Transverse cirrus bands in weather systems - Wiley Online Library

suitably to these natural hazards and risks. These facts show the importance of studying with more detail the impacts and consequences of an extreme rainstorm in Lisbon, in order to prevent future disasters and to propose suitable mitigation measures.

References

Correspondence to: Marcelo Fragoso, Centro de Estudos Geográficos, Alameda da Universidade, 1600-214 Lisboa, Portugal. [email protected] © Royal Meteorological Society, 2010 DOI: 10.1002/wea.513

Transverse cirrus bands in weather systems: a grand tour of an enduring enigma

Weather – February 2010, Vol. 65, No. 2

Fragoso M, Tildes Gomes P. 2008. Classification of daily abundant rainfall patterns and associated large-scale atmospheric circulation types in southern Portugal. Int. J. Climatol. 28: 537–544. Gimeno L, Trigo RM, Ribera P, Garcia JA. 2007. Editorial: Special issue on cut-off low systems (COL). Meteor. Atmos. Phys. 96: 1–2. Gumbel EJ. 1958. Statistics of extremes. Columbia University Press: New York.

Nieto R, Gimeno L, De la Torre L, Ribera P, Gallego D, García Herrera R, García JA, Nuñez M, Redaño A, Lorente J. 2005. Climatological features of cut-off low systems in the Northern Hemisphere. J. Climate 18: 3085–3103. Trigo RM, Trigo IM, Da Camara CC, Osborn TJ. 2004. Climate impact of the European winter blocking episodes from the NCEP/NCAR Reanalyses. Clim. Dyn. 23(1): 17–28.

Exceptional Lisbon rainfall event

thunderstorm systems were formed over the ocean near to the western coast of Portugal and moved slowly towards the Lisbon region throughout the night, between the late hours of 17 February and the early hours of 18 February. By the time those convective cells reached the Lisbon area, deep cloud systems had moved and passed slowly over the study area for more than eight hours, being responsible for intense precipitation activity that resulted in accumulated values exceeding 100 millimetres. The effects of the storm in the Lisbon metropolitan region were harmful and in some cases even ruinous. A large number of urban inundations and country flash-floods have caused loss of life and serious damage to property, demonstrating some lack of preparedness on the part of the city to respond

John A. Knox1 A. Scott Bachmeier2 W. Michael Carter1 Jonathan E. Tarantino1 Laura C. Paulik1 Emily N. Wilson1 Gregory S. Bechdol1 Mary J. Mays1 1

Department of Geography, University of Georgia, Athens, Georgia 2 Cooperative Institute for Meteorological Satellite Studies, Space Science and Engineering Center, University of Wisconsin-Madison Transverse cirrus banding (Figure 1) is defined by the American Meteorological Society Glossary of Meteorology (1999) as: Irregularly spaced bandlike cirrus clouds that form nearly perpendicular to a jet stream axis. They are usually visible in the strongest portions of the subtropical jet and can also be seen in tropical cyclone outflow regions. However, this definition raises more questions than it answers. Why are transverse bands irregularly spaced? Why are they perpendicular to a jet stream axis? Do these traits differentiate the bands from other cloud

Figure 1. Transverse bands in cirrus clouds associated with the polar jet stream over Cape Breton Island in the Maritime Provinces of Canada, as viewed from the Space Shuttle looking east. From http://www.solarviews.com/huge/earth/jet.jpg, adapted from http://eol.jsc.nasa.gov/scripts/sseop/ photo.pl?mission=STS039&roll=80&frame=60. (Image STS039–80–60 courtesy of the Image Science & Analysis Laboratory, NASA Johnson Space Center.)

