Determining Midlatitude Cyclone Structure and Evolution
from the Upper-Level Flow

David M. Schultz
NOAA/National Severe Storms Laboratory, Norman, Oklahoma

Heini Wernli
Swiss Federal Institute of Technology (ETH), Zürich, Switzerland

5 January 2001

Corresponding author address:
Dr. David M. Schultz
NOAA/National Severe Storms Laboratory
1313 Halley Circle
Norman, OK 73069

1  Introduction

The importance of surface weather-map analysis to marine interests cannot be overstated. Features analyzed on surface maps (e.g., low pressure systems, fronts) convey basic information about the type and intensity of weather present. Due to the sparseness of marine surface observations, however, the analysis of surface features in between observations often (1) is inferred from other information such as satellite imagery, (2) is drawn according to the analysts' previous experience, or (3) is envisioned in terms of a conceptual model. This paper illustrates a method to assist in surface analysis by incorporating information from the large-scale flow and using conceptual models of marine midlatitude cyclone structure and evolution.

Those who regularly perform surface frontal analyses of midlatitude cyclones often recognize that these storms can undergo a variety of different evolutions (e.g., Smigielski and Mogil 1995; Sienkiewicz 1996). Two of these evolutions have been observed frequently enough to attain conceptual-model status: the Norwegian and Shapiro-Keyser cyclone models (Fig. 1).

The Norwegian cyclone model (Fig. 1a), so named to honor the Norwegian meteorologists (e.g., Bjerknes, Bergeron, and Solberg) who first conceptualized the typical life cycle of midlatitude cyclones in the 1910s and 1920s, presents the evolution of a cyclone from an incipient frontal wave with cold and warm fronts (stage I), to a deepening cyclone with a narrowing warm sector as the cold front rotates around the cyclone faster than the warm front (II and III), and finally to a mature cyclone with an occluded front (IV). Typically, a Norwegian cyclone is oblong, oriented roughly north-south with the cold front more intense and longer than the weak and ``stubby" warm front.

The Shapiro-Keyser cyclone model (Fig. 1b) is named after the authors of the study that first presented this conceptual model of the frontal structure in some marine cyclones (Shapiro and Keyser 1990). As with the Norwegian cyclone model, an incipient cyclone develops cold and warm fronts (I), but in this case, the cold front moves roughly perpendicular to the warm front such that the fronts never meet, the so-called frontal T-bone (II). Also, a weakness appears along the poleward portion of the cold front near the low center, the so-called frontal fracture, and a back-bent front forms behind the low center (III). By stage IV, colder air encircles warmer air near the low center, forming a warm seclusion. Typically, a Shapiro-Keyser cyclone is oblong, elongated east-west along the strong warm front.

An important factor in determining which evolution will be preferred in a given situation has been determined recently to be the nature of the large-scale flow. For example, recent idealized numerical-model experiments (e.g., Davies et al. 1991; Thorncroft et al. 1993; Wernli et al. 1998) have indicated significant sensitivity to the profile of wind speed in the direction perpendicular to the flow (the cross-jet direction). Although these previous simulations have been valuable, the jet stream also has significant variations in wind speed along the flow (the along-jet direction). Other studies have indicated that the along-jet variations in wind speed can be important in influencing the structure and evolution of cyclones (Evans et al. 1994; Young 1995; Schultz et al. 1998). Along-jet variations in wind speed are often associated with confluence (flow lines coming together) and diffluence (flow lines spreading apart). Confluence frequently occurs in features known as jet-entrance regions and diffluence frequently occurs in jet-exit regions (Fig. 2). As will be shown in this article, cyclones embedded within diffluent flow tend to evolve like the Norwegian cyclone model, whereas cyclones embedded within confluent flow tend to evolve like the Shapiro-Keyser cyclone model.

