Coastal Meteorology: A Review of the State of the Science (1992)

As traveling disturbances pass over a coastline or ice edge, they experience a change in surface friction, heat fluxes, and possibly orography. The sudden change in surface conditions can modify the disturbance, but it also can give rise to entirely new phenomena that are peculiar to the coastal region. Similar effects can be produced indirectly, through interactions with the local coastal circulations discussed in preceding sections. Furthermore, variations in large-scale conditions can strongly influence coastal circulations, producing phenomena that are not found in steady-state large-scale conditions. The complexities introduced by such mesoscale and synoptic-scale interactions hinder conceptual understanding, and accurate forecasting of changing coastal conditions requires the simultaneous simulation of a variety of disparate processes and their interactions. Manifestations of these interactions are numerous; only examples are provided here.

One example of a weather system that is directly modified by the change in surface conditions across a coastline is the typhoon or hurricane. Hurricanes require a warm ocean for maintenance and intensification. Although hurricanes soon weaken over land, the weakening is accompanied by potentially devastating short-term phenomena. For example, orographic barriers influence the dynamics of land-falling hurricanes through the blocking and resultant disruption of the circulation of the storm. Torrential rainfall in coastal environments and large latent heat releases result from this interac-

tion. Even storm motion is influenced, as shown by Bender et al. (1987) for the island of Taiwan; there the storm moved north of the westward translation it would have had in the absence of the terrain.

During and after landfall, the near-surface winds associated with the storm decelerate as a result of increased surface friction, which causes an expansion in radius of the eye wall cloud as well as creation of a stably stratified atmosphere in the lower atmosphere due to lower tropospheric cooling. Above the lowest levels, however, the winds usually remain strong and can even increase temporarily as coupling with the surface is diminished. The resultant large vertical wind shear has been used to explain the frequent occurrences of tornadoes in these storms (Novlan and Gray, 1974, as described in Pielke, 1990, Figure 4.6). Some land-falling hurricanes, however, do not produce tornadoes. We need to understand better the complex interaction of the coastal environment and the hurricane that often, but not always, results in tornadoes (e.g., with Hurricanes Carla in 1961 and Celia in 1970). A first modeling study of deep cumulus convection in hurricane environments is reported by McCaul (1991). Farther inland the acceleration of winds aloft due to decoupling from the surface may explain localized areas of tornado and strong wind damage such as those that occurred in Charlotte, North Carolina (Hurricane Hugo, 1989), and Washington, D.C. (Hurricane David, 1979). Extratropical cyclones are also modified significantly as they strike land, as found, for example, on the Washington coast in the Cyclonic Extratropical Storms (CYCLES) project (Hobbs et al., 1980).

The presence of a coastline or ice edge at high latitudes appears to strongly favor the formation of intense storms. Such storms, called arcticfront-type polar lows by Businger and Reed (1989), form over water just beyond the ice edge, where the large temperature contrast between cooler air and warmer water leads to strong low-level baroclinicity and weak stratification. As documented by Reed and Duncan (1987), trains of polar lows can form along the strong temperature gradient (arctic front), which suggests an instability mechanism. On the other hand, the most intense polar lows form rapidly beneath upper-level troughs that move from the ice shelf to the open ocean. According to Businger (1991) and Emanuel and Rotunno (1989), some polar lows attain the intensity and structure commonly associated with hurricanes.

The range of possible mechanisms for polar low formation includes barotropic and baroclinic instability, conditional instability of the second kind, and air-sea interaction instability (Twitchell et al., 1989). Recent evidence suggests that even katabatic winds (see Chapter 4) can play a

critical role in the formation of some Antarctic lows (Bromwich, 1991). Progress has been hindered by the lack of observations on sufficiently small space and time scales over the lifetime of polar lows, although interesting case studies have recently emerged from the Arctic Cyclone Experiment (Shapiro et al., 1987) and the Ocean Storms Experiment (Bond and Shapiro, 1991). Although observations seem to indicate that more than one mechanism operates in individual polar low events, little is known at present about how the proposed mechanisms interact with each other.

Along midlatitude coastlines, a wide range of processes are often present and influence each other. Important hybrid mesoscale circulations can develop that are a combination of thermally forced (Chapter 3) and topographically forced (Chapter 4) processes. One example is locally known in the northeastern United States as the backdoor cold front (Bosart et al., 1973), in southeast Australia as the southerly buster (Colquhoun et al., 1985), and in New Zealand as the southerly change (Smith et al., 1991). When the sea (or other large body of water) is much colder than the land, a cold front moving parallel to the coast undergoes local intensification and advances rapidly along the coast while being inhibited farther inland. While the cold front appears to be orographically trapped, numerical simulations have shown that differential surface heat fluxes play a primary role in modification of the cold front (Howells and Kuo, 1988).

