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

Physical processes that affect the coastal environment are often unique because of the dominant effect of the interactions between the land and the sea. The thermal contrast between the land and the sea contributes to the formation of the land-sea breeze, coastal atmospheric fronts, and atmospherically induced coastal ocean currents and upwelling. The presence of a lateral boundary can substantially modify the wind field that drives currents, upwelling, and surface waves, creating highly variable fields. As a consequence, the processes that drive the air-sea exchange of heat, mass, momentum, and trace gases over the continental shelf are highly variable; they are dominated by spatial scales on the order of tens to hundreds of kilometers. Stratiform clouds, which affect the radiation balance significantly, can form over cooler upwelled water (e.g., Rogers and Olsen, 1990), while the convergence of marine air over the coastline can result in strong convection with heavy precipitation and runoff. These mechanisms affect pollutant dispersion, coastal erosion and beach development, the coastal ecological system, and numerous other processes.

The importance of air-sea interaction processes to the flow and thermodynamic structure of the ocean is well established (Charnock, 1979). The wind-driven circulation on the continental shelf is controlled by local wind forcing on time scales on the order of hours and by remote wind forcing on time scales on the order of days. A substantial fraction of the variability of coastal currents, sea level, and temperature results from large-scale ocean waves that are coherent with remote wind forcing and can propagate many hundreds of kilometers (Davis and Bogden, 1989; Denbo and Allen, 1987).

The lateral boundary is an important element in the coastal circulation on all scales because it can constrain the flow in both the atmosphere and the ocean, and there is often close correspondence between changes in the currents and the local wind forcing.

The coastal ocean is characterized by large variations in sea surface temperature and roughness and a nonequilibrium sea state. Upwelled water is often much colder than the ambient surface water, so sharp ocean fronts form between the colder water nearshore and the warmer water offshore. These regions represent areas of particularly intense air-sea interactions because of large inhomogeneities and nonequilibrium conditions. The structure of the atmospheric boundary layer becomes increasingly complex in the vicinity of ocean fronts. Changes in the heat flux produce large variations in the stability of the boundary layer, cloud cover, and radiative fields on scales up to 100 km. Recent atmospheric measurements in the vicinity of a weak ocean front in the open ocean have revealed complex wind stress, cloud cover, and boundary layer depth patterns (Friehe et al., 1991). More dramatic effects are likely in the coastal ocean where fronts are persistent and strong (Charnock and Businger, 1991). The high biological productivity of the coastal environment also emphasizes the importance of understanding trace gas exchange in the presence of these large horizontal gradients. In particular, the coastal ocean may be an important sink for carbon because the surface partial pressure of CO₂ is controlled by photosynthesis, which depletes carbon from the upper layers (Baes and Killough, 1985; Broecker, 1982; Sarmiento et al., 1988; and a review by Broecker et al., 1985). Present models of the CO₂ cycle, however, tend to be limited to larger spatial scales and do not distinguish the effect of the coastal ocean (Moore and Björkström, 1986; and a review by Crane, 1988). In addition, it has been postulated that dimethyl sulfide, produced by phytoplankton, may be an important source of cloud condensation nuclei (Bates et al., 1987; Charlson et al., 1987; Hegg et al., 1991).

Recent measurements have also shown that the response time of currents on the continental shelf is sufficiently short that changes in ocean circulation patterns can be driven by strong wind events, such as those caused by the passage of a strong atmospheric front (Lee et al., 1989) or the interaction between mesoscale or regional pressure gradients and topography (e.g., Lackmann and Overland, 1989), and by intense cooling of the upper ocean associated with cold air outbreaks (e.g., Bane and Osgood, 1989). The problem is complicated further by mesoscale circulation patters, like the sea breeze, which contribute to the small-scale variability of wind stress (Clancy et al., 1979; Elliot and O'Brien, 1977; Mizzi and Pielke, 1984), and by open ocean processes such as the Gulf Stream that can force circulation on the outer continental shelf (Lee et al., 1989). Interactions between the atmosphere and the coastal ocean are characterized by their

intensity and horizontal variability on very small scales. The lower atmosphere, in general, responds quickly to changes in sea surface temperature and roughness. Where upwelling is particularly intense, the cold near-shore water cools the atmosphere, forming a shallow stable marine layer, capped by a large temperature inversion. Often this layer is below the height of the coastal terrain, so the marine air behaves in a manner that is very sensitive to the orientation of the coastline (Winant et al., 1988). A complete understanding of the effect of the marine atmosphere on the circulation and thermal structure of the coastal ocean requires a better understanding of how the air and sea interact on scales less than 10 km and less than 1 hour. A key to this understanding is realization of the possibility of strong coupling between the ocean and the atmosphere.