2  Observed cases

Two marine cyclones are presented. The first (Fig. 3) occurred within diffluent flow, as seen from the 300-mb winds over the cyclone. Note also that the winds are stronger upstream of the cyclone than downstream of the cyclone. The cyclone evolves into a Norwegian cyclone with a strong cold front and a weak warm front. At the end, an occluded front develops, as seen by a tongue of warmer air extending from the warm sector towards the low center. The stretching deformation in the large-scale flow due to the diffluence leads to the dominance of the north-south-oriented cold front and the elongation of the cyclone.

In contrast, a cyclone in confluent flow has stronger 300-mb winds downstream of the cyclone than upstream of the cyclone (Fig. 4). This cyclone develops a much stronger warm front than cold front, with the fronts nearly perpendicular to each other. No warm tongue is present and the cyclone more closely resembles the Shapiro-Keyser model. In this case, however, no warm seclusion develops. As with the previous example, the deformation due to the confluence leads to the dominance of the east-west-oriented warm front and the elongation of the cyclone.

3  Idealized model

To test whether the structural and evolutionary differences between these two cyclones can be attributed to the large-scale flow, we perform idealized primitive-equation channel-model simulations, neglecting surface friction and condensation. [The model is the same one used in Wernli et al. (1998).] To abstract the large-scale flow, two-dimensional basic states representing strong and weak large-scale flows are joined together to obtain a confluent or diffluent flow regime. The STRONG basic state is represented by a maximum wind speed of 83 knots at 320 mb and strong temperature gradient throughout the troposphere, whereas the WEAK basic state is represented by a maximum wind speed of 65 knots at 320 mb and weak temperature gradient throughout the troposphere. The transition region between the STRONG and WEAK states occurs over 3000 km. A cyclonic disturbance embedded at upper levels this flow produces an initial perturbation surface wind field of 8.7 knots.

The DIFF simulation (STRONG basic state transitioning to the WEAK basic state) shows the development of the strong cold front and weak, short warm front, along with the narrowing of the warm sector (Fig. 5). The formation of the warm seclusion at the cyclone center is a result of the zero surface friction. This feature could be eliminated if the modeled surface friction was greater (e.g., Hines and Mechoso 1993). Other than the warm seclusion, this two-day evolution resembles the Norwegian cyclone model.

The CONF simulation (WEAK basic state transitioning to the STRONG basic state) shows the development of the strong warm front and weak cold front (Fig. 6). No occluded front forms because the fronts are nearly perpendicular to each other. This evolution is consistent with the Shapiro-Keyser cyclone model.

4  Conclusion

This investigation shows that cyclones embedded in diffluent large-scale flow tend to develop structures and follow evolutions similar to the Norwegian cyclone model, with a strong cold front, weak warm front, and narrowing warm sector to form an occluded front. In contrast, cyclones embedded in confluent large-scale flow tend to develop structures and follow evolutions similar to the Shapiro-Keyser cyclone model, with a strong warm front, weak cold front, and T-bone frontal structure.

Because the large-scale flow is not constant in time, cyclones can change from one type to another. For example, many initially Shapiro-Keyser-like cyclones may develop occluded fronts late in their life cycles, becoming more Norwegian-like, as the initially confluent flow becomes more diffluent during cyclogenesis.

Whereas limited surface data may make detailed analysis of the frontal structure of marine cyclones difficult, the large-scale flow may provide some guidance in recognizing these different cyclone evolutions and anticipating the relative strengths and orientations of the fronts. The Marine Prediction Center currently issues analyses and forecasts of the 500-mb flow, available on the World-Wide Web ( These graphics products can be used to infer the nature of the surface cyclones embedded within that flow.

There are other factors on the mesoscale that affect the strength of low-level fronts, so this proposed method cannot be used independently of other guidance. Nevertheless, it has shown some utility at the Marine Prediction Center as one of their analysis tools (e.g., Sienkiewicz 1996) and may be of use to other mariners and surface analysts.

Acknowledgments. The following individuals are gratefully acknowledged for their reviews of an earlier draft of this article: Charles Doswell, Matthew Wandishin, and Daphne Zaras.


Davies, H. C., C. Schär, and H. Wernli, 1991: The palette of fronts and cyclones within a baroclinic wave development. J. Atmos. Sci., 48, 1666-1689.