Another type of hybrid circulation is common along the east coast of the United States during winter storms. As the wind becomes easterly ahead of a winter cyclone, coastal frontogenesis occurs, and cold air damming occurs along the mountains. After a few hours, the trapped cold air is bounded to the east by the coastal front, whose inversion extends along the top of the cold dome toward the mountains. The combined circulation enhances the coastal front by maintaining a source of cold air and keeping the front nearly stationary. As the coastal front intensifies, the temperature difference across the inversion increases, strengthening the damming and making it more difficult for the cold air to advect over the mountains. The combination of shallow cold air against the mountains, warm moist air ascending over the cold dome, and an approaching winter storm often leads to low visibility and hazardous ice storm conditions for sites east of the Appalachians (Forbes et al., 1987). Details of the interaction between the coastal front and cold air damming, such as whether one tends to trigger the other, are still unclear.

Coastal regions of the eastern United States are a favored location for cyclogenesis, and the resulting coastal storms tend to be rather different from their counterparts over the open Pacific and Atlantic or in the Mid-

west. East Coast cyclones have been shown to be generated or strongly modified by the Appalachian Mountains, the land-sea contrast, and the sea surface temperature contrast concentrated along the north wall of the Gulf Stream, as well as such coastally confined circulations as cold air damming and coastal fronts. (For examples, see Hobbs, 1987, and Holt and Raman, 1990.) The close geographical proximity of all these influences within the coastal region makes the individual interactions difficult to separate. In complex winter storms, such as the Presidents' Day storm of 1979, cyclogenesis has been found to depend on all the above factors acting in concert (Lapenta and Seaman, 1990; Uccellini et al., 1987).

One modifying coastal influence not yet discussed is localized latent heat release. In the prestorm environment, cold dry continental air has typically been advected from the northwest. This air remains cold and dry over land, while over water it receives both heat and moisture from the sea surface in a process known as preconditioning. When a developing storm approaches, the extra energy ingested by the marine boundary layer encourages rapid intensification of the storm over water or along the coast. The heat fluxes that occur as the storm intensifies are generally less important (Kuo et al., 1991).

When a strong cyclonic circulation is not already present, any mechanism that tends to focus latent heat release has the potential to produce a small-scale cyclone. The dynamics of coastal cyclones and low-level jets that form along the Baiu (Mei-yu) front of eastern China and Japan appear to be strongly dominated by the direct effects of latent heat release focused by orography or a frontal wave (e.g., Chen and Yu, 1988; Nagata and Ogura, 1991). Along the east coast of the United States, it has been suggested that coastal fronts can supply the necessary focusing mechanism and cooperatively interact with the resulting cyclone. For example, Keshishian and Bosart (1987) documented a case in which a small-scale coastal cyclone propagated northeastward along an extensive coastal front. Because the circulations associated with the coastal low acted to enhance the temperature gradient and strengthen the coastal front, it was called a ''zipper low.'' The increased low-level baroclinicity and low-level moisture may have played an important role in the major cyclogenesis event that followed.

The triggering mechanism for coastal latent heat release may be large-scale upward motion or forced lifting by a coastal front and cold air damming. In either case, a large-scale trough is generally approaching from the west to provide the necessary flow configuration. The combination of large-scale ageostrophic circulations and forced lifting is thought to have caused the initial formation of the Presidents' Day low, but small-scale coastal

cyclogenesis has sometimes been triggered by large-scale ascent alone in the absence of a coastal front or cold air damming.

There are large gaps in our present understanding of coastal interaction processes. Theoretical studies tend to focus in isolation on specific processes that are most amenable to analytical treatment and interpretation. By contrast, observational and numerical studies tend to focus on the most extreme cases, and it has been found that the extreme cases tend to involve a wide range of mesoscale and synoptic-scale processes interacting cooperatively. The range of interactions also presents a modeling challenge, since successful forecasts or simulations must adequately handle both the individual processes and their interactions. Without a better understanding of the nature of the interactions, verification of model dynamics and improvement of numerical forecasts become difficult.

We recommend strongly focused numerical, observational, and theoretical investigations into specific mesoscale-synoptic interactions:

• Research should be conducted to determine the dynamics of the local intensification of cyclone winds by coastal topography and the resulting modification of storm intensity and motion.

• Further research should be implemented to identify the causes of strong local winds, tornadoes, and extreme precipitation within land-falling hurricanes, polar lows, and extratropical cyclones.

• Studies should be conducted to understand the role of the coastal baroclinic zone and katabatic winds in the formation and dynamics of polar lows.

• Studies should be undertaken to determine the dynamical influence of coastal heating discontinuities in the along-shore propagation and local intensification of cold fronts.

• Studies should be carried out to understand the role of topography in the formation and motion of coastal fronts.

• Further research to quantify the nature of the influence of coastal fronts on midlatitude coastal cyclogenesis should be supported.

• Research should be conducted to clarify the importance of coastally induced moisture inhomogeneities for small-scale cyclogenesis and low-level jet formation.