35

Transverse cirrus bands in weather systems Weather – February 2010, Vol. 65, No. 2

phenomena? Why is there a preference for bands occurring near the subtropical jet? If the bands are related to jets, why are they also seen in the outflow of tropical cyclones? Do these bands reveal the inner workings of jets and cyclones? Are transverse bands observed with other, non-tropical, phenomena? And, ultimately, what are the physical processes associated with the formation, maintenance and dissipation of transverse bands? Complete answers to these questions are not forthcoming from the scientific literature. As a first step towards answering these questions, in this article we survey the current state of knowledge regarding transverse cirrus bands. In particular, we emphasize the presence of these features in a wide variety of weather systems across the globe as seen in high-resolution satellite imagery. By juxtaposing these disparate observations, we hope to shed some light on the nature and possible causes of transverse bands and to differentiate them from other visually similar cloud phenomena. While this article represents one prong of a research effort on clear-air turbulence forecasting, anyone who has ever marvelled at these persistent features in cirrus clouds can appreciate a ‘grand tour’ of this frequently observed atmospheric enigma.

Transverse bands versus other phenomena

36

What transverse cirrus bands are, and are not, has been the subject of some debate. Ellrod (1989) used GOES visible satellite imagery to differentiate between three types of features in cirriform clouds: transverse bands on the equatorward flank of a jet, ‘billows’ apparently associated with gravity waves, and ‘scallops’ on the poleward side of the jet. Lester (1994) distinguished between two types of band patterns leading to turbulence near the jet stream: billow clouds … and transverse bands … [which] are usually longer and broader than billows. Figure 2 illustrates these differences in a case of convection over the central United States, in which the transverse bands extend radially away from the centre of convection for hundreds of kilometres. In contrast, the billows associated with gravity waves at the lower right portion of Figure 2 are oriented perpendicular to the transverse bands, are much smaller in extent, are found at lower levels and appear to be liquid water clouds, not ice-crystal cirrus clouds. Still smaller billow clouds associated with Kelvin-Helmholtz instability (Dutton and Panofsky, 1970) cannot be resolved in most satellite imagery (Overeem, 2002). Despite clear differences in size, shape and orientation, these distinctions between transverse cirrus bands and other bandlike cloud phenomena are not always fully appreciated (Beckman, 1981).

Figure 2. Visible satellite image of upper-level convective outflow over Kansas (left), Missouri (centre) and Iowa (top) at 1602 UTC on 24 May 2005. Irregularly spaced transverse bands are evident to the north of convection over northern Missouri and southern Iowa. In contrast, regularly spaced billows associated with gravity waves are seen to the northeast of convection over eastern Missouri. Note that the billows are oriented perpendicular to the transverse bands. From http://cimss.ssec.wisc.edu/goes/blog/wp-content/uploads/2005/10/ 050524_G12_VIS_ZOOM.GIF

Transverse bands in jet streams The earliest mentions of transverse bands relate them to jet streams and aviation interests. Frost (1953) advised pilots to locate the jet stream by identifying banded cloud patterns. Not long thereafter, transverse bands were associated with clear-air turbulence (CAT). The connection between transverse bands (especially the widest and thickest bands) and CAT is now a longstanding aviation forecasting rule-of-thumb (Ellrod, 1989). Understanding transverse banding is therefore of considerable interest to aviation forecasting. Ground-based photography also stimulated research on transverse bands. Schaefer and Hubert (1955) performed an investigation using photographs and radiosonde data of cloud formations around the polar jet stream. The duo’s research led to the conclusion that jet-stream clouds, including ‘ripples’, ‘waves’ and ‘banded cirrus’, will most often form in the warm, moist air equatorward of a jet stream. The advent of satellite meteorology revealed the frequent occurrence of transverse banding. A remarkable structure of transverse bands viewed from satellite was depicted in the March 1964 Monthly Weather Review (Anonymous, 1964). These bands, approximately 15 to 30 kilometres in width, were located along and slightly south of an

