Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 40
--> Chapter Four Importance of Buoy/C-MAN Observations to Research and Operational Services Coastal marine observations in general, and buoy/C-MAN observations in particular, have been essential to warning and forecast operations and to a variety of basic and applied research over the past few decades. In this section, and in this analysis of the buoy/C-MAN network, a scientific perspective has been applied. The separation of weather operations from research is no longer clear. Advances in science and technology are, if not in practice at least in intent, rapidly being transferred to warning and forecast operations. This section presents an overview of some of the significant contributions of buoy/C-MAN data in supporting research and in extending the range of warning and forecast services. The United States arguably faces the world's greatest weather challenges with the most formidable array of weather extremes and the highest economic stakes. Property losses from severe weather continue to rise significantly in inflation-adjusted dollars. The growth in these losses, allowing for inflation, is rapid, doubling or tripling in constant dollars every decade. One category 5 hurricane could conceivably result in $50 billion in damage (Pielke et al., 1997). The U.S. National Mitigation Strategy (Federal Emergency Management Agency, 1995) points out that from 1980 to 1993, the value of insurable property on the Atlantic and Gulf Coasts increased 179 percent. Three hurricanes, Hugo in 1989 and Andrew and Iniki in 1992, resulted in a total of 70 deaths and $41 billion in damages. This damage was over 40 percent of the total reported damage caused by natural disasters in the United States between 1980 and 1993. Significant variations in wind, waves, and weather in the region between 5 km and 50–100 km of the coast can occur. Likewise, equally significant variations can occur in the immediate coastal waters due to coastal land features. Such variations, often undocumented because of sparse offshore observations, may be critical to marine safety. While weather has never posed greater or more urgent threats to the Nation's public safety, property, economic growth, and national security,
OCR for page 41
--> reliable, accurate knowledge of the weather has enormous potential for benefit to agricultural productivity, construction, energy efficiency, national defense, transportation, water resource management, and the economy in general. This potential has never been so close at hand, given the advance in measurement technology, modeling, and understanding. Considering that much of the U.S. population lives in or near coastal zones, including the Great Lakes (Figure 9), more data leading to improved forecasts will result in improved economic efficiency. The following example from a report of the USWRP sixth prospectus development team (Pielke et al., 1997) concerning oil and gas exploration illustrates the point. Improved forecasts of tropical weather conditions (wind, waves, and rain) can reduce delays in drilling operations at a cost of up to $250,000 per rig per day. There are several thousand rigs in the Gulf of Mexico. Improved hurricane track predictions could reduce days of production shutdown, each day of which costs the industry and the U.S. treasury a combined $15 M. The biggest and potentially avoidable, impact on oil and gas production operations comes from the threat of tropical cyclones (TCs). Shut-downs impact operating companies severely: they incur the costs associated with deferral of production, transportation of crews to safety and back, downtime for protecting equipment and structures and, later, for damage assessment, and the costs of facility repairs prior to resumption of safe operations. While fixed structures usually come back on-line within 48 to 72 hours of evacuation, floating production systems, like those now operating in deep waters, may take up to a week to resume production. Most companies allow for 5 to 7 days of weather-related production losses each year in their business plans. While costs of extreme events grab the headlines, simple forecast errors in more benign circumstances can also lead to economic losses and transfer of wealth among economic sectors. Improved storm track and landfall forecasts can save lives, save days of production shutdown, and avoid false alarms and unnecessary shutdowns for oil and gas companies operating offshore. One of the most valuable operational uses of buoy/C-MAN observations is that they permit forecasters to monitor pressure, wind, wave, and temperature conditions continuously in the coastal and immediate offshore waters. These reports allow forecasters to fine tune model-generated forecast storm tracks and to evaluate critical derived information (such as the strength of the coastal baroclinic zone) necessary to issue watches and warnings promptly for marine and coastal interests. These same buoy/C-MAN reports have been used by a variety of researchers, often in cooperation with NWS operational meteorologists, to study coastal meteorological events of major interest to forecasters. The importance of the buoy/C-MAN
OCR for page 42
