242

WEATHER AND FORECASTING

VOLUME 21

On the Classification of Vertical Wind Shear as Directional Shear versus Speed Shear
PAUL MARKOWSKI

AND

YVETTE RICHARDSON

Department of Meteorology, The Pennsylvania State University, University Park, Pennsylvania

(Manuscript received 31 December 2004, in final form 30 June 2005)

ABSTRACT
Vertical wind shear is commonly classified as directional or speed shear. In this note, these classifications are reviewed and their relevance discussed with respect to the dynamics of convective storms. In
the absence of surface drag, storm morphology and evolution only depend on the shape and length of a
hodograph, on which the storm-relative winds depend; that is, storm characteristics are independent of the
translation and rotation of a hodograph. Therefore, traditional definitions of directional and speed shear are
most relevant when applied to the storm-relative wind profile.

1. Introduction
The Glossary of Meteorology defines vertical wind
shear as the local variation of the wind vector or any of
its components in a given direction (Glickman 2000).
Vertical shear is a required presence in the geostrophic
wind profile in a hydrostatic, baroclinic atmosphere.
The relationship between the vertical shear of the geostrophic wind and the temperature gradient is given by
the thermal wind relation:

1
R
vg

k p k pT,
p
p f
fp

where p is the pressure (used here as the vertical coordinate), vg (1/f ) k p is the geostrophic wind,
is the geopotential height, R is the gas constant for dry
air, pT is the temperature gradient on a pressure surface, k is the unit vector in the vertical, and f is the
Coriolis parameter.
Vertical wind shear also can be present in the absence of large-scale baroclinity. For example, friction
plays a role in creating vertical wind shear within the
boundary layer, owing to the fact that the effects of
friction decrease with height, becoming negligible at
the top of the boundary layer. An analytical wind pro-

Corresponding author address: Dr. Paul Markowski, Dept. of

Meteorology, The Pennsylvania State University, 503 Walker
Bldg., University Park, PA 16802.
E-mail: pmarkowski@psu.edu

2006 American Meteorological Society

file containing vertical wind shearthe Ekman spiralcan be derived by assuming the presence of
boundary layer friction and no ambient baroclinity
(e.g., Holton 2004, 127129). Large accelerations of the
horizontal wind also can contribute to vertical wind
shear in ways not predicted by the thermal wind relation given in (1). For example, near jet streaks or rapidly moving and/or intensifying cyclones, the observed
vertical wind shear can differ substantially from the
geostrophic vertical wind shear. A more lengthy discussion of the processes contributing to vertical wind shear
is provided by Doswell (1991).
Although the definition of vertical wind shear makes
no distinction regarding whether the shear is associated
with variations in wind direction and/or speed with
height, the literature abounds with classifications of
vertical wind shear as being either directional or
speed shear. However, a survey of the literature indicates that the classifications have not always shared
the same meaning from one study to the next. As classroom instructors, we have found the possible ambiguity
in the wind shear classifications to be a source of confusion on occasion. The purpose of this note is to provide some clarification on wind shear classifications.
Although vertical wind shear exerts important controls
on many atmospheric phenomena (e.g., boundary layer
rolls, lake-effect snowbands, gravity waves), our forthcoming discussion primarily will focus on the relevance
of wind shear classifications to the structure and evolution of convective storms.