upper-level jet core over the west coast of Mexico. Whitney et al. (1966) used transverse bands as an identifying characteristic of jet streams in TIROS satellite imagery, a relationship that is still widely recognized today (Galvin, 2007). By 1970, however, the research trail grew cold as transverse bands became an increasingly overlooked feature in satellite imagery. Another notable feature of transverse cirrus bands in jets is their persistence. Galvin (2007) commented that bandlike ‘ribbons’ may persist in association with the subtropical jet stream for days at a time, an observation consistent with our own experience. However, little research appears to have been done to quantify or explain this characteristic. Researchers have attempted to explain the presence of transverse bands by examining the dynamical characteristics of the jetstream winds. Attention has been focused mainly on horizontal and vertical wind shear. Schaefer and Hubert (1955, Figure 6) found cirrus cloud features to be coincident with the overlap of regions of negative absolute vorticity (i.e. negative values of the sum of the Earth’s rotation plus the flow’s rotation about the local vertical axis, or strongly anticyclonic flow) and low Richardson number (i.e. small values of the ratio of the squares of static stability and vertical wind shear). Viezee et al. (1967) determined that areas of transverse bands were associated with

Some of the most spectacular examples of transverse banding occur in the outflow of tropical cyclones. Nevertheless, a recent informal electronic survey of hurricane experts revealed no consensus explanation for their long-noted presence (Kerry Emanuel, Chris Velden and Raymond Zehr, personal communications, June 2007). According to Zehr (2004),


Transverse bands in tropical cyclones

Transverse cirrus bands in weather systems

higher wind speeds, larger vertical wind shear and greater static stability than were found in non-band regions. Galvin (2007) remarked that the transverse bandlike ‘ribbons’ are associated with large horizontal wind shear. A recent example of transverse banding associated with a jet stream is presented in Figure 3. In this enhanced infrared satellite image, two distinct sets of transverse bands are present: one grouping embedded in the cirrus shield just southwest of the California coast near San Francisco Bay and the other, larger, grouping located further south and west from the first. Both groups occur in advance of a single jet streak with winds exceeding 170 knots located roughly 200 kilometres upstream. The bands in both groupings are aligned roughly perpendicular to the 300mbar wind flow (not shown). Analysis reveals that the transverse bands in Figure 3 are on average 16 kilometres wide, with lengths ranging from 115 kilometres to approximately 240 kilometres. Average spacing between individual bands is 8 kilometres. These bands developed and persisted over a period of several hours. Also faintly visible just left of centre in Figure 3 is a packet of billows associated with gravity waves to the south of the trough axis. As in Figure 2, these gravity wave-associated clouds are easily distinguished from the transverse bands due to their regular spacing, smaller dimensions, and their orientation perpendicular to the transverse bands.

Figure 3. MODIS infrared image depicting a cirrus shield associated with a deep trough over the eastern Pacific Ocean at 2133 UTC on 4 January 2008. Notice the regions of transverse bands extending southwestward from the California coast. From http://cimss.ssec.wisc.edu/goes/blog/wp-content/ uploads/2008/01/080104_modis_ir_wv_anim.gif

wheel. This azimuthal symmetry is in contrast to the usual appearance of transverse bands in jet stream cirrus on one flank of the jet. Approximately 75 ‘spokes’ can be discerned around the CDO in Figure 4(b); based on contemporary estimates of Isabel’s eye width, this translates to a rough estimate of 10 kilometres spacing between the bands.

More often, transverse banding is seen in the periphery of the outflow of tropical cyclones. An excellent recent example is from Tropical Storm Fay near the Florida peninsula (Figure 5). In this image, the bands in Fay are spaced more widely than in previous examples; they are approximately 50 kilometres apart in the outermost band of outflow cirrus

The causes, implications, and role that such features play in tropical cyclone evolution have not been well documented. More investigations are needed to understand why they exist and what they mean. Figure 4, inspired by Zehr’s research, depicts Hurricane Isabel (2003) with and without significant transverse banding. In Figure 4(a), the outflow from Isabel on 12 September 2003 lacks significant banding near its centre. There are small bands at the outer periphery of the central dense overcast (CDO) northeast of the eye. Less than one day later (Figure 4(b)), there is an impressive amount of transverse banding present around the entire periphery of the CDO of Isabel, resembling spokes on a

Figure 4. NOAA AVHRR 1 km resolution infrared images of Hurricane Isabel on (a) 12 September 2003 at 1725 UTC and (b) 13 September 2003 at 1446 UTC. Note the limited transverse banding in the far northern outflow bands in (a), as contrasted with the extensive banding throughout the outflow in (b).