--> network for research and operational purposes has been noted by Rotunno and Pietrafesa (1996) who conclude that “an expansion and enhancement of current NDBC C-MAN and buoy systems is essential.” The proposed expansion and enhancement of the buoy/C-MAN network by Rotunno and Pietrafesa (1996) and the NOAA MAROB plan detailed earlier are consistent with the stated objectives of the U.S. Weather Research Program (USWRP). The USWRP has three top research priorities that are linked to operations: (1) quantitative precipitation forecasting; (2) TC forecasting; and (3) adaptive measurement strategies. These three USWRP research priorities are driven by an operational need to provide forecasters better guidance on hazardous weather threats so that warnings can be improved and by the desire to advance basic scientific knowledge in these areas, out of which will come improved forecasting capability. Research and operations are inextricably linked in the modernized and restructured NWS, and this linkage is fully recognized in the USWRP. The three research priorities of the USWRP will require a robust buoy/C-MAN network as noted by Rotunno and Pietrafesa (1996). In August 1997, a Pacific Coast workshop was held that dealt with the problems of coastal forecasting and the associated operational requirements. (A workshop synopsis can be found in Hirschberg, 1998.) Four categories of forecast problems were identified on various time scales: (1) short-term and long-term quantitative precipitation forecasting, (2) mesoscale wind forecasting from a few hours to a few days, (3) marine layer forecasts of ceiling and visibility, and (4) sea surface conditions as measured by wind waves, swell, and currents. These types of issues are common to all of the coastal areas of the United States, including the Great Lakes and other sizable inland water bodies. Along the Gulf and Atlantic coasts there are additional forecast complexities related to the occasional landfalling tropical storm. Explosive extra tropical cyclones (ETCs) cause additional forecast problems along parts of most coastal regions and over the Great Lakes. A denser buoy/C-MAN observational network in the coastal and offshore waters is central to progress in scientific understanding of coastal weather and marine processes and in improved operational prediction capabilities. Coastal Issues and Applications Coastal baroclinic zones, often called coastal fronts, have been documented by Bosart et al. (1972) and Bosart (1975). Coastal fronts possess operational significance because the frontal position often coincides with the rain/snow line and the axis of heaviest precipitation in advance of a coastal cyclone. Coastal marine observations (including buoy and C-MAN
OCR for page 43
--> reports) have been critical to the success of various scientific investigations of coastal front life cycles and behavior. These critical observations have enabled researchers and operational meteorologists to show that coastal fronts are marked by the confluence of marine and continental air masses and are associated with strong low-level temperature contrasts across a 10 km wide frontal zone that can exceed 10°C. Coastal fronts are most commonly found in winter along portions of the Atlantic and Gulf coasts (e.g., New England, the Carolinas, the lower Texas coast). They tend to form in regions of differential heating between the land and ocean, adjacent to regions of differential heating between the Gulf Stream and the cooler continental shelf waters, and in some instances, by differential surface roughness (e.g., Bosart, 1981; Keshishian and Bosart, 1987; Nielsen 1989; Roebber, 1989; Nielsen and Neilley, 1990; Riordan et al., 1985, 1995; Riordan, 1990; Gyakum, 1991; Holt and Raman, 1992). Coastal fronts may also form inland in response to the formation of extended thermal gradients produced by differential heating associated with precipitation and evaporation or the blocking and lifting of moist air by the Appalachian Mountains (e.g., Branick et al., 1988; Bell and Bosart, 1988; Bosart and Dean, 1991; Nielsen and Dole, 1992; Fritsch et al., 1992). The buoy/C-MAN observations have been especially helpful in studies of the marine environment in which coastal fronts form in situ offshore. The availability of NWS NWS88D Doppler radar observations in conjunction with buoy/C-MAN reports and other conventional data sources has enabled researchers to better understand how differential diabatic heating across near-shore and offshore oceanic thermal gradients has contributed to in situ coastal frontogenesis and cyclogenesis (e.g., Davis and Dolan, 1992; Cione et al., 1993; Davis et al., 1993). The results from research on coastal storms and related phenomena have been used in turn by operational meteorologists to make improved weather forecasts (e.g., Gurka et al., 1995; Keeter et al., 1995). Similarly, research on oceanic coastal cyclone frequencies and cyclogenesis processes has also benefited from the availability of a regular network of coastal and offshore marine observations with regard to defining cyclone warm sector conditions, offshore deepening rates, and coastal wind and wave conditions (e.g., Bosart and Sanders, 1991). Modeling studies of coastal frontogenesis and coastal cyclogenesis (e.g., Ballentine, 1980; Stauffer and Warner, 1987; Lapenta and Seaman, 1990, 1992; Doyle and Warner, 1993a, b, c, d; Chien et al., 1997) have helped to uncover important mesoscale aspects of coastal front life cycles for which confirmatory evidence has been obtained from coastal marine observations. In a recent study, Khandekar and Lalheharry (1996) used data from American and Canadian moored buoys to help evaluate Environment
OCR for page 44