APRIL 2006

NOTES AND CORRESPONDENCE

2. Popular definitions of directional shear and

speed shear
Probably the most popular usage of the term directional wind shear has referred to the changing of the
angle of the wind velocity vector with height, expressed
in degrees per vertical distance. Speed shear correspondingly has been defined as the variation of wind
speed with height. To the best of our knowledge, these
are the only definitions that have been used outside of
the severe storms community. For example, these definitions have been used by investigators of lake-effect
snow (e.g., Cooper et al. 2000; Steenburgh et al. 2000),
boundary layer convection (e.g., Balaji and Clark 1988;
Redelsperger and Clark 1990; Weckwerth et al. 1997;
Atkins et al. 1998; Kristovich et al. 1999), gravity waves
(e.g., Kim and Mahrt 1992; Hauf 1993; Shutts 2003),
and tropical convection (e.g., Halverson et al. 1999;
Rickenbach 1999). Even within the severe storms community, these definitions frequently have been employed in the past (e.g., Schlesinger 1975, 1978; Johns
1984; Giordano and Fritsch 1991; Hagemeyer and
Schmocker 1991).
On the other hand, an alternative meaning of directional shear also has been used or implied within the
context of severe local storms, whereby directional
shear is the change in direction of the wind shear vector,
rather than the horizontal wind vector, with height
(e.g., Cotton and Anthes 1989, 515520; Houze 1993,
289291). Moreover, the terms unidirectional or
one-directional shear commonly have been used to
describe a shear vector that does not change direction
with height, in contrast to a directionally varying
shear vector. These classifications of the vertical wind
shear, which describe whether a hodograph is curved
(directional or directionally varying shear) or straight
(unidirectional shear), have been used extensively
within the severe storms community (e.g., Toutenhoofd
and Klemp 1983; Peterson 1984; Klemp 1987;
Wakimoto et al. 1998; Cai and Wakimoto 2001), especially in modeling studies of thunderstorms (e.g.,
Klemp and Wilhelmson 1978a,b; Wilhelmson and
Klemp 1978; Weisman and Klemp 1982, 1984, 1986).

3. Importance of hodograph shape and length

It probably is not coincidental that the severe storms
community has had to develop terminology to describe
the variation in the direction of the wind shear vector
with height. The seminal three-dimensional numerical
simulations of supercell thunderstorms in the late 1970s
and early 1980s (Schlesinger 1975, 1978; Klemp and
Wilhelmson 1978a,b; Wilhelmson and Klemp 1978;

243

Weisman and Klemp 1982, 1984) demonstrated the dynamical importance of the shape and length of the
hodograph characterizing the storm environment,
rather than the orientation of the hodograph relative to
the ground. Only the shear and its variation with height
have dynamical importancethe variation of the
ground-relative wind direction (or speed) with height
has no dynamical relevance to thunderstorm structure,
at least for simulations in which a free-slip lower
boundary is prescribed (almost universally the case
throughout the 1970s and 1980s).1 Indeed, some recently developed storm motion predictors have recognized the approximate invariance of gross storm dynamics to hodograph rotation and translation (Rasmussen and Blanchard 1998; Bunkers et al. 2000).
A rotation or translation of a hodograph in the real
atmosphere is complicated somewhat by surface drag,
which tends to extend a hodograph toward the origin
[(u, ) (0, 0) m s1, where u and are the zonal and
meridional wind components, respectively]; that is, the
surface layer often contains strong vertical wind shear
due to the requirement that wind speeds approach zero
at the ground. Stated another way, because of surface
drag, a hodograph generally cannot be rotated or translated without an attendant modification of the
hodograph length and shape. It currently is not known
what importance the vertical wind shear within this
relatively shallow layer may have on storm dynamics,
principally because this layer has not been well resolved
in past numerical simulations. In spite of the complications arising from surface drag, the gross characteristics
of convective storms are largely independent of the
magnitude and degree of veering of the ground-relative
winds as long as the hodograph length and shape outside this layer are held constant.
The fact that it is the hodograph shape and length
that dictate the structure and evolution of a convective
storm is related to the fact that hodographs dictate the
characteristics of the storm-relative wind profile. The
importance of storm-relative winds has been stressed
since the early days of severe storms research (e.g.,
Browning 1964), as well as in climatological (e.g., Maddox 1976; Darkow and McCann 1977) and theoretical
(e.g., Davies-Jones 1984) studies of severe storms environments. Davies-Jones (1984) applied the traditional
definitions of directional and speed shear to stormrelative wind profiles. This correctly implies that it is
the hodograph shape and length that matter; a given
1

It is entirely possible that the profile of ground-relative winds

could be relevant in considering the likelihood of thunderstorm
initiation. A detailed discussion of this matter is beyond the scope
of this paper.