Evans, M. S., D. Keyser, L. F. Bosart, and G. M. Lackmann, 1994: A satellite-derived classification scheme for rapid maritime cyclogenesis. Mon. Wea. Rev., 122, 1381-1416.

Hines, K. M., and C. R. Mechoso, 1993: Influence of surface drag on the evolution of fronts. Mon. Wea. Rev., 121, 1152-1175.

Schultz, D. M., D. Keyser, and L. F. Bosart, 1998: The effect of large-scale flow on low-level frontal structure and evolution in midlatitude cyclones. Mon. Wea. Rev., 126, 1767-1791.

Shapiro, M. A., and D. Keyser, 1990: Fronts, jet streams and the tropopause. Extratropical Cyclones, The Erik Palmén Memorial Volume, C. W. Newton and E. O. Holopainen, Eds., Amer. Meteor. Soc., 167-191.

Sienkiewicz, J. M., 1996: Surface analysis at NCEP's Marine Prediction Center. Preprints, Fifteenth Conf. on Weather Analysis and Forecasting, Norfolk, VA, Amer. Meteor. Soc., 487-490.

Smigielski, F. J., and H. M. Mogil, 1995: A systematic satellite approach for estimating central surface pressures of mid-latitude cold season oceanic cyclones. Tellus, 47A, 876-891.

Thorncroft, C. D., B. J. Hoskins, and M. E. McIntyre 1993: Two paradigms of baroclinic-wave life-cycle behavior. Quart. J. Roy. Meteor. Soc., 119, 17-55.

Wernli, H., R. Fehlmann, D. Lüthi, 1998: The effect of barotropic shear on upper-level induced cyclogenesis: Semigeostrophic and primitive equation numerical simulations. J. Atmos. Sci., 55, 2080-2094.

Young, M. V., 1995: Types of cyclogenesis. Images in Weather Forecasting, M. J. Bader, G. S. Forbes, J. R. Grant, R. B. E. Lilley, and A. J. Waters, Eds., Cambridge University Press, 213-286.

Figure 1: Conceptual models of cyclone evolution showing lower-tropospheric geopotential height and fronts (top), and lower-tropospheric potential temperature (bottom). (a) Norwegian cyclone model: (I) incipient frontal cyclone, (II) and (III) narrowing warm sector, (IV) occlusion; (b) Shapiro-Keyser cyclone model: (I) incipient frontal cyclone, (II) frontal fracture, (III) frontal T-bone and bent-back front, (IV) frontal T-bone and warm seclusion. Panel (b) is adapted from Shapiro and Keyser (1990, their Fig. 10.27) to enhance the zonal elongation of the cyclone and fronts and to reflect the continued existence of the frontal T-bone in stage IV. The stages in the respective cyclone evolutions are separated by approximately 6-24 h and the frontal symbols are conventional. The characteristic scale of the cyclones based on the distance from the geopotential height minimum, denoted by L, to the outermost geopotential height contour in stage IV is 1000 km. From Schultz et al. (1998, Fig. 15).

Figure 2: Schematic illustrating confluence at the jet-entrance region and diffluence at the jet-exit region. Strongest wind speed is denoted by arrow. Geopotential height lines are solid.

Figure 3: Evolution of cyclone in diffluent flow: 0000 UTC 9, 1200 UTC 9, and 0000 UTC 10 February 1993: 850-mb height (dashed lines, every 3 dam), 850-mb temperature (solid lines, every 2C), and 300-mb winds (knots).

Figure 4: As in Fig. 3 except for cyclone in confluent flow: 1200 UTC 23, 0000 UTC 25, and 1200 UTC 26 February 1989.

Figure 5: Idealized primitive-equation simulation of cyclone in diffluent flow (DIFF; days 2, 3, and 4): surface temperature (solid lines, every 2C), surface pressure (dashed lines, every 4 mb).

Figure 6: Same as Fig. 5 except for confluent-flow simulation (CONF).

This page created on 2/25/99 by David Schultz .

Return to Current Research.
Return to David Schultz's homepage.