--> Canada's ocean wave prediction model. The buoy data indicate that the wave model underpredicts the highest wave heights. Verification of model forecasts is a critical use of moored buoy data. In a related study, Bouws et al. (1996) investigated the recent trend for observed wave heights to increase over the North Atlantic Ocean. The scientific issue was whether wave height trends were related to increases in swell size as opposed to changes in the heights of wind-driven waves. Hurrell (1995) showed that the positive phase of the North Atlantic Oscillation (NAO) has increased, favoring an increase in the strength of the westerlies across the North Atlantic Ocean basin since 1960. Bouws et al. (1996) used the information on the positive time tendency of the NAO to conclude that the wave height increase could be attributed to increased swell, a conclusion confirmed by Kushnir et al. (1997). Clearly, wave spectra information obtained from moored buoys are of critical importance to research studies of this nature, and which have important weather and climate ramifications. Note that Nowlin et al. (1996) have concluded that measurements of SST, wind, and wind stress measurements are of the highest priority in the construction of an ocean observing system for climate studies. The Gulf of Mexico is also a very important cyclogenesis region for North America. An important recent example occurred in March 1993 when the Gulf of Mexico spawned a vicious extratropical storm, eventually known as Superstorm 1993 (e.g., NOAA, 1994). Severe weather in the form of heavy snow, heavy rains, high winds, tornadoes, and coastal storm surges associated with Superstorm 1993 killed many people, injured many more, caused hundreds of millions of dollars in property damage, and generally disrupted human activity across much of eastern North America from Cuba to eastern Canada (e.g., Caplan 1995; Uccellini et al., 1995; Alfonso and Naranjo, 1996; Bosart et al., 1996; Dickinson et al., 1997; Schultz et al., 1997). Although Superstorm 1993 was well forecasted several days in advance over the northeastern United States (e.g., Caplan, 1995; Kocin et al., 1995; Uccellini et al., 1995), the initial cyclogenesis over the northwestern Gulf of Mexico was poorly forecasted (e.g., Dickinson et al. 1997). Gilhousen (1994) showed that the buoy/C-MAN observations over the northwest Gulf of Mexico were crucial for establishing the initial (and unforecasted) intensity of Superstorm 1993. Dickinson et al. (1997), besides showing that Superstorm 1993 was the deepest ETC to occur over the Gulf of Mexico in the 40-year period 1957–1996, also established that the NCEP and to a lesser extent operational prediction models performed poorly in predicting the intensity of the initial cyclogenesis over the Gulf of Mexico. They showed that the initial poor model forecasts of Superstorm 1993 could probably be attributed to a com-
OCR for page 45
--> bination of: (1) a significant underestimation of the initial strength of the surface baroclinity over the northwest Gulf of Mexico that was apparent in the buoy/C-MAN observations reported by Gilhousen (1994), (2) a misrepresentation of the strength of the warm oceanic ring situated over the northwestern Gulf of Mexico (Gilhousen, 1994), (3) a failure of the NCEP models to represent properly the bulk effects of cumulus convection associated with the massive convective outbreak over the Gulf of Mexico during the incipient and rapidly intensifying phase of the storm, and (4) an underestimation of the intensity of a dynamical tropopause disturbance embedded in the subtropical jet stream that helped to trigger the massive convection and initial cyclogenesis over the Gulf of Mexico. It is quite likely that a better representation of the marine environment over the Gulf of Mexico at the time of the incipient cyclogenesis would have resulted in better model forecasts of the initial storm development. The Gulf of Mexico has also been the target of many warm season research investigations that are also critically dependent upon availability and reliability of marine observations. Weiss (1992), Thompson et al. (1994), and Breaker et al. (1997) have investigated the return flow of moisture from the Gulf of Mexico to the continent as winds turn poleward behind retreating anticyclones. Given the importance of the timing of the return flow of moisture to episodic storm and precipitation events, sometimes severe, over the Plains and lower Mississippi Valley, it is crucial that offshore marine observations be readily available to forecasters charged with multiple public safety responsibilities (e.g., the evacuation of personnel from offshore oil platforms before hazardous weather arrives) and to researchers trying to document and understand the life cycles of the often complex convective systems that form in this region (e.g., Fritsch et al., 1986; Johns and Doswell, 1992; Hagemeyer, 1997; Laing and Fritsch, 1997). Coastal frontogenesis along the lower Texas coast can often be associated with exceptionally heavy rains and, occasionally, the spin up of tropical cyclones (e.g., Bosart, 1984; Bosart et al., 1992). Without the availability of buoy/C-MAN observations, these research case study investigations, with results that may also have potential operational utility, would not be possible. TC research has also benefited significantly from the availability of coastal marine observations. Powell and Black (1990), Breaker et al. (1994), Houston and Powell (1994), Dobos et al. (1995), Powell (1996), and Powell and Houston (1996) have used these observations along with base reflectivity data from the NWS operational coastal Doppler radar network (WSD88D) and wind observations derived from low-level passes by NOAA research aircraft to construct detailed offshore wind analyses accompanying landfalling tropical storms. The buoy/C-MAN observations are of critical importance