244

WEATHER AND FORECASTING

hodograph will be associated with the same amount of

directional and speed shear of the storm-relative wind
regardless of how the hodograph is oriented, assuming
that storm motions are affected only by internal storm
dynamics and not by external forcings, such as terrain
or an inhomogeneous environment.
On the other hand, hodograph translation and rotation lead to changes in the variation of ground-relative
wind speed and direction with height. Directional and
speed shear defined by such ground-relative wind variations ultimately depend on the mean wind velocity,
which governs hodograph translation with respect to
the origin. One would not expect such quantities to be
dynamically relevant. Directional and speed shear defined by ground-relative wind variations also depend on
the orientation of the thermal wind, which largely controls the orientation of a hodograph. The hodographs
shown in Fig. 1, which very roughly have similar shapes,
span a variety of mean wind velocities and deep-layer
shear orientations. Hodographs that are confined to the
first quadrant (i.e., u, 0 at all levels) are commonly
observed in eastern United States severe weather outbreaks [Figs. 1a and 1b; e.g., Kaplan et al. (1998)],
whereas hodographs commonly span the first and second quadrants (i.e., 0 at all levels, but u changes
sign from negative to positive in the lower troposphere)
in severe weather outbreaks in the U.S. Great Plains
region [Fig. 1d; e.g., Maddox (1976)]. Hodographs occasionally span more than two quadrants, for example,
in northwest flow events [Fig. 1c; e.g., Johns (1982),
(1984)]. Ground-relative veering is smaller when hodographs are confined to a single quadrant compared to
when hodographs span multiple quadrants, yet the
magnitude of storm-relative wind speeds and veering is
largely independent of hodograph orientation and position with respect to the origin (storm motions tend to
be faster when hodographs are confined to a single
quadrant). In the specific examples of Fig. 1, the cases
with the smallest ground-relative wind veering (Figs. 1a
and 1b) actually are associated with the largest stormrelative wind veering.
The arguments presented above are further clarified
by the idealized hodographs depicted in Fig. 2. Hodographs A and A are identical in length and shape (both
are straight), but hodograph A has been shifted by
adding 5 m s1 (10 m s1) to the zonal (meridional)
wind speeds of the points composing hodograph A.
Hodographs A and A both have unidirectional shear
as defined by the variation of the wind shear vector
with height (the shear is westerly at all levels), but have
very different degrees of directional wind shear if directional shear is defined as the angular turning of the
ground-relative wind vector with height (0 of veering

VOLUME 21

FIG. 1. (a)(d) Hodographs observed in relative proximity to

outbreaks of supercell thunderstorms having a wide variety of
mean wind velocities and deep-layer shear (thermal wind) orientations. Labels along the hodographs indicate heights above
ground level in km. Although wind data in only the lowest 6 km
are plotted [the winds are missing above 5 km in (b)], this does not
imply that the winds above 6 km are unimportant. The symbols
indicate the approximate average storm motions observed on
each day. Profiles of ground-relative (g-r) and storm-relative (s-r)
winds (half barb, 2.5 m s1; full barb, 5 m s1; flag, 25 m s1) are
displayed to the right of each hodograph.

in the case of hodograph A; 90 of veering in the case

of hodograph A). Likewise, hodographs B and B are
identical in length and shape (both are half-circles), but
hodograph B has been shifted by adding 20 m s1

APRIL 2006

NOTES AND CORRESPONDENCE

245

FIG. 2. Idealized (top left) straight and (bottom left) curved hodographs and their corresponding (top
right) ground-relative (g-r) and (bottom right) storm-relative (s-r) wind profiles (half barb, 2.5 m s1; full
barb, 5 m s1; flag, 25 m s1). Labels along the hodographs indicate heights above ground level in km.
The symbols indicate the storm motions predicted by the Bunkers et al. (2000) technique. Hodographs
A and A are identical, but hodograph A has been shifted by adding 5 m s1 (10 m s1) to the zonal
(meridional) wind components of the points composing hodograph A. Hodographs B and B are identical, but hodograph B has been shifted by adding 20 m s1 (8 m s1) to the zonal (meridional) wind
components of the points composing hodograph B. The storm-relative wind profiles are identical for the
A and A hodographs and the B and B hodographs.

