Ocean voyagers want the most accurate forecasts they can get. Marine weather forecast accuracy depends in large part on the extent to which the weather models used by forecasters recognize and accommodate the close coupling that exists between the ocean and the atmosphere.
Efforts to evaluate the probable accuracy of a weather forecast often fail to adequately consider the possible effects of boundary characteristics. Inland, the presence of mountain ranges, dry desert plains or large lakes often significantly affects local air mass properties and the associated weather. Moving offshore, marine weather regimes can be particularly sensitive to the presence of a coastline or significant variations in sea surface temperatures associated with warm water currents such as the Gulf Stream or the East Australian Current.
Sensitivity to boundary characteristics is one reason why some models do a better job forecasting snowfall events than squall lines. This issue of boundary characteristics is of particular importance when one is attempting to assess risks of a small boat passage where concerns are focused on an area much smaller than the synoptic scales (about 1000 kilometers) or even the mesoscale features (10 to 100 kilometers) embedded in the larger area maps provided by most marine weather services.
Gulf Stream effects
It was not too many years ago when forecasts were often in error in the vicinity of the Gulf Stream. The available models failed to adequately treat the effects of the abrupt increase in heat and moisture rising upward along the inshore edge of the stream — a condition only made worse by the limited observational data that were available at that time. The situation today is much improved, but there are still “surprises.”
Ocean-atmosphere coupling comes in a variety of forms, both physical and chemical, and a variety of spatial scales. NASA’s famous image of the “Blue Marble” provides graphic evidence that the extent of the coupling is nearly global since the ocean does cover more than 71 percent of the Earth’s surface. The evident cloud patterns represent the condensation of water vapor rising from the surface. This condensation process is a central element of the coupled ocean-atmosphere system, and it serves to induce atmospheric circulation that facilitates additional coupling.
The Great Ocean Conveyor Belt transports heat from the tropics to the higher latitudes.
In introductory meteorology, we learned that a portion of incoming solar radiation reaches the Earth’s surface and warms a thin layer of the upper surface of the continents (about 1 to 2 meters in New England). But, facilitated by wind-driven mixing, this results in a much deeper layer of warmed ocean water (about 100 to 500 meters) worldwide. This warmed mass of high-heat-content water provides a valued reservoir of heat that affects the temperature of the air in contact with the ocean’s surface, inducing rising temperatures and ameliorating local seasonal climate variability.
In addition to the direct conduction of heat across the air-sea boundary, the solar warming of the surface of the ocean induces evaporation or a change of state from the liquid to a vapor. This water vapor mixes with the rising air and, in time, depending on the vertical temperature gradient in the air over the water, will condense to form clouds. In the condensation process, the heat required to induce the initial evaporation at the surface will be released aloft, warming the adjacent air. This variety of warming due to direct conduction and the release of “latent heat” can affect the local density of the air that, in turn, can modify local atmospheric pressure inducing horizontal pressure gradients sufficient to move masses of air.
The central element
This relatively simple sequence of warming, evaporation and condensation represents the central element of the coupled ocean-atmosphere system producing the winds we care about and, on a larger scale, the circulation of the atmosphere moving heat from the equator to the poles and maintaining climatic conditions suitable for human life on Earth.
The condensation process responsible for the release of latent heat in the upper atmosphere also causes rain. Scientists estimate that approximately 78 percent of the global rainfall occurs over the ocean. This input of fresh water affects the salinity of the seawater and the associated density of the water column. Just as temperature changes in the atmosphere can affect density and induce horizontal pressure gradients and the winds, so too can changes in water column density affect ocean flows.
Again, using the Gulf Stream as an example, we know that the stream is marked by an abrupt change in water temperature. It is also a distinct salinity boundary with salinities increasing from a practical salinity unit (PSU) of 33 inshore along the adjacent continental shelf to more than 36 PSU in the stream and into the Sargasso Sea. The combination of temperature and salinity gradients produces spatial variations in water column density and the associated pressures that induce a significant fraction of the Gulf Stream current field.
An extreme example of wind-driven waves during a hurricane.
These warm, salty waters proceed across the Atlantic and — in the vicinity of 45° W — split with the northern limb, proceeding northward into the Norwegian Sea, where in time they sink and form an essential portion of the subsurface limb of the Great Ocean Conveyor Belt. Although this feature has little relevance to a small boat navigator, it serves to transport significant quantities of heat; so, it is of great interest to those involved in the ongoing debates regarding global climate change.