OCR for page 46
--> for calibration purposes in the construction of these wind analyses. The derived wind analyses are also of critical importance to forecasters at NOAA's TPC in the fine-tuning of hurricane landfall positions and for coordinating with federal, state, and local emergency management officials on evacuation plans. In a related study, Mettlach et al. (1994) have used moored buoy measurements to compute significant wave heights (3–8 m) and dominant wave periods (20–25 s) along the west coast of North America (Alaska to California) associated with supertyphoon Flo in the central North Pacific Ocean in September 1990. Most recently, Cione and Black (1998) have used historical buoy/C-MAN observations taken in the vicinity of tropical storms to construct profiles of thermodynamic variables as a function of radial distance from the storm center. In recent years there has also been a renewed interest in the problems of winter weather forecasting over the Great Lakes, particularly for the forecasting of lake-effect snow (e.g., Niziol et al., 1995). Passarelli and Braham (1981) showed that low-level convergence over Lake Michigan associated with a winter land breeze off snow-covered terrain helped to drive a thermally direct snow-band circulation over the lake. More recently, Reinking et al. (1993) and Byrd et al. (1991) have reported on new lake-effect snow studies over Lake Ontario. With the advent of NWS Doppler radar (88D) observations, forecasters can now monitor lake-effect snow bands much more carefully than previously. They have also been able to take advantage of research that has established the importance of vertical wind shear profiles to the type, intensity, location, and duration of lake-effect snow bands (e.g., Niziol et al., 1995). However, critical measurements of temperature, pressure, wind, and maritime air profiles over the Great Lakes during lake snow events are lacking. These measurements are required to help assess quantitatively the convergence into the lake bands and the strength of the thermally direct frontogenetical circulations that help to drive the bands. Meadows et al. (1997) remark that “the buoys are providing high-quality, long-term wave measurements for the Great Lakes. Unfortunately, the buoys are usually removed during the heavy icing season of November to March….” Given the importance of weather information over the lakes, it is hoped that future enhancements to the buoy program will include construction and deployment of moored buoys capable of withstanding the severe winter conditions on the Great Lakes. There is also a long history of research that has taken advantage of buoy/C-MAN observations along the west coast of North America for a variety of synoptic and mesoscale studies, a number of which have had operational utility. The Catalina Eddy, a mesoscale circulation that occurs in the coastal waters of southern California in response to airflow over and
OCR for page 47
--> around prominent topographic barriers, is of local significance because the marine layer deepens with the onset of the offshore cyclonic circulation (e.g., Bosart, 1983; Mass, 1989; Wakimoto, 1987; Thompson et al., 1997). The deepening of the marine layer is associated with a lessening of the concentration of pollutants in the coastal atmosphere of the Los Angeles basin, but may be associated with the longer range transport of pollutants northwestward along the coast toward Santa Barbara and inland through the mountain passes to the lower deserts of California, Nevada, and Arizona. Buoy/C-MAN observations are of critical importance to forecasters and air pollution officials who monitor the life cycle of the Catalina Eddy circulation. Severe storm and heavy precipitation studies over coastal and interior North America would also benefit from the maintenance and enhancement of a critical buoy/C-MAN network. The timing, frequency, and intensity of low-level moisture surges into the southwestern United States are of crucial importance to the onset and severity of the summer monsoon precipitation in this region. Studies by Douglas (1993, 1995), Douglas et al. (1993), Douglas and Li (1996), and Stensrud et al. (1997) have shown that the Gulf of California is an important low-level moisture source for the summer monsoon. These same studies, however, have shown that low-level time-mean flow is not properly simulated in the NCEP/ECMWF prediction models, because surface observations are lacking. Given the potential devastation and loss of life associated with flash floods in this region, it is important that the buoy/C-MAN network be enhanced to provide more reliable low-level wind information in the Gulf of California and in the Pacific south and west of San Diego. Finally, the occasional episodic severe weather outbreak (usually in the cooler half of the year) across coastal and interior California (e.g., Monteverdi and Johnson, 1996) might be easier to forecast if more information were available on wind, temperature, and moisture conditions from moored buoys on air masses approaching coastal California. Mass and Albright (1987), Dorman et al. (1995), Mass and Bond (1996), and Bond et al. (1996), among others, have investigated the mesoscale aspects of coastal cool surges along the West Coast that are associated with southerly wind reversals and coastal mesoscale pressure ridges. These authors have demonstrated that the mesoscale circulation aspects of the onshore marine “pushes,” which often end coastal interior valley heat waves, would be very difficult to deduce without the additional meteorological information provided by the buoy/C-MAN network. Bond et al. (1997) have presented an overview of the Coastal Observation and Simulation with Topography Experiment (COAST) conducted recently in the Pacific