37


the first definitive hurricane-strength tropical cyclone in the South Atlantic Ocean, which affected Brazil in March 2004. The image, from the International Space Station on 27 March 2004, is approximately one day after Catarina’s period of maximum intensification to a ‘category 2’ storm on the SaffirSimpson scale. In this image, transverse


extending from Alabama to the Atlantic Ocean. The bands developed during a period when Fay generally maintained its strength despite its centre being located over land. Transverse banding is not limited to North Atlantic hurricanes. In Figure 6, transverse banding is prominent in Cyclone Catarina,

Transverse bands in mid-latitude mesoscale convective systems

Figure 5. MODIS infrared image of transverse bands associated with Tropical Storm Fay at 1558 UTC on 21 August 2008. Note the banding throughout the entire northern semicircle of outflow, from over Alabama through Georgia and South Carolina and over the Atlantic Ocean.

38

banding is present throughout the spiral bands of the storm and into the interior of the storm. The most prominent banding is in the southwest region of the storm, but some banding also appears on the eastern side of the storm. Similar banding has been observed in numerous recent tropical cyclones around the world. For example, during 2007 the following storms exhibited pronounced transverse banding: Super Cyclone Gonu in the Arabian Sea, Hurricane Flossie near Hawaii, Cyclone Sidr in the Bay of Bengal, and Hurricane Noel off the southeast US coast (see http://cimss.ssec.wisc.edu/goes/ blog/ for more details on each storm). In most of these cyclones, transverse banding appeared to occur episodically, rather than persistently for days as in the case of jet streams. Furthermore, the bands often (but not always) developed in concert with intensification, a theme we explore in more detail in the next section.

Figure 6. International Space Station photograph of Cyclone Catarina near the Brazilian coast on 27 March 2004, looking approximately northward. Note the transverse banding along the western, southern and eastern fringes of the outflow. From http://eol.jsc.nasa.gov/sseop/images/EO/highres/ ISS008/ISS008-E-19646.JPG. (Image courtesy of the Image Science & Analysis Laboratory, NASA Johnson Space Center.)

In addition to their long-recognized presence in jet streams and tropical cyclones, transverse cirrus bands also develop in the upper-level outflow of mesoscale convective systems (MCS) in the mid-latitudes. Research is currently being focused on the relationship between these banding events, rapidly expanding cumulonimbus anvils, and clearair turbulence (Bedka et al., 2007). As in some tropical cyclones such as Isabel, transverse bands associated with MCS activity may develop as ‘spokes on a wheel’. More often, however, the bands in MCS upperlevel outflow form in an asymmetric pattern up to several hundred kilometres away from convective cores (Figures 7–9). In Figure 7, sharply defined transverse bands (with approximate spacing of 20 kilometres) are accompanied at their far edges by a linear cloud feature across South Dakota and Minnesota. Finding nothing in the research literature on this cloud feature, we tentatively call it a convectively generated mesoscale cirrus streak. This cirrus streak is almost exactly perpendicular to the transverse bands, conforming to the shape of the MCS cirrus shield and related to accelerating upper-level winds (not shown). The cirrus streak developed by 1100 UTC, about one hour prior to the most spectacular banding, and persisted for at least four hours while the bands faded and redeveloped along the northwest side of the outflow. A hint of a similar cirrus streak is visible at the top of Figure 8, which depicts a recent transverse banding event over Minnesota and Wisconsin. The exceptional resolution of the MODIS instrument permits visual contrast between upper-level