(8 m s1) to the zonal (meridional) wind speeds of the

points composing hodograph B. Hodographs B and B
have identical amounts of directional shear if directional shear is defined as the veering of the shear vector
with height or veering of the storm-relative winds with
height. But, if directional shear is defined as the veering
of the ground-relative wind vector with height, then
hodographs B and B have very different degrees of
directional shear (180 of veering in the case of
hodograph B; 45 of veering in the case of hodograph
B). Furthermore, hodograph B has no speed shear, yet
hodograph B has 25 m s1 of speed shear, if speed
shear is defined as the variation of ground-relative wind
speed with height. It also is easily shown that the storm-

relative wind profiles (and thus the implied dynamics)

are identical within each pair of hodographs, assuming
that the storm motion is the same function of the environmental wind profile (see the storm-relative wind
profiles in Fig. 2).

4. Summary
Vertical wind shear routinely has been categorized as
directional shear or speed shear, but the criteria used in
making such distinctions have varied throughout the
literature. We have attempted to summarize the range
of interpretations permitted by these shear classifications, as well as provide some clarification regarding
the dynamical implications of these classifications.

246

WEATHER AND FORECASTING

Wind shear has been described as directional most

often, and almost exclusively outside of the severe local
storms context, when the ground-relative wind velocity
vector turns with height. With regard to severe local
storms, this terminology also has been applied to the
storm-relative winds, and in a few cases, to describe
environments in which the wind shear vector turns with
height, as in the case of a curved hodograph. To avoid
potential confusion with the traditional terminology,
curved hodographs have been described in most studies
as having directionally varying wind shear, whereas
straight hodographs often have been described as having unidirectional or one-directional shear. Correspondingly, the term speed shear traditionally has been
used to denote ground-relative winds that increase in
speed with height. As is the case with the directional
shear terminology, speed shear also has been used in
reference to storm-relative winds.
In this review it has been shown that vertical variations in ground-relative wind speed and direction
change as a given hodograph is rotated or translated
with respect to the origin; that is, the magnitude of
ground-relative wind direction and the speed variations
with height depend on the orientation of the thermal
wind (which largely governs hodograph orientation)
and mean wind (which governs hodograph position
with respect to the origin). Thus, definitions of directional and speed shear that depend on vertical variations of the ground-relative winds are not the most dynamically relevant. Conversely, the traditional definitions of directional and speed shear applied to the
storm-relative winds (or hodographs described as containing directionally varying or unidirectional shear)
are insensitive to hodograph rotation and translation
and are directly relevant to storm dynamics.
Finally, when the effects of surface drag are considered, a hodograph generally cannot be rotated or translated while retaining its original shape and length. In
other words, changes in the mean wind or mean shear
orientation unavoidably cannot be considered independently of a hodograph that includes winds within the
surface layer, all the way to the ground. No systematic
study of the effects of near-surface wind shear changes
has been undertaken yet.2 We believe that further research into the influence of near-surface wind shear and
surface drag on convective storms is well warranted,

The finding in a recent study (Markowski et al. 2003) of

ground-relative wind speeds being a relatively skillful discriminator between tornadic and nontornadic supercell thunderstorms
(speeds tended to be larger in tornadic supercell environments)
might be an indication of the importance of surface drag and
near-ground wind shear.