OCR for page 48
--> Northwest. They make the point that the success of research investigations like COAST depend very much upon the synthesis of disparate data sources and types of which buoy/C-MAN observations are one important component. Finally, we note that recent studies of the synoptic and mesoscale structures of frontal systems approaching and crossing the Washington and Oregon coast (e.g., Steenburgh and Mass, 1996; Colle and Mass, 1996; Chien and Mass 1997; Chien et al., 1997) have validated the importance of the research philosophy governing the COAST experiment. Research Related to Tropical Weather Applications NOAA and its predecessor agencies, through the offshore moored buoy platforms and C-MAN sites have provided invaluable data on numerous tropical storms and hurricanes. Forecasters at NOAA's National Hurricane Center (NHC), now the TPC, and researchers at AOMLHRD, formerly the National Hurricane Research Laboratory, make extensive use of these data. Initially the buoy platforms were intended simply to replace the ocean weather ship program, and to provide a “Maginot” line of defense along coastal regions to aid in the warning of approaching storms and to supplement the network of satellite observations with in situ measurements. Over the years, users have come to rely on the augmented network of open ocean and coastal automatic weather stations for day-to-day operations and for in-depth research. Tropical Cyclone Winds Data acquired by the prototype experimental data buoys deployed by NDBC in the Gulf of Mexico and off the east coast, beginning in the 1970s, provided the first accurate marine surface measurements in the core of intense tropical cyclones (TCs) (e.g., Gulf of Mexico Hurricane Eloise, 1975 at Environmental Buoy #71; EB71; East Coast Hurricane Belle, 1976 at EB04). These data sets and the many additional data from hurricanes that followed have greatly improved our understanding of the structure of the marine surface wind fields in intense TC wind fields and have had a significant impact on TC forecasts and warnings, TC boundary layer research, diagnostic studies, three-dimensional dynamic models, ocean response modeling, and specification of design criteria for offshore and coastal structure. First, the surface wind measurements provided the first direct evidence that Monin/Obukov planetary boundary layer (PBL) similarity theory could be applied to describe the vertical shear of the horizontal wind in the ma-
OCR for page 49
--> rine boundary layer of tropical cyclones (e.g., Ross and Cardone, 1978; Cardone and Ross, 1979; Powell, 1980; Powell and Black, 1990; Liu et al., 1997). This finding has had a major impact on diagnostic studies and modeling of tropical cyclones with full three-dimensional dynamical models. These studies also have provided a basis for identifying the relationship of the flight-level wind observations (acquired by reconnaissance aircraft at flight levels typically of 850 mb and 700 mb) to winds near the surface, which impact life and property. Second, the buoy data have provided the first reliable data sets against which models of the TC PBL wind field may be evaluated and refined. Before the deployment of the buoy array, virtually no such data were available. For example, the first practical dynamically based numerical model of the wind field in the TC marine PBL (Cardone et al., 1976) could be compared only to winds from an oil platform (rig 50) in Hurricane Camille 1969, manually scaled off a strip chart whose time tics were uncertain because of a sticking clock drive, that was recording the output of an anemometer of uncertain calibration mounted on the top of the drilling rig. The measured wind data sets obtained from the NDBC buoys in the Gulf of Mexico and off the U.S. east coast have allowed validation and refinement of that first model and of alternative parametric models (e.g., Holland, 1980; Georgiou, 1985). Typically, during the 1970s and 1980s, only one of the available NDBC buoys would encounter a given TC, so that only one or at most two quadrants of the storm circulation could be investigated. The expansion of the buoy array and the addition of the C-MAN array in the late 1980s and 1990s has made possible, especially near landfall, an accurate synoptic analysis of the entire wind field (Powell, 1982; Powell et al., 1991), thereby allowing resolution of secondary wind maxima and storm asymmetries. Such analyses are now produced routinely in real-time at HRD for the NHC/TPC (Powell and Houston, 1997; Houston, et al., 1997). It is important to note that, in general, advanced satellite remote wind sensors such as radar scatterometers, radar altimeters, and passive microwave sensors, lack the sensitivity to resolve winds above about 20 m s-1, which means that remote sensors are able to resolve the intensity of TCs only up to category 1 on the Saffir-Simpson scale. In situ sensors are needed to correctly resolve the intensity of storms of greater intensity. Ocean Response Modeling The enhanced ability enabled by the NDBC buoy measurements to specify surface wind and stress fields in TCs through synoptic analysis and
