COMMENTARY

Coastal Ocean Pollution, Water Quality, and Ecology AUTHOR Robert H. Weisberg College of Marine Science, University of South Florida

Overview

R

eal estate investing adheres to the adage that all that matters is location, location, location. For coastal ocean pollution, water quality, and ecology, the analog is all that matters is connection, connection, connection. Whereas pollution may be initiated by point sources, the connections occurring across space, time, and trophic levels are what determine the evolution of water properties, including pollutants and nutrients, and their effects on ecology. Thus, coastal ocean ecology is not only biology. It is a comprehensive, multidisciplinary subject, one that requires an understanding of how the large-scale, coastal ocean system works in its entirety, which brings us back to connectivity and the reason why a physical oceanographer is offering a commentary on coastal ocean pollution, water quality, and ecology. Ocean circulation is the fundamental determinant of such connectivity. It is what unites nutrients with light, fueling primary productivity and thence all higher trophic level interactions. As is true for any coastal ocean region, the Gulf of Mexico’s coastal ocean consists of three interconnected domains: (1) the deep ocean or the region seaward from the shelf break, (2) the coastal ocean or the continental

shelf region between the shelf break and the shoreline, and (3) the estuaries, where the rivers transition to the sea. The workings of the coastal ocean depend on the connections between these three regimes, as illustrated by Figure 1. Depth and density variations, coupled with the Earth’s rotation, provide constraints on these connections. The deep Gulf of Mexico is governed by the Loop Current–Florida Current–Gulf Stream system, which connects the Caribbean, the Gulf of Mexico, and the southeastern United States as part of the Atlantic Ocean’s western boundary current complex (Figure 2). The coastal ocean is governed locally by wind forcing, heat and fresh water fluxes, and interactions with both the deep ocean and the estuaries. The estuaries are governed by density differences between the river and ocean waters, with tides being important in how these waters mix. With these different subsystem workings, we are challenged to understand and

predict the overall functionality of the Gulf of Mexico System. However, it is not possible to understand and predict water properties and, thereby, have scientifically defensible rationale for regulation, without first understanding and predicting these connections, as the following examples will demonstrate. First, consider a glider transect taken just north of Tampa Bay, Florida, in July 2010 (Figure 3). The across-shelf distributions of chlorophyll, colored dissolved organic matter, temperature, and salinity can only be attributed to the fully three-dimensional ocean circulation. Chlorophyll is largest at depth and connected with the deep ocean, demonstrating nutrient coalescence with light from above. These across-shelf distributions occurred through upwelling of relatively deep water onto the continental shelf in response to both the Loop Current and its (shed) eddy interactions with the shelf slope. These interactions caused the

FIGURE 1 Three regions (deep ocean, continental shelf, and estuary), each with different controlling physics, whose connections, along with nonconservative processes, determine coastal ocean water properties.

March/April 2011

Volume 45

Number 2

35

FIGURE 2 Connectivity between the Caribbean, the Gulf of Mexico, and the southeastern United States provided by the Loop Current–Florida Current–Gulf Stream system. This example is for virtual surface drifters deployed monthly along the three transects shown in red beginning in August 2005 and tracked for 60 days. The surface currents are derived from satellite altimetry (sea surface height anomalies plus a mean field) using the geostrophic approximation. The technique follows that of Alvera-Azcárate et al. (2009), and the plot is courtesy of L. Zheng. (Color versions of figures available online at: http://www.ingentaconnect.com/content/mts/mtsj/2011/00000045/00000002.)

FIGURE 3 Temperature, salinity, chlorophyll, and colored dissolved organic matter observed across the continental slope and shelf by a July 2010 glider transect (courtesy of C. Lembke, CMS-USF profiler/glider team).

36

Marine Technology Society Journal

advection of nutrients onto the shelf and then both along and across the shelf, contributing to the local productivity evident along the bottom. Such occurrences for the West Florida Continental Shelf (WFS) are a general finding (e.g., Weisberg et al., 2004). How productivity occurring seaward from the shoreline may affect the near shore is evident from the evolution of a K. brevis red tide event in 2005. Figure 4 illustrates the observed movement of Karenia brevis (K. brevis) cells compared with a numerical circulation model simulation (Weisberg et al., 2009). We see that the connection between the initial, offshore bloom of K. brevis and the near shore was via the bottom Ekman layer. Thus, nearshore water properties can neither be understood nor predicated by nearshore observations alone. Connections, owing to the entire coastal ocean system, are involved. Not all years exhibit similar nearshore manifestations. When comparing two successive years of WFS initiated K. brevis blooms, we found that red tide was retained on the WFS in 2006, whereas it was exported to Florida’s east coast in 2007 (Figure 5) (Walsh et al., 2009). This study also demonstrated the connection across trophic levels. Because K. brevis makes its living, in part, by killing fish, the observed fish kills mimicked the retention and export findings for these two years. Turning our attention to the estuaries, we argue that their connections with the shelf are also omnipresent and fully three-dimensional. Figure 6 shows the subtidal, mean circulation at the mouth of Tampa Bay with inflows of coastal ocean water at depth and outflows of estuarine water at the surface (Weisberg and Zheng, 2006). The flushing from Tampa Bay of

