, which they referred to as the gradient-wind level. It's not surprising, then, that hurricane forecasters would routinely look at the vector difference in winds between 200 mb and the gradient-wind level to get a sense for the bulk change in winds in this rather deep layer. Of course, as the accuracy of computer analyses increased, forecasters could get a sense for bulk shear in layers of air between mandatory pressure levels. But the practice of using the vector difference between winds at the top and bottom of a specified layers of air had already been established.
The 0915Z enhanced IR image of Major Hurricane Raymond on October 21, 2013. Courtesy of NOAA's Regional and Mesoscale Meteorology Branch (RAMMB).
Indeed, bulk vertical wind shear has become a standard forecasting tool. For example, NOAA's Regional and Mesoscale Meteorology Branch (RAMMB) routinely calculates the bulk vertical shear for the layers, 850 mb to 500 mb and 850 mb to 200 mb. Check out (below) the time-series graphs of these area-averaged bulk shears for Major Hurricane Raymond (the 09Z enhanced IR satellite image of Raymond off the west coast of Mexico on October 21, 2013, is shown above). For the record, the magnitudes of the vector differences for the two layers are plotted in red, and the directions of the bulk shear vectors, expressed in degrees, are plotted in blue.
The time-series plots of bulk vertical wind shear between 850 mb and 200 mb (top) and 850 mb and 500 mb (bottom). The magnitude (expressed in knots) of the bulk shear is plotted in red, and the direction of the bulk-shear vector, expressed in degrees, is plotted in blue. Courtesy of NOAA's RAMMB.
Research continues to flourish is the area of bulk shear and tropical cyclones. Indeed, the abstract (below) of an upcoming paper by Dr. Chris Velden and John Sears of the Cooperative Institute for Satellite Meteorological Studies (University of Wisconsin) will show that a simple vector difference between the winds at the top and bottom of a specified layer of air (in this case, 850 mb to 200 mb) can sometimes leave a lot to be desired (it will be very helpful to see the research that proves our general impressions about the vulnerability of simple bulk shear over a deep layer). I note that bulk vertical wind shear ("VWS" in the abstract) serves as a predictor in the SHIPS model. The take-away here is that the simple approximation (vector difference between winds at two levels) can lead to significant misinterpretations of the shear environments of tropical cyclones.
The abstract of an upcoming paper written by Chris Velden and John Sears of the Cooperative Institute of Meteorological Satellite Studies.
Any way you slice it, the profession is replete with examples of equating the bulk vertical wind shear as a vector difference between winds at the top and bottom of a specified layer of air (and knots or meters per second as the units for the magnitude of this bulk shear). There is both history and precedence to support this operational forecasting practice.
I have the advantage of having been an operational forecaster and being old enough to have gained a sense of history for the development of bulk vertical wind shear in both the mesoscale and tropical forecasting communities.
Bruzote...you made a relevant point about the units...this controversy arose frequently while I was teaching at Penn State. Next time, just try to keep in mind that there's a big difference between vertical wind shear (the gradient of the vertical wind, as defined in the AMS glossary) and bulk vertical wind shear as defined and implemented by the forecasting community (the vector difference between the wind vectors at the top and bottom of a specified layer of air). Deal?
Dusting myself off from being under the bus, :-)
Comments will take a few seconds to appear.