OCR for page 50
--> modeling has had, and continues to have, a large impact on ocean response modeling. Numerous modeling studies have reported on storm surge, mixed-layer, and three-dimensional current response to hurricanes (e.g., Forristall et al., 1980). The buoys have made an even greater impact on understanding ocean wave generation under extreme wind forcing and on ocean wave modeling, because the buoys also provide accurate measurements of wave height, period, and the frequency spectrum. For example, Ochi (1994) used buoy wind and wave measurements acquired in hurricanes Eloise, Frederic, Anita, Belle, Kate, and Gloria to establish useful new functional relationships between the wind and wave parameters. First and second generation wave models developed in the 1970s and early 1980s were often validated against wave data acquired by NDBC buoys (e.g., Ross and Cardone, 1978). The advanced community third-generation spectral wave prediction model (WAM), which now is used at most operational numerical weather prediction centers around the world for global and regional wave forecasting, was not released and published until it was thoroughly validated against measured NDBC wave data acquired in Gulf of Mexico hurricanes Anita in 1977 and Frederic in 1979 (Wave Advanced Dynamics Model 1; WAMD1 Group, 1988). Many new studies are underway utilizing data sets acquired in intense U.S. east coast storms during the active 1995 (notably Felix and Luis) and 1996 (Fran and Bertha) seasons. The availability of directional wave sensors on many NDBC buoys during these years are allowing further refinement of source terms representing physical mechanisms of wave growth and dissipation. The NDBC buoys and their Canadian cousins moored on the Scotian Shelf and Grand Banks have measured record high sea states (significant wave heights up to 17 m and maximum wave heights up to 30 m) in recent severe hurricanes, as well as in severe extratropical storms such as the Storm of the Century (1993) and Halloween Storm (1991) (Cardone et al., 1996). These data sets have revealed a tendency for even the most advanced wave models to under-specify extreme sea states, a finding that has stimulated further research in ocean wave dynamics. In fact, an important field experiment and modeling program Surface Wave Advanced Dynamics Experiment (SWADE) in 1990/1991, was based on the existing U.S. East coast NDBC buoy array (Cardone et al., 1995). Another reliable use of buoys may be in research and forecasting of coastal upwelling. Upwelling regions are key to much of the ocean's biological productivity, and buoy measurements provide critical data for coupling meteorological, oceanographic, and biological models in developing the means to manage coastal food sources.
OCR for page 51
--> Coastal Oceanographic Research and Coastal Weather The coastal buoy network is an important source of continuous weather and wave information to fishing, shipping, and recreational interests. In addition, the real-time data from these buoys and the climatologies and time series subsequently generated from them are essential to ongoing oceanographic and marine research. For example, from available buoy data scientists have determined the typical annual signal of wind and SST along the U.S. west coast. Three distinct geographical regions have been identified (Dorman and Winant, 1995; Schwing et al., 1997a, b). Each region has a unique biological signature (with respect to primary production), zooplankton, and fish communities (Parrish et al., 1981, 1983; U.S. GLOBEC, 1994). Buoy data can also define seasonal changes such as the “spring transition” into upwelling favorable conditions along the coast (Strub et al., 1987a). The Coastal Ocean Forecast System, jointly developed by NWS, National Ocean Service, and Princeton University researchers shows the value of in situ observations. In areas where high-quality, relatively dense observational networks are available, forecasts of surface and subsurface ocean temperature are more accurate (e.g., Kelley et al., 1997). Schwing et al. (1997a) show regional differences in annual alongshore wind and SST at selected buoys off the west coast. Winds within the Southern California Bight are weak and variable throughout the year, particularly in summer, relative to those north of Point Conception. South of Cape Mendocino, winds are equatorward throughout the year. Minimum SSTs off much of California occur in late spring, due to the cooling effect of coastal upwelling. Winds are typically highly variable on shorter synoptic (3–14 day) time scales in all months, creating brief periods of downwelling (southerly winds along the California coast) and warmer SSTs. Buoy winds and SSTs also display significant interannual differences. Based on this well-defined seasonality, conditions vary on synoptic as well as interannual scales in ways that are not possible to forecast. Local wind stress is an important factor in driving coastal currents and determining their variability on synoptic scales (Strub et al., 1987b; Winant et al., 1987; Chelton et al., 1988; Rosenfeld et al., 1994; Paduan and Rosenfeld, 1996). This ocean response to wind-forcing also influences the position of water mass features, such as coastal jets and eddies, and the distribution of heat, salt, nutrients, carbon, and marine organisms (Huyer, 1983; Kosro et al., 1991). All these studies used data collected by coastal buoys. More recent studies have shown the spatial patterns of wind stress and air-sea heat
OCR for page 52