An important consideration in the coastal environment is the relative contribution of local and remote wind forcing to the ocean circulation. Remote forcing consists of the generation of coastally trapped long waves that have wavelengths on the order of 1000 km. The local wind velocity consists of two components, one part driven by the pressure gradients and another part driven by the stress divergence, which is confined to a shallow layer called the Ekman layer. Along the west coast of North America, the response of sea level to fluctuations in along-shore wind stress at large-scales accounts for a substantial fraction of the total sea-level variance. Halliwell and Allen (1984, 1987) have shown that the most effective wind stress forcing was confined to two regions, along the northern California and Oregon coasts, and northern Baja, California. The strongest wind-stress-forced fluctuations in sea level propagated northward, with the maximum correlation between sea-level fluctuations and the wind stress having lagged about 500 km equatorward 1 to 2 days earlier. Thus, large-scale processes play a major role in determining the wind-driven shelf circulation. Variance of the wind stress is largest near capes (Enriquez and Friehe, 1991; Halliwell and Allen, 1984), where strong local forcing may trigger wind-driven large-scale coastal-trapped waves.

One important coastal phenomenon is the wind-driven cross-shelf circulation that drives upwelling and downwelling (Figure 6.1). An equatorward along-shore wind on a west coast produces an offshore surface flow driven by turbulent stresses (Ekman layer). This offshore flow is compensated for by an onshore flow deeper in the water column. In addition to the horizontal cross-shelf currents generated by the wind stress, there is a compensating vertical velocity pattern that results in irreversible incorporation

Figure 6.1 Relationship of along-shore winds and coastal upwelling and of wind stress curl and divergence/convergence of surface Ekman transpore offshore of the primary upwelling zone (from Nelson, 1977).

of cold nutrient-rich deep water into the surface layer. The dynamics of the return flow that feed the Ekman layer are not well understood. An important question is whether this return flow could be geostrophic. For the CODE area, Davis and Bogden (1989) have shown that the pressure gradient is inconsistent with a geostrophic return flow to balance the Ekman transport, so that some other ageostrophic component is indicated. This suggests that the effect of the topography and local thermally driven atmospheric circulation on the wind flow may be particularly important in determining the wind-driven circulation. This problem, however, has hitherto not been addressed.

In a wind-driven upwelling environment, a sharp discontinuity in density typically develops between the less dense surface water offshore and the denser upwelled water near shore. Although an along-shore wind stress is essential to the formation of a coastal upwelling front, the effect of the wind stress on the stability of a front is uncertain. The atmosphere may respond to this ocean thermal structure on very small scales (<10 km), thereby affecting the wind speed, stress, and heat exchange between the atmosphere and ocean. In turn, there may be further feedback from the atmosphere to the ocean, also indicating that the coastal ocean and atmosphere may be

strongly coupled. Friehe et al. (1991) have shown that the wind and wind stress field can vary dramatically in response to the atmospheric stability differences that exist across an ocean front. The result is a highly variable time-dependent stress pattern that alters the basic flow field in the oceanic Ekman layer. Interactions with this flow are likely to affect the stability of a coastal upwelling front. Thus, time-dependent changes in the wind field and the Ekman layer should be considered in studies of upwelling and coastal frontal dynamics.

Atmospheric forcing also affects the structure of the temperature and salinity fronts that develop between cool shelf water and the western boundary currents that flow adjacent to the eastern continental shelves of most continents at midlatitudes. This large temperature or salinity discontinuity is maintained by momentum and buoyancy fluxes across the air-sea interface and by changes in the path of the boundary current. Rapid cooling of the sea surface may intensify the frontal structure because the shallow shelf water cools at a faster rate than the water offshore, which, in any case, is resupplied with heat by the poleward advection of warmer water. There is substantial feedback from this warm current to the atmosphere, where storms can develop, resulting in further interaction with the ocean. This is discussed later in more detail. Air-sea interactions also modify frontal boundaries that develop in buoyancy-driven and tidally dominated coastal environments where the heat exchange between the air and the sea affects the strength of the front. Of particular importance is the effect of episodic events, such as cold air outbreaks and hurricanes.