VOLUME 21

especially given the computational capabilities of the

present day.
Acknowledgments. This paper was motivated by
many difficult questions asked by inquisitive students in
the mesoscale meteorology class at Penn State University. We thank the three anonymous reviewers for their
constructive reviews.
REFERENCES
Atkins, N. T., R. M. Wakimoto, and C. L. Ziegler, 1998: Observations of the finescale structure of a dryline during VORTEX 95. Mon. Wea. Rev., 126, 525550.
Balaji, V., and T. L. Clark, 1988: Scale selection in locally forced
convective fields and the initiation of deep cumulus. J. Atmos. Sci., 45, 31883211.
Browning, K. A., 1964: Airflow and precipitation trajectories
within severe local storms which travel to the right of the
winds. J. Atmos. Sci., 21, 634639.
Bunkers, M. J., B. A. Klimowski, J. W. Zeitler, R. L. Thompson,
and M. L. Weisman, 2000: Predicting supercell motion using
a new hodograph technique. Wea. Forecasting, 15, 6179.
Cai, H., and R. M. Wakimoto, 2001: Retrieved pressure field and
its influence on the propagation of a supercell thunderstorm.
Mon. Wea. Rev., 129, 26952713.
Cooper, K. A., M. R. Hjelmfelt, R. G. Derickson, D. A. Kristovich, and N. F. Laird, 2000: Numerical simulations of transitions in boundary layer convective structures in a lake-effect
snow event. Mon. Wea. Rev., 128, 32833295.
Cotton, W. R., and R. A. Anthes, 1989: Storm and Cloud Dynamics. Academic Press, 883 pp.
Darkow, G. L., and D. W. McCann, 1977: Relative environmental
winds for 121 tornado bearing storms. Preprints, 10th Conf.
on Severe Local Storms, Omaha, NE, Amer. Meteor. Soc.,
413417.
Davies-Jones, R. P., 1984: Streamwise vorticity: The origin of updraft rotation in supercell storms. J. Atmos. Sci., 41, 2991
3006.
Doswell, C. A., 1991: A review for forecasters on the application
of hodographs to forecasting severe thunderstorms. Natl.
Wea. Dig., 16, 216.
Giordano, L. A., and M. Fritsch, 1991: Strong tornadoes and flashflood producing rainstorms during the warm season in the
mid-Atlantic region. Wea. Forecasting, 6, 437455.
Glickman, T. S., Ed., 2000: Glossary of Meteorology. 2d ed. Amer.
Meteor. Soc., 855 pp.
Hagemeyer, B. C., and G. K. Schmocker, 1991: Characteristics of
east-central Florida tornado environments. Wea. Forecasting,
6, 499514.
Halverson, J. B., B. S. Ferrier, T. M. Rickenbach, J. Simpson, and
W.-K. Tao, 1999: An ensemble of convective systems on 11
February 1993 during TOGA COARE: Morphology, rainfall
characteristics, and anvil cloud interactions. Mon. Wea. Rev.,
127, 12081228.
Hauf, T., 1993: Aircraft observations of convection waves over
southern GermanyA case study. Mon. Wea. Rev., 121,
32823290.
Holton, J. R., 2004: An Introduction to Dynamic Meteorology. 4th
ed. Elsevier/Academic Press, 535 pp.
Houze, R. A., 1993: Cloud Dynamics. Academic Press, 573 pp.
Johns, R. H., 1982: A synoptic climatology of northwest flow se-

APRIL 2006

NOTES AND CORRESPONDENCE

vere weather outbreaks. Part I: Nature and significance. Mon.