--> flux from buoy and ship data to be much more complex than previously thought (Winant and Dorman, 1997). Buoy winds are useful in separating the effects of regional winds and the remote ocean connection to tropical ENSO activity (Ramp et al., 1997). They also are incorporated in larger data sets, for example, the Comprehensive Ocean Atmosphere Data System (Woodruff et al., 1987) which have been used to identify decadal-scale fluctuations in wind forcing and SST (Schwing et al., 1997b). Because these variations are unpredictable, information from coastal buoys is essential for monitoring changing conditions for synoptic and climate forecast models. Wind stress and heat flux information from buoys are critical elements for diagnostic and prognostic atmosphere, ocean circulation, mixed layer, and wave modeling (Blumberg and Mellor, 1987; WAMD1 Group, 1988; Clancy, 1997). Some of these data are included in the model forcing fields, incorporated into assimilation schemes in prediction models, and used to verify model output. Other models that rely on input from buoys include those used to predict the evolution of hazardous coastal events, such as oil spills and harmful algal blooms. Bosart (1981) showed that the failure of the NCEP (then NMC) models to predict the Presidents' Day snowstorm on February 19, 1979, could be attributed in part to the failure of the NMC analyses to accurately depict the strength of the temperature contrast from the coast eastward to the Gulf Stream. In addition, the NMC surface analyses were systematically too cold in the offshore waters, a situation that precluded the models from representing the existence of unstable air offshore. Bosart and Sanders (1991) examined the NMC model failures in a case of a surprise snowstorm across part of New York and New England on October 4, 1987. Again, Bosart and Sanders (1991) were able to show that, if offshore marine observations had been used, they would have helped place the surface cyclone correctly and define the strong surface temperature gradient. And in the case of the March Superstorm, Gilhousen (1994) and Dickinson et al. (1997) showed that NCEP initial analyses misplaced the location of the region of strong thermal contrast over the Gulf of Mexico and underrepresented its intensity, which may have contributed to the poor model forecasts of the initial cyclogenesis. Huo, et al. (1998) ran a numerical experiment with the Canadian Regional Element (RFE) Model in which the original RFE surface temperature analysis was replaced by a new objective analysis in which all available temperature observations from buoys and ships of opportunity over the Gulf of Mexico were included. A comparison of the RFE runs using the original and modified objective analyses suggested that forecasts of the incipient cyclogenesis were significantly improved with the better surface tem-
OCR for page 53
--> perature analysis. The authors conclude that although the results of their limited study cannot be taken as definitive, they do suggest that attempts to improve the lower tropospheric thermal field in model initial analyses in data sparse regions might lead to improved forecasts. Coastal buoy data are needed for “ground-truthing” satellite data (e.g., scatterometer, SST) and refining algorithms. Very few alternative permanent sites in the ocean are available for this purpose. These ground measurements will become even more critical as we advance technologically toward greater use of remote sensing and as new space, air-borne, and shorebased observational systems are developed. Buoy data can also be integrated with satellite data to create a more complete and accurate picture of the coastal environment. In addition to direct studies of the coastal ocean, buoy data have been applied to studies of coastal meteorological processes, such as the structure of the marine boundary layer (Beardsley et al., 1987) and coastal gravity currents (Dorman, 1987). These observations in real-time are important in tracking the development and movement of storms before landfall. Meteorological and surface ocean variables are important for identifying conditions within critical reproductive periods for living marine resources. U.S. GLOBEC (1994) lists a number of studies showing correlation between coastal environmental conditions, typically wind and temperature, and recruitment in fishery stocks, as well as many of their prey species. Measurements from buoys provide data to quantify this relationship and to characterize individual years by their effect on reproductive and recruitment success. Wind reversals, or relaxations in upwelling, are linked to recruitment in nearshore marine species (Farrell et al., 1991). Winter storms in California and Oregon, with the possibility of increased rainfall that some may link to E1 Niño, are being studied by government and university scientists hoping to improve forecasts of heavy rain, snow, and wind along the west coast. The study, called CALJET (California Land-Falling Jets Experiment), begun on December 1, 1997, includes researchers and forecasters from NOAA, and U.S. Navy, and various universities. The study will run through March 1998, which is the wet season in that area. This field experiment is aimed at scientifically examining the required observations to improve one type of coastal forecast problem. Studies of this type are needed for many other phenomena as well. The CALJET study illustrates two points particularly relevant to the analysis of the buoy/C-MAN system herein. First, experiments of opportunity may be tapped for additional data that could augment the core buoy/C-MAN operational network. Second, CALJET uses a great variety of instruments and platforms that should be considered when looking for al-