Severe storms affect many coastal regions during the winter and, to a lesser extent, in summer. Regional differences exist, with ocean storms affecting the coast from California to Alaska and extratropical cyclones moving from the interior of the continental United States to the east coast. Tropical cyclones, including hurricanes, may affect almost any part of the coast during summer and autumn. Particularly important effects are caused by cold air outbreaks along the east coast and by rapidly intensifying storms over the Gulf Stream.

Cross-shelf transport off the eastern coast of the United States is substantially affected by a combination of wind and Gulf Stream forcing (Lee and Atkinson, 1983; Lee et al., 1989). The atmospheric effect is most significant during cold air outbreak episodes in winter. This atmospheric forcing is strongly influenced by the large gradients in surface temperature that lead to low-level frontogenesis (Bane and Osgood, 1989). This is accompanied by substantial changes in the surface stress that drives the cross-shelf Ekman transports in the surface layer. The response time of the

shelf currents is sufficiently short that changes in the shelf circulation can be driven by rapidly moving atmospheric events (Lee et al., 1989). The response time of the inner-and middle-shelf regions in these conditions is between 3 and 12 hours.

Interaction between the regional or mesoscale pressure gradients and topography leads to complex wind and turbulence fields over the coastal ocean (Macklin et al., 1988; Overland, 1984; Chapter 4, this report). Strong ageostrophic winds occur in topographically restricted channels when synoptic-scale disturbances produce an along-channel pressure gradient (Lackmann and Overland, 1989). These gap winds produce a highly variable surface stress field through local enhancement of the flow downstream of the topographic gap. Large variations in the wind field and corresponding changes in the surface pressure over short spatial scales have been observed along the coast of California (Winant et al., 1988; see Figure 4.5). This pattern, which occurs with generally northerly winds, is characterized by a strong low-level temperature inversion at the top of the marine atmospheric boundary layer, and the spatial structure of the surface wind is correlated with the coastal topography.

There have been extensive studies of topographic effects on the development of internal waves in a stable stratified fluid; however, the effect of topographically forced gravity waves on the coastal marine atmospheric boundary layer and sea surface layer has received relatively little attention (Dorman, 1987; Gossard and Munk, 1954; Wald and Georgopolous, 1984). In a stable coastal environment, mesoscale variations in the flow pattern may be dominated by topographically forced waves that propagate offshore, in the lee of elevated coastal terrain, or that propagate parallel to the coast as trapped waves. Gossard and Munk (1954) observed a slight effect of a propagating atmospheric gravity wave field on sea level using measurements of surface pressure and sea level from the pier at Scripps Institution of Oceanography. Interaction of atmospheric waves with the sea surface was observed by Wald and Georgopolous (1984) using Advanced Very High Resolution Radiometer (AVHRR) satellite data. Alternate light and dark features seen downwind of islands in the Aegean Sea were accounted for by changes in the sea surface temperature and roughness. Reversals in the wind direction and variations in the depth of the marine layer along the west coast of the United States have been related to gravity current surges along the coast (Dorman, 1987). The effect of the sea surface temperature and stress patterns is uncertain, but the observations of Enriquez and Friehe (1991) indicate that sea surface temperatures along the northern California coast respond quickly to changes in reversals in wind direction due to changes

in the upwelling pattern. Herman et al. (1990) have shown that a drop in sea surface temperature of 1° to 3°C from 15 km offshore out to approximately 150 km is accompanied by a rise of similar magnitude within approximately 15 km of the shore in response to highly nonlinear bores or gravity currents.

The exchange of heat, mass, moisture, momentum, trace gases, and particulates between the sea and the air is fundamental to an understanding of the ocean-atmosphere system and global climate. Although such processes are ubiquitous over the ocean, the coastal environment poses certain difficulties. The land-sea boundary, ocean temperature fronts, surface wave field, variations in water depth, and biological activity play an important role in determining the magnitude and variability of air-sea fluxes over the coastal ocean.

Momentum flux is particularly important in determining the effect of the atmosphere on coastal circulation. Direct measurement is generally difficult, so drag coefficients that relate the turbulent stress to the mean wind are often used (Fairall and Larsen, 1986; Smith, 1988). Accurate estimates of the exchange coefficients of momentum between the atmosphere and the ocean depend strongly on atmospheric boundary layer stability, wind shear, and the ocean surface wave field (Donelan et al., 1985; Smith et al., 1990a). This is further complicated by the limited fetch and water depth in the coastal zone, and the heterogeneity of the coastal atmosphere that limits application of the steady-state assumptions inherent in the formulation of bulk parameterizations of the surface exchange processes (Geernaert, 1988, 1990; Chapter 2, this report).