FIGURE 4 Observed K. brevis, red tide evolution in winter 2005 (top and bottom right) along with a model simulation of virtual drifter movements from January 13, 2005 to February 18, 2005 (left). The virtual drifter color coding represents sigma coordinate level (deep blue being near bottom and deep red being the surface). Thus, virtual drifters initiated near-bottom offshore upwelled toward the surface as they approached the shoreline (from Weisberg et al., 2009). (Color versions of figures available online at: http://www.ingentaconnect.com/content/mts/mtsj/2011/00000045/00000002.)

FIGURE 5 Simulated trajectories for virtual drifters deployed on the WFS in 2006 (right) and 2007 (left). The virtual drifter color coding represents the drifter age after deployment (deep blue being the time of deployment and deep red being some 70 days later; from Walsh et al., 2009). (Color versions of figures available online at: http://www.ingentaconnect.com/content/mts/mtsj/2011/00000045/00000002.)

river- and land-derived inputs is primarily a consequence of this densitydriven, estuarine circulation. Similarly, the advection of coastal ocean water properties into the estuary (including plankton and larvae) is also a consequence of this density-driven, estuarine circulation. Not only are the overall water properties of the estuary determined in a three-dimensional manner, but so are the water properties on much smaller scales. Consider, for instance, the flushing of a residential channel (Figure 7). The flushing is virtually nil with tides alone, but with the introduction of winds (that cause a three-dimensional circulation), the flushing is rapid (Zheng and Weisberg, in preparation). These foregoing illustrations demonstrate that coastal ocean state variables (i.e., anything for which an equation of state may be written such temperature, salinity, nutrients, phytoplankton, larvae, fish, specific pollutants, etc.) result from spatial and temporal connections between the coastal ocean and both the deep ocean and the estuaries, along with trophic level interactions manifest, in large measure, by these physical connections. It follows that the implementation of water quality regulations or concepts such as Ecologically Based Management or Marine Spatial Planning must be predicated on understanding these connections. In other words, we must understand how the coastal ocean system works, in a comprehensive, multidisciplinary manner, if we are to manage it, plan for its utilization, and predict the consequences of human actions.

A Pathway Forward Understanding the workings of the coastal ocean system is a large, but tractable problem, given a few guiding March/April 2011

Volume 45

Number 2

37

FIGURE 6 The mean (September to November 2001) velocity component normal to the Egmont Channel and South Pass cross sections at the mouth of Tampa Bay and the associated mean salinity distributions (from Weisberg and Zheng, 2006).

principles. The first is that any study must be complete enough, in other words, comprehensive and multidisciplinary. Fisheries provide an example. Fisheries science cannot advance by just studying fish. Instead, the entire context in which fish make their living must be understood. Being that fish (as upper trophic level creatures) integrate the effects of all other coastal ocean processes, if we can understand and manage fisheries resources in a scientifically defensible way, then we can make application to most other coastal ocean problems of societal importance. Fisheries, in other words, present an excellent focal point for coastal ocean science, but only if engaged in a comprehensive, multidisciplinary manner.

38

The second guiding principle is that we must combine extensive observations with science-based models, in essence, adhere to the scientific method. There can never be enough observations, and this requires models for integration, but models without observations for initialization, boundary values, and veracity testing are nearly useless. The two (observations and models) must be coordinated. Third, no single sensor (for measuring any of the state variables) or sensor delivery systems (for housing sensors, such as moorings, profilers, gliders, ships, side scan sonars, high-frequency (HF) radars, satellites, etc.) are adequate. A judicious mixture of these is needed, plus the continual investment in evolv-