Wea. Rev., 110, 16531663.
, 1984: A synoptic climatology of northwest-flow severe
weather outbreaks. Part II: Meteorological parameters and
synoptic patterns. Mon. Wea. Rev., 112, 449464.
Kaplan, M. L., Y.-L. Lin, D. W. Hamilton, and R. A. Rozumalski,
1998: The numerical simulation of an unbalanced jetlet and
its role in the Palm Sunday 1994 tornado outbreak in Alabama and Georgia. Mon. Wea. Rev., 126, 21332165.
Kim, J., and L. Mahrt, 1992: Momentum transport by gravity
waves. J. Atmos. Sci., 49, 735748.
Klemp, J. B., 1987: Dynamics of tornadic thunderstorms. Annu.
Rev. Fluid Mech., 19, 369402.
, and R. B. Wilhelmson, 1978a: The simulation of threedimensional convective storm dynamics. J. Atmos. Sci., 35,
10701096.
, and , 1978b: Simulations of right- and left-moving
storms produced through storm splitting. J. Atmos. Sci., 35,
10971110.
Kristovich, D. A., N. F. Laird, M. R. Hjelmfelt, R. G. Derickson,
and K. A. Cooper, 1999: Transitions in boundary layer
meso- convective structures: An observational case study.
Mon. Wea. Rev., 127, 28952909.
Maddox, R. A., 1976: An evaluation of tornado proximity wind
and stability data. Mon. Wea. Rev., 104, 133142.
Markowski, P. M., C. Hannon, J. Frame, E. Lancaster, A. Pietrycha, R. Edwards, and R. Thompson, 2003: Characteristics of
vertical wind profiles near supercells obtained from the
Rapid Update Cycle. Wea. Forecasting, 18, 12621272.
Peterson, R. E., 1984: A triple-Doppler radar analysis of a discretely propagating multicell convective storm. J. Atmos. Sci.,
41, 29732990.
Rasmussen, E. N., and D. O. Blanchard, 1998: A baseline climatology of sounding-derived supercell and tornado forecast
parameters. Wea. Forecasting, 13, 11481164.
Redelsperger, J., and T. L. Clark, 1990: The initiation and horizontal scale selection of convection over gently sloping terrain. J. Atmos. Sci., 47, 516541.
Rickenbach, T. M., 1999: Cloud-top evolution of tropical oceanic

247

squall lines from radar reflectivity and infrared satellite data.

Mon. Wea. Rev., 127, 29512976.
Schlesinger, R. E., 1975: A three-dimensional numerical model of
an isolated deep convective cloud: Preliminary results. J. Atmos. Sci., 32, 934957.
, 1978: A three-dimensional numerical model of an isolated
thunderstorm. Part I: Comparative experiments for variable
ambient wind shear. J. Atmos. Sci., 35, 690713.
Shutts, G., 2003: Inertiagravity wave and neutral Eady wave
trains forced by directionally sheared flow over isolated hills.
J. Atmos. Sci., 60, 593606.
Steenburgh, W. J., S. F. Halvorson, and D. J. Onton, 2000: Climatology of lake-effect snowstorms of the Great Salt Lake.
Mon. Wea. Rev., 128, 709727.
Toutenhoofd, V., and J. B. Klemp, 1983: An isolated cumulonimbus observed in northeastern Colorado: Comparison of field
observations with results of a three-dimensional simulation.
Mon. Wea. Rev., 111, 468478.
Wakimoto, R. M., C. Liu, and H. Cai, 1998: The Garden City,
Kansas, storm during VORTEX 95. Part I: Overview of the
storms life cycle and mesocyclogenesis. Mon. Wea. Rev., 126,
372392.
Weckwerth, T. M., J. W. Wilson, R. M. Wakimoto, and N. A.
Crook, 1997: Horizontal convective rolls: Determining the
environmental conditions supporting their existence and
characteristics. Mon. Wea. Rev., 125, 505526.
Weisman, M. L., and J. B. Klemp, 1982: The dependence of numerically simulated convective storms on vertical wind shear
and buoyancy. Mon. Wea. Rev., 110, 504520.
, and , 1984: The structure and classification of numerically simulated convective storms in directionally varying
wind shears. Mon. Wea. Rev., 112, 24792498.
, and , 1986: Characteristics of convective storms. Mesoscale Meteorology and Forecasting, P. S. Ray, Ed., Amer.
Meteor. Soc., 331358.
Wilhelmson, R. B., and J. B. Klemp, 1978: A numerical study of
storm splitting that leads to long-lived storms. J. Atmos. Sci.,
35, 19741986.