Figure 7. GOES-8 visible image of transverse bands on the northwest side of the mesoscale convective system outflow over Minnesota and South Dakota at 1245 UTC on 21 July 1998. From http://cimss. ssec.wisc.edu/goes/misc/980721_1245_vis.GIF. See http://cimss.ssec.wisc.edu/goes/misc/980721_ vis_java.html for a four-hour animation.

transverse banding and the low-level flowparallel cumulus cloud streets – illustrating the diversity of linear band features in the Earth’s atmosphere (Kuettner, 1959, 1971). An extraordinary transverse banding event associated with an MCS occurred over the upper Midwest United States on 14 August 2007 (Figure 9). Two pulses of banding were observed over Wisconsin and the upper peninsula of Michigan, roughly 300 kilometres to the northeast of the coldest cloud tops and strongest convection in each case. Each banding event developed up to two hours after a peak in convection (as determined by cloud-to-ground lightning data) and persisted for at least five hours after the convective maximum. These events collectively exhibit characteristics similar to those found by other researchers. Trier et al. (2008) note that while the bands can form on all sides of an MCS, they are commonly found on the northern edges, as seen in Figures 7–9. In a survey of turbulence in transverse bands as detected by aircraft during summer 2006, Lenz (2008) concluded that the band spacing was generally between five and ten kilometres with initiation of banding occurring during mature convective events about seven hours after storm development. The presence of transverse bands along the northern fringe of MCS outflow, as in Figures 7–9, may be connected to the influence of jet stream winds on the MCS outflow. Fritsch and Maddox (1981) determined that the strongest winds are located on the northern side of MCSs because upperlevel winds are enhanced when anticyclonic outflow of an MCS is oriented in the same direction as the environmental flow. This can increase the vertical wind shear; Trier et al. (2008) have suggested a role for vertical shear instabilities in the creation of the bands and subsequent reports of turbulence from aircraft. However, clear-air turbulence can also be related to horizontal shear (Knox, 1997). Lenz (2008) noted the relationship between the bands and horizontal shear in the form of relative vorticity; the bands typically aligned with the direction of the relative vorticity gradient. Research is ongoing to discern the relationships of both vertical and horizontal shear to transverse banding events.

Transverse bands in extra-tropical cyclones

Figure 8. MODIS 250m resolution true colour visible image of transverse cirrus bands associated with a mesoscale convective system over northeast Minnesota and northern Wisconsin at 1741 UTC on 27 July 2008. (For orientation, the western shoreline of Lake Michigan is visible at right.) Low-level cumulus cloud streets can be seen to the east of the bands. Adapted from http://ge. ssec.wisc.edu/modis-today/images/terra/true_color/2008_07_27_209/t1.08209.USA3.143. 250m.jpg

Transverse cloud bands can be found embedded within the structures of extratropical cyclones, in the polar front jet (Galvin, 2009) and often (but not exclusively) in the cirrus shield associated with the warm conveyor belt (WCB) (Carlson, 1980). Figure 10 presents one such example, which occurred during a period of rapid cyclone development over the eastern Pacific. The

39


Figure 9. MODIS infrared image with overlaid negative (yellow) and positive (aqua) 15-minute lightning data for a mesoscale convective system over Wisconsin on 14 August 2007. Note the extensive transverse banding along the northern and northeastern fringe of the outflow. From http://cimss.ssec.wisc. edu/goes/blog/wp-content/uploads/2007/08/070814_modis_ir_ltg.jpg. For a 12-hour animation, see http://cimss.ssec.wisc.edu/goes/blog/wp-content/uploads/2007/08/ir_lightningmovie.gif