Marine Technology Society Journal

ing new technologies. Fourth (and similar to observations), no single model is adequate. In analogy to hurricane landfall prediction, we require an ensemble of models for oceanatmosphere interactions, ocean circulation (nested deep ocean, coastal ocean, and estuarine), and the complex biological interactions that, together with the circulation, comprise ecology. There is much to do, and this requires many partners amongst the agencies (federal, state, and local), the academics, and the private sector, each with their own expertise and capabilities, brought together in a multidisciplinary, multiinstitutional, cooperative manner. The starting point is with existing observing and modeling resources, which must be sustained and built upon. Numerical weather forecasting provides an example of how success is achievable. The early results when engaged in the 1950s were poor, but as more observations were steadily added and sustained, model workings were better understood (e.g., Thompson, 1990) and computational power increased; our ability to predict weather steadily improved to the point where most television viewers now eagerly await the evening update. The same can be applied to the development of coordinated ocean observing and modeling systems for the Gulf of Mexico, its coastal ocean regions and estuaries, with the added benefit of further improving marine (and adjacent land) weather prediction through more observations. Being that a major limitation to modeling the coastal ocean circulation (and hence all water properties) is the coastal ocean wind field (e.g., He et al., 2004), improving upon coastal ocean weather prediction goes hand in glove with improving predictions for all coastal ocean state variables.

FIGURE 7 The top two panels show the flushing of a simulated tracer from a residential channel under the action of tides alone. The top left is the initial condition, and the top right shows the tracer concentration 19 days later. The bottom two panels show the flushing of a simulated tracer from a residential channel under the action of both tides and winds. The bottom left is the initial condition, and the bottom right shows the tracer concentration 3 days later. Color coding represents normalized tracer concentration, red being one and deep blue being zero. Flushing is slow by tides alone, whereas it is fast under the influence of winds (because winds generate a fully three-dimensional circulation). (Color versions of figures available online at: http://www.ingentaconnect.com/ content/mts/mtsj/2011/00000045/00000002.)

The utility of sustaining and building upon existing observing and modeling resources was recently demonstrated through responses to the Deepwater Horizon oil spill. External to the Incident Command structure, several academic and private sector groups mar-

shaled their resources in assistance. At the University of South Florida College of Marine Science, we had a capability for observing and modeling the eastern Gulf of Mexico (with emphasis on the WFS) built over many years, and these resources were engaged immediately

after the drill rig exploded and sank (e.g., Liu et al., 2011). Our starting point was a WFS nowcast/forecast circulation model, whose particle trajectory forecasts were modified for the oil spill. Within a few days of the incident, we served oil spill trajectory forecasts on the Internet (http://ocgweb. marine.usf.edu) and distributed these broadly as PowerPoint presentation briefings, along with other information. Within a few weeks, we expanded our tracking to include a total of six different models (five by others whose circulation field forecasts were available on the Internet) to produce an ensemble of surface oil trajectories. Our own WFS model fields and trajectory forecasts became part of the regular, daily National Oceanic and Atmospheric Administration (NOAA) forecasts used by the Incident Command, and we made the entire ensemble available to the agencies for their use. A representative example of a forecast from four of the six models employed is provided in Figure 8. We also tracked oil beneath the surface with our WFS model, and these trajectories were used by academic scientists aboard the R/V Weatherbird II to select stations at which to successfully sample subsurface hydrocarbons that were subsequently fingerprinted to the Deepwater Horizon spill source (D. Hollander and E. Peebles, personal communication). Along with modeling, we served real-time observations of velocity on the WFS from moorings and HF radar, analyses of surface geostrophic currents from satellite altimetry (also at http://ocgweb.marine.usf.edu), plus oil location from analyses of ocean color (http://optics.marine.usf. edu/events/GOM_rigfire). The combination of observing and modeling resources, plus knowledge gained

March/April 2011

Volume 45

Number 2

39

FIGURE 8 An ensemble oil trajectory forecast example using four of six diagnosed models. The oil location was initialized using a 6/12 satellite image interpretation (courtesy of C. Hu) and the forecast used either NOAA NWS forecast winds (for the left panels) or Navy NOGAPS winds (for the right panels). Black and purple represent virtual drifter initial positions and forecast swept areas, respectively (from Liu et al., 2011). (Color versions of figures available online at: http://www.ingentaconnect. com/content/mts/mtsj/2011/00000045/00000002.)

from previous scientific investigations, enabled us to provide guidance on what to expect near term, along with explanations for why the trajectories behaved as they did. These contributions, at a time of crisis, could not have been realized without the commitment over many years for building and sustaining a coastal ocean observing system. The framework for building and sustaining coastal ocean observing systems (COOS) nationwide presently exists in the context of an Integrated Ocean Observing System (IOOS) that was promulgated in May 2002 by the (former) interagency group, Ocean.US. IOOS, now under the auspice of NOAA, was conceived to consist of regional associations (RA), each with Regional Coastal Ocean Observing Systems (RCOOS). Whereas our University of South Florida (USF) WFS COOS activities predated

40

IOOS and included funds marshaled from a variety of sources, they continue under the aegis of the Southeast Atlantic Coastal Ocean Observing Regional Association (SECOORA) because SECOORA was predicated on the connectivity between the Gulf of Mexico and the southeastern United States by virtue of the Loop Current– Florida Current–Gulf Stream system. SECOORA cooperates with the neighboring RAs, the Gulf of Mexico Coastal Ocean Observing System (GCOOS), and the Mid-Atlantic Coastal Ocean Observing Regional Association (MACOORA). Despite its 2004 endorsement by the U.S. Commission on Ocean Policy (http://www.oceancommission.gov), the 2002 Ocean.US vision (with 11 RAs in total) putters at idle due to dwindling (versus increasing) funding. The general concept remains valid; its utility is made clear by the Deepwater

Marine Technology Society Journal

Horizon oil spill, but, in view of its failure to take hold, its implementation must be reassessed if IOOS is to garner the political will to better serve the nation. For IOOS to be successful, there must be a reassessment of the roles to be played by all of the partners. Whereas operations are largely the purview of the agencies, research and development are mainstays of the academics and the private sector. In the same way that all politics are local, so are many aspects of applied research. Our knowledge of local waters has been generated largely by local scientists with more than a casual interest in those waters. At a time when accountability is the operant word, we must also empower those who demonstrate performance, and this should begin by sustaining what has been proven to be useful and building upon such proven resources. In other words, tearing down to build anew is not acceptable if existing infrastructure is scientifically defensible and demonstrates utility. A corollary is that practitioners must be involved in the process of observing system decision-making. Finally, it is my opinion that IOOS took a wrong turn when the leadership prioritized interoperability and data management over observations and models. The scientific method begins with observations, and hypothesis testing begins with models. Without these two elements, there is little to be interoperable or to manage. The Deepwater Horizon oil spill offers a case in point. We were able to quickly adapt our information feed for use by the Unified Command. The limitation was the science (observations and models), not the management. We need to reposition the horse before the cart. IOOS is poised for success with reassessment, as suggested above.

The Gulf of Mexico and the southeastern United States are particularly vexing, because despite the potential threat from (necessary) oil and gas activities, the IOOS infrastructure there tends to lag that in other parts of the nation. Two improvements are necessary. The first is to increase coordinated observing and modeling coverage for the coastal ocean where societal impacts are potentially the greatest. The second is to expand into the deep Gulf of Mexico where oil and gas activities will focus in the future. Items that are needed include (and are not limited to) are as follows: ■ telemetering surface moorings for real-time surface meteorology and in-water currents, temperature, salinity, plus maturing biological and chemical sensors; ■ subsurface moorings for in-water variables; ■ high-frequency radars for sea surface currents and waves; ■ gliders and profiling floats for threedimensional, synoptic mapping of a broad suite of ocean state variables; ■ support for acquisition and analyses of remotely sensed satellite information; ■ support for an ensemble of physical and coupled physical-biological (i.e., ecological) models tailored to Gulf of Mexico needs (i.e., the deep ocean, the connections between the deep ocean, the coastal ocean and the estuaries, and the connections between the Caribbean, Gulf of Mexico and the southeastern United States, plus improvements to mesoscale atmosphere models); ■ support for passive and active acoustics studies of living marine resources; ■ support for new sensor developments and testing;

support for structural mapping of the ocean bottom; ■ research vessels for deployments, maintenance, and emergency response; ■ port facilities to support research vessels, sensors and sensor delivery systems, and data transmission systems; ■ support for data management and dissemination; and ■ applications of sensors on oil and gas platforms where deemed useful. Whereas all of these items are important, the priority rests with building a coordinated set of observations and models for system understanding and prediction. Sustaining existing and implementing new resources should be based on scientifically defensible rationale, not convenience. Hence, while platforms of opportunity are attractive, there must be a strong scientific rationale for investing observing capital at such locations. ■

Closing Remark The coastal ocean is where society meets the sea. It is a complex, interconnected system; the workings of which must be understood if we are to predict the consequences of human actions and distinguish these from natural occurrences. Such understanding comes through adequate observations and hypothesis testing via science-based models, in other words, the application of the scientific method. Priority must therefore be given to implementing a coordinated, multidisciplinary program of coastal ocean observing and modeling, including the interactions that occur between the coastal ocean and the deep ocean and between the coastal ocean and the estuaries. That was the essence of

my testimony before the U.S. House of Representatives Committee on Natural Resources, Subcommittee on Insular Affairs, the Oceans and Wildlife on June 15, 2010 (at the height of the Deepwater Horizon oil spill), and it remains valid today. This is the pathway toward becoming better environmental stewards of the coastal ocean. Only in this manner will we be better prepared to assess the potential impacts of pollutants, engage in ecologically based management and marine spatial planning, or deal more effectively with future, unintended assaults on the coastal ocean environment.

Acknowledgments Together with a long-standing commitment to coastal ocean observing and modeling, this commentary was strongly influenced by the Deepwater Horizon oil spill. My immediate USF colleagues in this endeavor were Drs. C. Hu, Y. Liu, and L. Zheng. Other members of my immediate research team, Messrs. J. Donovan, J. Law, D. Mayer, and P. Smith, unselfishly committed time and energy to the crisis response, as did my USF CMS Center for Ocean Technology colleague Mr. C. Lembke and his profiler/glider team. Previous research team members (Dr. A. Barth, Dr. A. Alvera-Azcárate, Dr. J. Virmani, and Mr R. Cole) also contributed substantively. Our WFS observing and modeling activities benefited from many funding sources over the years. Present support is by the State of Florida (USF), NOAA grants NA07NOS4730409 (IOOS through SECOORA) and NA06NOS4780246 (ECOHAB through FWC), ONR grants N00014-05-1-0483 and N00014-10-1-0785, NSF grant

March/April 2011

Volume 45

Number 2

41

OCE-0741 705, and F IO g ran t 4710110105 (via BP research funds). Previous (pre-2005) earmarked funds through NOAA (Coastal Ocean Program) and ONR (the SEACOOS Program) were essential for building infrastructure (hence, not all earmarks are bad, as demonstrated by our ability to respond to the Deepwater Horizon oil spill). This is Center for Prediction of Red Tide (CPR) contribution 12.

Author: Robert H. Weisberg College of Marine Science University of South Florida St. Petersburg, FL 33701 Email: [email protected]

References

with implications for coastal fisheries over both source regions of the West Florida Shelf and within downstream waters of the South Atlantic Bight. Progr Oceanogr. 80:51-73. doi: 10.1016/j.pocean.2008.12.005. Weisberg, R.H., Barth, A., Alvera-Azcárate, A., & Zheng, L. 2009. A coordinated coastal ocean observing and modeling system for the West Florida Shelf. Harmful Algae. 8:585-98. doi: 10.1016/j.hal.2008.11.003. Weisberg, R.H., He, R., Kirkpatrick, G., Muller-Karger, F., & Walsh, J.J. 2004. Coastal ocean circulation influences on remotely sensed optical properties: A west Florida shelf case study. Oceanography. 17:68-75. Weisberg, R.H., & Zheng, L. 2006. Circulation of Tampa Bay driven by buoyancy, tides, and winds, as simulated using a finite volume coastal ocean model. J Geophys Res. 111:C01005. doi: 10.1029/2005JC003067.

Alvera-Azcárate, A., Barth, A., & Weisberg, R.H. 2009. The surface circulation of the Caribbean Sea and the Gulf of Mexico as inferred from satellite altimetry. J Phys Oceanogr. 39:640-57. doi: 10.1175/2008JPO3765.1. He, R., Liu, Y., & Weisberg, R.H. 2004. Coastal ocean wind fields gauged against the performance of a coastal ocean circulation model. Geophys Res Lett. 31:L14303. doi: 10.1029/2003GL019261. Liu, Y., Weisberg, R.H., Hu, C., & Zheng, L. 2011. Tracking the Deepwater Horizon oil spill: A modeling perspective. EOS, Trans Am Geophys Un. 92(6):45-6. Thompson, P.D. 1990. Charney and the revival of numerical weather prediction. In: The Atmosphere Challenge: The Science of Jule Gregory Charney, eds. Lindzen, R.S., Lorenz, E.N., & Platzman, G.W., 93-119. Boston, MA: American Meteorological Society. Walsh, J.J, Weisberg, R.H., Lenes, J.M., Chen, F.R., Dieterle, D.A., Zheng, L., … Landsberg, J.H. 2009. Isotopic evidence for dead fish maintenance of Florida red tides,

42

Marine Technology Society Journal