The estimate of transfer velocities for trace gas exchange at the air-sea interface remains difficult in the ocean and is further complicated in the coastal environment by the variability of the wind and temperature fields and the structure of the upper ocean. Estimates of transfer velocities can be obtained using various methods (for a review, see Liss, 1988).

Assuming that production of natural ¹⁴CO₂ in the atmosphere is time invariant and that a steady state exists in the ocean-atmosphere system, the rate at which ¹⁴CO₂ enters the oceans across the air-sea interface must balance the rate at which ¹⁴C decays in the oceans, and a global estimate of the transfer velocity can be obtained (Broecker and Peng, 1974; Liss, 1988). Bomb-produced ¹⁴C can also be used to estimate the air-sea transfer velocity. Another method uses the change in oxygen concentration obtained from

a time series of dissolved oxygen. The change in concentration is attributed to biological production and consumption due to photosynthesis-respiration and decomposition and exchange across the air-sea interface. The biological component is also related to nutrient changes, so the oxygen concentration can be corrected for biological activity. The residual signal is then attributed to the transfer velocity for oxygen. Radon deficiency has also been used. This method relies on measuring the radioactive disequilibrium between the nuclides ²²²Rn and ²²⁶Ra in near-surface water. In the deep oceans the two isotopes are found to be generally in radioactive equilibrium. Near the surface, however, the gas ²²²Rn is lost to the atmosphere because its activity in air is much less than in seawater. By assuming steady state, for any measured vertical profile the rate loss of ²²²Rn to the atmosphere must equal the depth-integrated deficiency in ²²²Rn activity, and a transfer velocity can be estimated. None of these methods attempt to estimate the flux of a tracer across the air-sea interface directly; all of them depend on steady-state assumptions that are difficult to relate to environmental parameters such as wind speed, sea state, and trace gas concentrations in the surface water, all of which vary over short time scales. Ideally, direct flux estimates are needed that are based on atmospheric measurements of the covariance of the vertical velocity and trace gas concentration in the atmosphere. This method has been used to estimate the air-sea flux of CO₂ (Smith and Jones, 1985; Wesley et al., 1982), but the results are controversial because they give a transfer velocity for CO₂ that exceeds the oceanographic estimates by a factor of 100. A significant problem in the eddy correlation approach is the high signal-to-noise ratio where direct estimates of the variance of CO₂ are required. This has been overcome recently with a more sensitive CO₂ sensor (S.D. Smith et al., 1991). These observations indicate upward and downward fluxes of CO₂ in response to daily variations in CO₂ in the coastal surface seawater where there are changes in the salinity of the surface water due to freshwater runoff. Transfer velocity estimates are an order of magnitude larger than the open ocean measurements made using radon but are similar to the wind tunnel and lake studies of Liss and Merlivat (1986). Recent advances in chemical flux estimates using conditional sampling techniques that combine estimates of the variance of the vertical velocity and the mean concentration of the trace gas may reduce much of the uncertainty in atmospheric measurements of tracer fluxes across the air-sea interface (Businger and Delany, 1990; Businger and Oncley, 1990).

Coupled air-sea interactions occur when, for example, the atmosphere is forced by the ocean, resulting in a feedback to the ocean that modifies

further response of the atmosphere. Coupling occurs throughout the ocean-atmosphere system; however, it is more pronounced in the coastal environment where the temporal and spatial scales in the ocean and the atmosphere are similar. While surface exchange processes directly control the interaction of the ocean and the atmosphere, they depend greatly on the mesoscale structure of the entire boundary layer. The sea breeze, the interaction of a stable boundary layer with topography, boundary layer rolls, and clouds may all contribute to the variability of the surface fluxes. The importance of the sea breeze for enhancement of the surface wind field and subsidence offshore has been demonstrated along the Oregon coast (Clancy et al., 1979; Elliot and O'Brien, 1977; Mizzi and Pielke, 1984). Feedback to the mesoscale circulation is likely if surface temperature gradients are modified by the sea breeze circulation. Modification of the surface stress field in the vicinity of a cape along the California coast has recently been observed in the Surface Mixed Layer Experiment (SMILE) (Enriquez and Friehe, 1991). The structure of the boundary layer may be further modified by radiative processes associated with cloud development, resulting in variations of the surface fluxes due to changes in the vertical structure of the marine atmosphere (see Chapter 2).