pattern and were coincident with strong upper-level divergence and possible gravity wave activity. Dixon et al. (2000), in an examination of striations with wavelengths of 30 to 55 kilometres at 5.5 to 6.5 kilometres elevation in the cloud head of an extra-tropical cyclone, identified several possible mechanisms which might result in such features, including conditional symmetric instability, convective rolls similar to those found in boundary-layer convection, and KelvinHelmholtz instability. Dixon et al. (2000) settled on convective rolls as the most likely explanation. The similarity of Dixon et al.’s striations to transverse bands is not entirely clear, however. In Figure 11 the relationship between transverse bands and turbulence reports from aircraft is illustrated in the context of a mid-latitude cyclone. Banding developed along the northwest edge of an occluded cyclone’s comma cloud over and near Lake Erie. Large commercial aircraft equipped with sensors providing estimates of eddy dissipation rate (an objective measure of turbulence) reported turbulence up to severe levels (red symbol in Figure 11) while flying through the bands. Clearly, the dynamical processes leading to transverse banding and aviation turbulence are not limited to any one type of weather phenomenon.

Discussion This ‘grand tour’ of transverse cirrus bands in a variety of weather systems reveals many commonalities among banding events: •

•

• •

• Figure 10. GOES-11 water vapour image of an extra-tropical cyclone over the eastern North Pacific Ocean on 13 January 2008 at 1200 UTC exhibiting transverse banding in the warm conveyor belt. Adapted from http://cimss.ssec.wisc.edu/goes/blog/wp-content/uploads/2008/01/conveyor_ belts.070.png

40

transverse banding in the WCB in Figure 10 is present in a region of anticyclonic curvature, on the equatorward flank of the system. At least two other studies have noted transverse bands in extra-tropical cyclones.

Feren (1995) identified transverse cirrus banding with wavelengths of about 25 to 70 kilometres as a precursor to major cyclogenesis in the eastern Australian-western Tasman Sea region. In Feren’s climatological analysis, the bands occurred in a ‘delta’ cloud

•

There is no consensus on why transverse bands form, although both horizontal and vertical wind shear are often implicated. Whether in equatorward jet-stream flanks, tropical cyclone or MCS outflow, or in warm conveyor belts, these bands frequently occur in anticyclonic flow situations. The bands often occur during or after intensification of weather systems. The bands are typically perpendicular to lower-level billows caused by mesoscale gravity waves, and are also frequently both more irregularly spaced and more extensive in spatial coverage than gravity wave-associated billows. The bands are sometimes accompanied by a ‘convectively generated mesoscale cirrus streak’ oriented perpendicular to the bands at their farthest extent. Many of these events would not have been fully captured by satellite imagery as recently as a decade ago, such have been recent improvements in horizontal resolution.

We are currently conducting research on case studies of MCS-related transverse

Acknowledgements Our grateful thanks to Kris Bedka and Bob Sharman for data, assistance and suggestions, and to the reviewers for constructive comments. Stephen Corfidi, Gary Ellrod, Stephen Jascourt, David Schultz and Steven Silberberg provided thoughtful reviews of drafts of the manuscript. Wayne Feltz, Josh Durkee, Holly Carrico, and Chris Fuhrmann also contributed useful comments.

References American Meteorological Society. 1999. Glossary of Meteorology. 2nd edition. American Meteorological Society: Boston, MA. Anonymous. 1964. Picture of the Month. Mon. Weather Rev. 92: 104.

Beckman SB. 1981. Wave clouds and severe turbulence. National Weather Digest 6: 30–37. Bedka K, Feltz W, Mecikalski J, Sharman R, Lenz A, Gerth J. 2007. Satellite signatures associated with significant convectivelyinduced turbulence events. Proceedings of the Joint 2007 EUMETSAT Meteorological Satellite Conference and the 15th Satellite Meteorology & Oceanography Conference of the American Meteorological Society, Amsterdam, The Netherlands. EUMETSAT. http://www.eumetsat.int/Home/Main/ Publications/Conference_and_Workshop_ Proceedings/groups/cps/documents/ document/pdf_conf_p50_s4_01_bedka_ p.pdf [Accessed 20 January 2009]. Carlson TN. 1980. Airflow through midlatitude cyclones and the comma cloud pattern. Mon. Weather Rev. 108: 1498–1509. Dixon RS, Browning KA, Shutts GJ. 2000. The mystery of striated cloud heads in satellite imagery. Atmos. Sci. Lett. doi:10.1006/asle.2000.0001. Dutton JA, Panofsky HA. 1970. Clear air turbulence: A mystery may be unfolding. Science 167: 937–944. Ellrod GP. 1989. A decision tree approach to clear air turbulence analysis using satellite and upper air data, NOAA Technical Memorandum NESDIS


banding in order to better document the dynamical conditions that lead to their development. An explanation for transverse cirrus bands could benefit our understanding of a variety of mesoscale and large-scale weather phenomena, as well as provide a direct boost to efforts to forecast turbulence for the aviation industry.