Direct physical effects
In addition to the heat exchange and rainfall effects across the air-sea interface, the atmosphere exerts a direct physical influence on the ocean as the winds distort the surface due to friction and small-scale spatial variations in atmospheric pressure. Frictional drag causes setup — the “mounding” of water — which increases sea level elevations along shore or within large-scale oceanic gyres such as that affecting the Sargasso Sea in the North Atlantic. These mounds often support some amount of “downhill” flow, with speeds depending on the slope of the hill. Friction-induced drag also directly induces flows with speeds equaling approximately 2 to 3 percent of the wind speed. Wind-driven flows represent the other major component of the Gulf Stream current field, acting in combination with the density-driven flows.
As winds blow along the air-sea interface, the initially planar surface becomes progressively more irregular as a variety of waves form. The character of this surface wave field will vary as a function of wind speed and direction, the duration that the wind blows and the fetch — the overwater distance on which the wind is acting. The progressing waves produce an energetic velocity field that serves to mix the water column, distributing heat and a variety of chemical constituents.
Breaking waves entrain air into the water and the resulting bubbles, upon bursting, eject water particles into the air above the interface. Oceanfront residents as well as offshore sailors are only too familiar with this process, since the exchange often leaves a residue of salt particles on a boat’s brightwork or on vehicles parked along the shore. This is an additional factor affecting the local humidity of the atmosphere, beyond simple evaporation. Also, the suspended salt particles serve as condensation nuclei for clouds.
The potential for wave-breaking and the enhancement of air-sea exchange will be significantly increased in areas where the winds blow against the current. While winds blowing with the current will tend to increase current speeds, adverse winds will similarly slow currents, causing associated waves to steepen and become somewhat shorter in wavelength. The process is nonlinear, with adverse wind effects on surface wave characteristics dominant.
The relationship between wave height, wave length and current flow.
From Van Dorn’s Oceanography and Seamanship
Current and wind interaction
These effects were quantified by William G. Van Dorn and the results presented in his book, Oceanography and Seamanship. The data show that in the presence of a 5-knot current — similar in speed to Gulf Stream maxima — and waves with a 10-second period (similar to many open ocean waves), adverse winds have the potential to cause wave heights to be two to five times higher than expected with no adverse wind. Under those conditions, wavelength might be cut in half. The combination favors early onset of whitecaps and ultimately the breaking of steep waves. Such conditions at the least make for a nasty sea state and, in the limit, hazardous conditions that can roll or capsize small boats — such was the sea state in the 1979 Fastnet race.
The variety of factors affecting air-sea exchange, and ultimately the collision of masses of air causing the weather we experience, represents a particular forecasting challenge. The flows are turbulent and are best described statistically. This is the reason why we typically hear not that it will rain, but that there’s a 20 or 40 percent chance of rain. In addition, the future evolution of a weather system is uniquely dependent on initial conditions. The complexity of the air-sea exchange and the number of factors involved, as well as the limited number of direct observations in the open ocean especially, makes it difficult to precisely specify these conditions. Forecasters address this deficiency using ensembles of model runs in which the boundary conditions are changed slightly in each run to test the sensitivity of the results to selected factors. The results are used to provide the probabilistic guidance now available on the Ocean Prediction Center website.
Despite the complexity of the processes governing marine weather, there is no doubt that forecasts are improving daily due to a combination of improved computers, forecaster skills and increasing observations. For most conditions, forecast accuracy out to four to seven days is high. However, in recognition of the extent to which air-sea coupling governs weather and its evolution, it should not be a surprise if forecasts in the vicinity of a major warm water current such as the Gulf Stream are sometimes inaccurate.
Similarly, coastal forecasts where systems cross from land to sea might be more likely to err. This potential for error should be included in assessments of risk. The “what ifs” and a degree of skepticism should figure prominently in the evaluations. In addition, continuing personal observations during a passage is essential. An eye on the barometer, checking the clouds and a sense of how rapidly things are changing may be the best indication of a departure from forecast conditions. Complete reliance on some distant source of info, however well qualified, is never a good idea.
W. Frank Bohlen is a physical oceanographer and professor emeritus in the Department of Marine Sciences at the University of Connecticut. He has participated in 20 Newport Bermuda Races, five trans-Atlantic crossings and a variety of cruises from the Arctic to the Caribbean to Chile. He is a member of the Bermuda Race Organizing Committee. He sails his Ohlson 38 from Mystic, Conn.