The horizontal variability of surface temperature fields and land-sea contrasts leads to modification of the air and development of internal boundary layers (IBLs). As air flows over an abrupt change in surface properties, an internal boundary layer develops within an existing boundary layer. For the case of cold air flowing over a warm surface, an unstable IBL develops, rapidly replacing the existing boundary layer; for the case of warm air flowing over a cold surface, a stable IBL develops that may persist as a shallow layer until either the air is cooled by radiation or the surface temperature increases to reverse the stability of the air. This stable boundary layer accounts for many of the more important effects that occur in the interaction of the marine boundary layer with topography (Dorman, 1987; Winant et al., 1988).

It is clear that the coastal ocean and atmosphere are inherently interdependent or coupled because of the dependence of the atmosphere on the heat and moisture source at the sea surface and the dependence of the ocean circulation on the wind. Traditionally, however, studies have rarely combined investigations of both environments to determine the extent of the feedback between the ocean and the atmosphere or the extent to which these fluids are coupled. This applies also to the exchange of trace gases between the atmosphere and the ocean that depends on the structure of the upper ocean and atmospheric boundary layer. Direct interaction between the at-

mosphere and biological or geological processes in the ocean is more tenuous; however, the physical interaction between the atmosphere and the ocean modifies the cross-shelf circulation and therefore plays an important, albeit indirect, role.

Studies of the coupled ocean-atmosphere system in the coastal environment depend largely on understanding the scales of interaction between the two fluids and the processes that provide the strongest feedbacks. A focus on small-scale interactions would provide the opportunity to elucidate important physical processes that control the coastal ocean and atmospheric circulation and the larger-scale fields. A number of basic air-sea interaction problems are of particular interest. Parameterization of the surface fluxes in shallow water remains a problem. The surface wave field on the shelf is substantially different from that over the open ocean, and the effect of this on air-sea exchanges is not well known (Geernaert, 1990). The relative effect of topographically forced flows and thermally driven circulation on the coastal ocean circulation has not been explored. There is compelling evidence for both topographically forced and sea-breeze-driven wind stress variations along the west coast of the United States, although little is known about the effect of the thermally driven circulation on the structure of the marine layer that controls the hydraulically driven flow near the ocean surface.

The high wind speed zone in the lee of capes and points is associated with a large wind stress curl pattern (see Figure 6.1) and lower sea surface temperature in the vicinity of the cape. It is unclear whether this is due to increased surface cooling from the high winds, increased upper ocean mixing, or upwelling at small spatial scales. Similarly, these capes appear to be sources of the largest wind stress variability that propagates along the coast and may be responsible for the remote forcing of sea level and the coastal circulation. However, the relative importance of the local and remote wind forcing of the coastal ocean remains uncertain.

Another uncertainty is the effect of the increased surface fluxes associated with cold air outflow from the continent on coastal frontogenesis over the coastal ocean. At present, it is unclear whether the development of coastal frontogenesis is related to variations in the shelf circulation, larger-scale atmospheric forcing, or a combination of both. The extent of the feedback from coastal frontogenesis to the ocean via variations in the wind stress also is uncertain. The growth and decay of the atmospheric or oceanic boundary layer also are not well understood when internal waves, shear flow instabilities, and temperature fields are spatially highly variable.

Advances in our understanding of coastal phenomena will benefit from research to understand interactions between the ocean and the atmosphere over a large range of spatial and temporal scales. While it is apparent to oceanographers that the wind-driven ocean circulation is of fundamental

importance on the continental shelf, it may be less clear to meteorologists that the highly heterogeneous state of the coastal ocean, in association with the complex topography, produces a highly variable atmospheric structure on very small spatial and temporal scales across the continental shelf. Efforts to understand the extent of the coupling between the atmosphere and the ocean in this highly variable environment will be very rewarding, providing new insight into the basic physics of the atmosphere and the ocean and improving our ability to predict the mesoscale ocean and atmosphere circulations.

In conclusion, the panel recommends the following:

• Studies should be conducted to determine the relative effects of topographically forced flows and thermally driven circulation on coastal ocean circulation.

• Further research should be conducted to understand the coupled ocean-atmosphere processes that control the interactions between the wind field, atmospheric boundary layer structure, and upper ocean.

• Support should be given to encourage development of coupled ocean-atmosphere models that use integrated field measurements in the coastal ocean and atmosphere.