Figure 11. Aviation turbulence (symbols) associated with transverse bands over and near Lake Erie (centre of image) on 28 March 2005. Colour codes: red = severe turbulence; green = moderate turbulence; blue = light turbulence. Letter codes: aircraft is: A = above cloud; I = in cloud; B = below cloud; C = in clear skies. Wind barbs indicate ambient wind speed and direction in knots. (Figure courtesy Kris Bedka, University of Wisconsin; EDR data courtesy Bob Sharman, National Center for Atmospheric Research.)

23. Satellite Applications Laboratory: Washington DC. Feren G. 1995. The ‘striated delta’ cloud system—a satellite imagery precursor to major cyclogenesis in the eastern Australian-western Tasman Sea region. Weather Forecast. 10: 286–309. Fritsch JM, Maddox RA. 1981. Convectively driven mesoscale weather systems aloft. Part I: Observations. J. Appl. Meteorol. 20: 9–19. Frost BC. 1953. Flying in jet stream wind. Shell Aviation News 186: 4–8. Galvin JFP. 2007. The weather and climate of the tropics: Part 2 – The subtropical jet streams. Weather 62: 295–299. Galvin JFP. 2009. Weather image: Two cases of turbulence over western Europe. Weather 64: 83–84. Knox JA. 1997. Possible mechanisms of clear-air turbulence in strongly anticyclonic flow. Mon. Weather Rev. 125: 1251–1259. Kuettner JP. 1959. The band structure of the atmosphere. Tellus 11: 267–294. Kuettner JP. 1971. Cloud bands in the earth’s atmosphere. Observations and theory. Tellus 23: 404–425. Lenz A. 2008. Identification of aviation turbulence signatures from mesoscale convective features using satellite imagery. Unpublished senior thesis, University of Wisconsin-Madison: Wisconsin, USA. Lester PF. 1994. Turbulence: A new perspective for pilots. Jeppesen: Englewood, Colorado. Overeem A. 2002. Verification of clear-air turbulence forecasts. Technisch rapport, KNMI. http://www.knmi.nl/~overeem/ CAT_TR.pdf [Accessed 20 January 2009]. Schaefer VJ, Hubert WE. 1955. A case study of jet stream clouds. Tellus 7: 301–307. Trier S, Sharman RD, Fovell RG. 2008. Case studies of widespread turbulence in the vicinity of mesoscale convective systems using 13-km RUC analyses. Proceedings 13th Conference Aviation, Range and Aerospace Meteorolology, New Orleans, LA. American Meteorological Society: Boston, MA. Viezee W, Endlich RM, Serebreny SM. 1967. Satellite-viewed jet stream clouds in relation to the observed wind field. J. Appl. Meteorol. 6: 929–935. Whitney Jr LF, Timchalk A, Gray Jr TI. 1966. On locating jet streams from TIROS photographs. Mon. Weather Rev. 94: 127–138. Zehr RM. 2004. Satellite products and imagery with Hurricane Isabel. Proceedings 13th Conference Satellite Meteorology and Oceanography, Norfolk, VA. American Meteorological Society. http://ams.confex. com/ams/pdfpapers/79158.pdf [Accessed 20 January 2009].

Correspondence to: Dr John A. Knox, Department of Geography, University of Georgia, Athens, GA 30602, USA. [email protected] © Royal Meteorological Society, 2010 DOI: 10.1002/wea.417

41