OCR for page 54
--> ternative or complementary observing methods to the moored buoy or C-MAN station network. Rapidly intensifying coastal storms occur over the mid-Atlantic during the winter season (Sanders and Gyakum, 1980; Sanders, 1986; Rogers and Bosart, 1986). These storms produce heavy snow, gale force winds, and crippling ice, and have been known to batter the eastern seaboard of North America from the Carolinas to Newfoundland. On average, there are 13–15 substantial storms per year, which annually cost over a billion dollars in property damage (Dirks et al., 1988). The toll on human life has also been great, for example, 50 lives were lost in the April 6–7, 1982, snowstorm and 70 deaths were attributed to the February 11–12, 1983, “Megalopolitan” snowstorm (e.g., Sanders and Bosart, 1985a, b; Bosart and Sanders, 1986; Chang et al., 1989). The Presidents' Day Storm on February 18–19, 1979, (e.g., Bosart, 1981; Uccellini et al. 1984, 1985, 1987) and the March 12–14, 1993, superstorm (e.g., Kocin et al., 1995; Uccellini et al., 1995; Bosart et al., 1996; Dickinson et al., 1997) produced copious amounts of snowfall and effectively paralyzed much of the eastern United States. There is also a long history of these storms disrupting naval operations, commercial shipping, and fishing, occasionally resulting in extreme damage to marine vessels, including sinking. During the late fall through early spring, strong horizontal temperature contrasts often arise due to the offshore presence of the Gulf Stream and cold dry air over the adjacent land. Many times, the average distance to the Gulf Stream from Cape Hatteras, North Carolina is less than 60 km, with typical air temperature above the Gulf Stream front ranging between 22°C and 24°C, while nearby land-based air temperatures range between -20°C and 10°C. As a result, large gradients in low-level air temperature are often observed. In fact, during periods of strong offshore cold advection, where average land surface temperatures can remain at or below 0°C for extended periods of time, horizontal gradients of air temperature can exceed 30°C 50 km-1 or 0.6°C km-1. These large horizontal gradients of air temperature translate into large horizontal gradients of surface latent and sensible heat fluxes. Under strong cold air outbreak (CAO) conditions, total surface turbulent heat flux values have been observed to exceed 1500 W m-2 (Wayland and Raman, 1989; Riordan, 1990; Vukovich, 1991). Even under moderate CAO conditions, the degree of vertical heat and moisture transport that occurs within the lower troposphere is enough to quickly destabilize the low-level offshore Gulf Stream environment. Fantini (1990) has shown that this pre-storm destabilization may act to significantly increase the likelihood for subsequent rapid cyclogenesis or reintensification. The apparent importance of pre-storm periods on future cyclonic intensification prompted an investigation to look at the effects of pre-storm
OCR for page 55
--> baroclinicity on wintertime coastal cyclogenesis. The climatological study conducted by Cione et al. (1993) incorporated cold season (November-April) cyclonic episodes from the period 1982 through 1990 along the Carolina and Virginia coasts. Storms entering or spawned in this area are subjected to the highly variable baroclinic zone associated with the lateral, onshore-offshore meanders of the Gulf Stream (Pietrafesa and Janowitz, 1980), and buoy data are essential to the study strategy. Cione et al. (1993) show that the pre-storm baroclinicity was strongly associated with the intensification of coastal cyclones. This is potentially very useful for operational winter storm forecasting. The findings of this study and their potential value to forecasting depend greatly on the availability of satellite-derived SST observations. But when clouds or aerosols interfere, buoy data are the only reliable source of data to drive the promising new forecast tools. Another example of the utility of marine buoy data is in the prediction of flooding along the North Carolina coast and in coastal sounds during the passage of severe TCs and ETCs (Xie et al., 1997). Collaborating researchers from North Carolina State University and the NWS operational forecasters in Raleigh use a 1 km resolution version of the Princeton Coastal Ocean Model. Predicted storm tracks with marine buoy and C-MAN winds are used to direct inputs in both forecasting and, following storm passage, post-analysis assessment. This flooding model, while in a developmental mode, was cited as a success story and the Raleigh Forecast Office was given a NOAA Unit Citation Award in 1997. In addition to the above, use of the buoy and C-MAN data becomes important, particularly during storm conditions; they initialize and verify ocean circulation and mixed-layer wave models, predict the path and spread of toxic spills, ground-truth satellite measurements, and in studying marine biology. The buoy/C-MAN data are useful also in studies of the marine boundary layer, distant locations to monitor ENSO signatures, coastal climatology studies, wind stress and air-sea heat flux, coastal flooding, and beach erosion, to name several applications. Also, the personal views of one experienced forecaster was solicited (Ainsworth, personal communication, 1998) in Appendix G.
Representative terms from entire chapter: