Скачать 1.55 Mb.
where N is an index of stratification strength known as the buoyancy frequency, is the water density, g is the gravitational acceleration (9.8 m/s2), and V/z is the vertical shear of horizontal current. The gradient Richardson number, Ri, expresses the ratio of turbulence suppression by stratification (numerator) relative to vertical shear production of turbulence (denominator). When Ri > ¼, turbulence is suppressed, and vertical transport of oxygen from surface to bottom layers by turbulent mixing is unlikely to occur. Thus, strong vertical density gradients (for example, when freshwater sits on top of salty water) and/or weak current shears can suppress vertical mixing and be favorable to hypoxia. Key physical factors that produce stronger vertical density gradients (/z) and thus reduce vertical mixing include freshwater inputs from rivers or precipitation, warmer surface temperatures from absorption of solar radiation or sensible heat input, and near-bed suspended sediment (which causes benthic stratification). Factors responsible for producing enhanced vertical shear (V/z) and enhanced vertical mixing include tidal and wind-driven currents, inertial waves, internal tides, surface waves and Langmuir cells (Kantha and Clayson, 2000). Although no field studies of vertical mixing by microstructure measurements of the turbulent dissipation rates of velocity, salinity and temperature fluctuations have been reported for the NGOM, many of the physical mechanisms described on the New England shelf (MacKinnon and Gregg, 2005) and in Monterrey Bay (Carter et al., 2005) are at play on the NGOM as well.
While the tributaries within the Mississippi River basin are the sources of nutrient loading to the river trunk, the distributaries within the Mississippi Delta are critical to the final dispersal of nutrients, buoyancy and sediment into the Gulf of Mexico. The multiple distributary mouths of the Mississippi and Atchafalaya Rivers are, for the most part, highly stratified “salt wedge” estuaries, and their combined effluent debouches onto the shelf as a discrete layer of fresh water that is spread into the surface layer. Exceptions occur where smaller distributaries enter shallow bays where salinity is nearly uniform from top to bottom. Total buoyancy fluxes are, of course, proportional to river discharge and cause the turbulence suppressing stratification of the upper water column that is strongly implicated in hypoxia. In most inner shelf environments, tidal currents are the major source of mixing, and the position of temperature fronts (sharp horizontal temperature gradients) can often be accurately predicted from the h/Ut3 criterion of Simpson and Hunter (1974), where h is the local depth and Ut represents the depth-averaged tidal velocity. Unfortunately, the Simpson-Hunter criterion of tidal mixing has not yet been mapped for the northern Gulf of Mexico. Nevertheless, it is generally agreed that tidal mixing over the Louisiana-Texas shelf is very weak because the tidal range is only about 40 cm and tidal currents typically do not exceed 10 cm/s (Kantha, 2005). So the contribution of tidal mixing to the vertical exchange of oxygen is minimal over the shelf, particularly off the mouths of the larger distributaries, such as Southwest and South Passes, which debouch into deep water. Wind-driven currents are stronger than tidal currents but occur episodically (Ohlmann and Niiler, 2005). Winds also cause breaking and white capping waves as well as vertical circulation (Langmuir) cells (Thorpe, 2004) that contribute to mixing in the upper water column.
The hydrologic regime of the Mississippi River and the spatial distribution and timing of freshwater inputs to the shelf relative to the occurrence of energetic currents and waves are critical to vertical mixing intensity, stratification, and hypoxia. These influences were recognized in the CENR report (Rabalais et al., 1999). Using oxygen measurements within 2 m of the bottom and vertical profiles of temperature and salinity collected during the 1992-1994 LaTex experiment on the Louisiana-Texas shelf and during the 1996-1998 NECOP (Northe36astern Gulf of Mexico Chemical Oceanography Program) in the region east of the Mississippi delta and north of Tampa Bay, Belabassi (2006) performed an evaluation of the empirical relationships between the maximum value of the buoyancy frequency Nmax in the water column, bottom silicate concentration as a proxy of phytoplankton remineralization, and the occurrence of hypoxic waters (< 2 mg/L) or low-oxygen waters (< 3.4 mg/L). She found that low-oxygen and hypoxic bottom waters only occurred when Nmax, evaluated at a vertical resolution of 0.5 m was greater than 40 cycles per hour (cph), which corresponds to a buoyancy period shorter than 1.5 minutes. This result confirms that strong density stratification is a prerequisite for hypoxia occurrence on the northern Gulf of Mexico shelf. She also found that low-salinity water from the Mississippi and Atchalafaya Rivers was generally the main contributor to stratification in spring and summer, although temperature was more important than salinity in determining stratification during summer at all depths west of Galveston Bay and at depths greater than 20 m between Galveston Bay and Terrebonne Bay. Interestingly, stations with strong stratification (Nmax greater than 40 cph) but low bottom silicate concentrations (less than 18 mmol m-3) did not have low-oxygen or hypoxic bottom waters. The analyses of Belabassi (2006) thus indicate that strong stratification (Nmax greater than 40 cph) is a necessary but not sufficient condition for bottom layer hypoxia; a second necessary condition for hypoxia occurrence is high bottom water remineralisation as indicated by the proxy of high concentrations of bottom water silicates (greater than 18 mmol m-3). Simply put, there cannot be hypoxia without both density stratification and degradation of labile organic matter.
Stow et al. (2005) attempted to disentangle the relative contributions of eutrophication and stratification as drivers of hypoxia in the NGOM. Their analysis indicates that the probability of observing bottom hypoxia increases rapidly when the top to bottom salinity difference reaches a threshold of 4.1. Stow et al. (2005) also showed that this salinity threshold decreased from 1982 to 2002. Concurrently, they highlighted that surface temperature had increased, while surface dissolved oxygen decreased, suggesting that changes in surface mixed layer properties may be partly responsible for oxygen decrease in the bottom layer.
Changes in Mississippi River hydrology and their effects on vertical mixing
By far the most important change in local hydrology has been the increased flow of the Atchafalaya River during the 20th century. Available data show that in the early 1900’s the discharge from the Atchafalaya River accounted for less than 15% of the combined Atchafalaya-Mississippi River discharge (Figure 4). This proportion progressively increased to reach about 30% in 1960, peaked at 35% in 1975 and since then was reduced to 30% by means of regulatory measures (Bratkovich et al., 1994). To understand the significance of this change on circulation patterns and on the strength of stratification on the Louisiana-Texas shelf, it must be kept in mind that the Mississippi River plume enters the shelf near the shelf edge and typically does not extend to the bottom, even near the river mouth. On the other hand, the Atchafalaya River plume enters a broader shelf, is more diffuse, and extends to the bottom over a larger distance from the river mouth.
Figure 4: Change in the relative importance of the Atchafalaya flow to the combined flows from the Mississippi and Atchafalaya Rivers over the 20th Century. Reprinted from Bratkovich et al. (1994).
The short distances (10 to 30 km) separating Mississippi River delta passes from the shelf break facilitate the export of plume waters offshore and to the east by sporadic wind events or by eddies present on the upper continental slope, some of which may have been spun off by the Loop Current (Ohlmann and Niiler, 2005; Oey et al., 2005a, 2005b). The modeling study of Morey et al. (2003a) shows that a prime export pathway for river freshwater during the summer months is to the east, and offshore of the Mississippi River delta. During non-summer months, the main freshwater export pathway consists of a coastal jet flowing westward to Texas and then southward. Etter et al. (2004) estimate that 43% ± 10% of the Mississippi River discharge is carried westward to the Louisiana-Texas continental shelf, the remainder being carried offshore and/or eastward. While this proportion is slightly lower than the earlier estimate of 53% ± 10% from Dinnel and Wiseman (1986), both studies indicate that roughly half of the freshwater from the Mississippi River goes westward, toward the Louisiana-Texas continental shelf.
In contrast, 100% of the Atchafalaya River discharge of freshwater, nutrients and sediments is delivered to the Louisiana-Texas continental shelf. Moreover, the very broad shelf near Atchafalaya Bay implies longer residence times of this freshwater source on the shelf compared with freshwater from the Mississippi River delta. A “back-of-the-envelope” calculation helps capture the full significance of the increased Atchafalaya River flow. In the early 1900’s, for every 100 m3 of water discharged, 85 m3 took the Mississippi River delta route. Of these, roughly 42.5 m3 went westward and 42.5 m3 went offshore or eastward. The 42.5 m3 that went westward were added to the 15 m3 that took the Atchafalaya River route to give a grand total of 57.5 m3 of freshwater on the Louisiana-Texas continental shelf. By contrast, in the post-1970’s, for every 100 m3 of combined Atchafalaya and Mississippi River outflows, 70 m3 took the Mississippi River route. Of these, roughly 35 m3 went westward, and 35 m3 went offshore or eastward. The 35 m3 that went westward were added to the 30 m3 that took the Atchafalaya River route to give a grand total of 65 m3 of freshwater on the Louisiana-Texas continental shelf. This simple calculation reveals two things. First, it suggests that even in the absence of a temporal trend in combined Atchafalaya-Mississippi River freshwater discharge, the amount of freshwater delivered to the Louisiana-Texas continental shelf would have increased by 13% (65/57.5 = 1.13). Second and more importantly, it reveals that in the 1920s, the Atchafalaya River contributed about one quarter (15/57.5 = 0.26) of the freshwater discharge to the Louisiana-Texas continental shelf. Between 1920 and about 1960, the Atchafalaya River’s contribution markedly increased to about one half (30/65 = 0.46) of the freshwater discharge to the Louisiana-Texas continental shelf. While this probably made the Louisiana-Texas continental shelf more prone to hypoxia, the timing of this change occurred 15 to 20 years earlier than the onset of regular summer hypoxia (Section 2.1.1).
Future physical modeling studies are needed to investigate the effects of past and proposed future changes in the distribution of freshwater flows, including inputs to Atchafalaya Bay some 200 km to the west of the Mississippi River delta, on changes in the spatial distribution of surface salinity, temperature, and stratification on the Louisiana-Texas continental shelf and on the Mississippi Sound to the east of the birdfoot delta. Physical oceanographic models that can adequately answer such questions about the impacts of flow diversions already exist but have only been run using the post-1970s flow conditions (30% Atchalafaya River, 70% Mississippi River). One such modeling study by Hetland and DiMarco (2007) suggests that the freshwater plumes from the Atchafalaya and Mississippi Rivers are often distinct from one another (Figure 5) and that both contribute significantly to the development of hypoxia (Figure 1) on the shelf through their influence on stratification and nutrient delivery (Rabalais et al., 2002a). In addition, maps of observed surface salinity and satellite images of chlorophyll (e.g., figure 9), show the same result. It thus appears likely that increases in freshwater discharge from the Atchafalaya River and resulting increased stratification from the early 1900’s to the mid-1970’s have increased the area of the Louisiana-Texas continental shelf that is prone to bottom layer hypoxia.
Figure 5: Modelled surface salinity showing the freshwater plumes from the Atchafalaya and Mississippi Rivers during upwelling favorable winds (top panel) and during downwelling favorable winds 8 days later (bottom panel). Adapted from Hetland and DiMarco (2007).
Recently evolved plans for protecting coastal Louisiana (CPRA, 2007) propose significant diversions of the water, nutrients, and sediment outflow from the Mississippi River into the Gulf. Figure 6 illustrates a diversion scenario that involves redirecting a large part of the outflow into shallow bays upstream of the present day “bird’s foot” delta. This scenario could alter the shelf hydrodynamics, particularly if more of the buoyancy is directed into shallow water instead of the deep water off the active river mouths, which are near the shelf edge. It is important that three-dimensional numerical circulation models be applied to these scenarios. Future management strategies may be able to utilize engineered modulations of the timing of freshwater releases to coincide more closely with more energetic waves and current conditions, thereby reducing the strength of stratification (i.e., Ri). This approach will, of course, rely on engineering innovations and effective diversion management. The opportunity exists for EPA and other federal and management agencies to urge flow diversion strategies that also consider the goal of reducing the volume and bottom area of hypoxic waters on the NGOM shelf without endangering other estuarine and coastal waters. The CPRA/U.S. Army Corp of Engineers proposals also highlight the need for interagency coordination and for an integrated approach to management strategies for jointly addressing multiple issues including hypoxia, coastal protection, and coastal inundation.
Figure 6: Proposed diversions of Mississippi effluents for coastal protection. From Coastal Protection and Restoration Authority (CPRA) of Louisiana, 2007 Integrated Ecosystem Restoration and Hurricane Protection: Louisiana’s Comprehensive Master Plan for a Sustainable Coast. CPRA, Office of the Governor (La) 117 pp.
Zones of hypoxia controls:
The resulting stratified region influenced by the Mississippi and Atchafalaya River plumes exerts strong control on the extent and spatial distribution of hypoxia and is an important factor in determining where hypoxia may occur (Rabalais and Turner, 2006). The buoyancy fluxes from the rivers also contribute to regional circulation in the form of baroclinic flows (Morey et al., 2003a, 2003b). Following a similar line of reasoning used in earlier work by Rhoads et al. (1985) off the mouth of the Changjiang (Yangtze) River, Rowe and Chapman (2002) defined three zones of hypoxia control in the NGOM. The boundaries between these three zones are admittedly fuzzy, and change through time; however Figure 7 illustrates the SAB Panel’s view of these concepts as represented by 4 zones. In zone 1, which is most proximal to river mouth sources, strongly stratified and light- as well as nutrient-limited, respiration of organic carbon coming both directly from the river efflux and from nutrient-dominated eutrophication dominates. The relative importance of these organic carbon sources as the cause of hypoxia remains somewhat uncertain, although the model of Green et al. (2006b) indicates a major dominance by in situ phytoplankton production even in the immediate plume of the Mississippi River. In the intermediate zone 2, stratification is also strong; light limitation is less than in zone 1; very high rates of phytoplankton production occur; and water column respiration fuels bottom layer hypoxia. Farther along the coast from the river mouths but within the low-salinity coastal plume (zone 3), local phytoplankton production is less, but labile organic matter may have been imported from zone 2 and deposited on the bottom. In zone 3, stratification remains strong, and oxygen consumption in the sediment is more important than water column respiration in driving hypoxia. Zone 4 depicts the highly productive, coastal current, as suggested by Boesch (2003).
Boesch (2003) strongly criticized the physical, biological and chemical reasoning behind the delineation of the Louisiana-Texas continental shelf into these three distinct zones of hypoxia control. He also argued that these zones did not capture well the physics and biology of the Louisiana coastal current, which is characterized by low salinities and high nutrient and chlorophyll levels (Wiseman et al., 2004). Nevertheless, Rowe and Chapman (2002) stimulated new research into the role that stratification plays in the reduction of vertical mixing rates and the flux of oxygen through the pycnocline in the regions of the Louisiana-Texas continental shelf under the influence of the Mississippi and Atchafalaya River plumes. Using realistic three-dimensional physics (equation 1) with simple representations of water column and benthic respiration for the zones A, B and C of Rowe and Chapman (2002), Hetland and DiMarco (2007) were able to represent the bottom area, thickness, and volume of hypoxic waters over the NGOM fairly well.
Figure 7: An illustration depicting different zones (Zones 1-4, numbered above) in the NGOM during the period when hypoxia can occur. These zones are controlled by differing physical, chemical, and biological processes, are variable in size, and move temporally and spatially. Diagram created by D. Gilbert.
So far as we are aware, time series measurements of physical oceanographic parameters are inadequate to support or refute hypotheses regarding changes in shelf circulation, stratification, and vertical mixing during the 20th century. Initial planning for a Gulf of Mexico Coastal Ocean Observing System (GCOOS) has begun (for additional information see: http://www.gcoos.org). As these GCOOS plans continue to evolve and implementation begins over the next few years, it is important that physical parameters relevant to oxygen dynamics be included among the measurements. Empirical parameterizations of vertical eddy diffusivity Kz as a function of vertical shear and density stratification are available for shallow continental shelf environments (MacKinnon and Gregg, 2005). These parameterizations enable quantification of vertical mixing (term 4 in equation 1) with vertical shear measurements from moored Acoustic Doppler Current Profilers (ADCPs) and vertically profiling conductivity, temperature, and depth instrumentation (CTDs) tethered on a cable. Ship-based microstructure measurements of the turbulent rates of dissipation of velocity, salinity, and temperature fluctuations (Gregg, 1999) should also be conducted occasionally to complement the moored ADCP and profiling CTD measurements. Physics-based models of ocean mixing and turbulence exist today and are part of 3-D circulation models (Mellor and Yamada, 1982). These models need to be rigorously tested using ADCP, CTD, and microstructure data because vertical mixing is the most important physical process to model correctly when hypoxia is under consideration.
Shelf circulation: local versus regional
Circulation in the NGOM can be considered on two scales: Gulf-wide deep-sea circulation and shelf circulation near the coast. Among the most prominent features of the large-scale Gulf-wide circulation are the Loop Current and the Loop Current Eddy System (Oey et al., 2005a, 2005b). Although these features impinge on and affect the outer shelf, Rabalais et al. (1999) conclude that local wind forcing and buoyancy are more important to shelf circulation inshore of the 50 meter isobath. Direct ship-board observations by Jarosz and Murray (2005) during five separate cruises led those authors to conclude that the momentum balance on the inner and mid shelf to the west of the active birdfoot delta is indeed dominated by wind stress. During summer, alongshore sea-surface slope caused by buoyancy forcing was also important in forcing currents. On the 20 m isobath off Terrebonne Bay, ADCP measurements (Wiseman et al., 2004) show periods of several days with negligible vertical shear followed by other periods of a few days with much more elevated vertical shear and reduced density gradients, suggestive of more intense vertical mixing.
Several physical oceanographic models taking into account the crucial baroclinic effects that typify the Louisiana-Texas continental shelf are now available (e.g., Morey et al., 2003a, 2003b; Zavala-Hidalgo et al., 2003). The model results of Hetland and DiMarco (2007) show that the plume from the Mississippi River, which enters the shelf near the shelf edge, forms a recirculating gyre in Louisiana Bight and does not interact with the seabed, whereas the Atchafalaya River plume interacts with the shallow coastal topography (Hetland and DiMarco, 2007). Both plumes respond directly to local winds and are advected seaward during upwelling-favorable winds (Figure 5). The distinct plumes from the Mississippi and Atchafalaya Rivers influence the spatial pattern of bottom hypoxia on the Louisiana-Texas continental shelf. This influence is clearly seen on the 1985-2005 map of hypoxia frequency of occurrence (Figure 1) and is even more obvious in certain years (e.g., 1986, Rabalais and Turner 2006). Given this interaction, planned diversions of Mississippi River and Atchafalaya River flow may alter shelf circulation and the spatial pattern of bottom hypoxia. Applications of 3-D baroclinic models to future scenarios such as that portrayed in Figure 6 are thus important to planning for future strategies for coastal restoration (CPRA, 2007).
In their analysis of low-frequency (occurring over a time scale greater than 24 hours) currents over the shelf, Nowlin et al. (2005) distinguished between currents that respond within the “weather band” of 2-10 days and those within the mesoscale band of 10-100 days corresponding to large-scale eddies off the shelf. Inshore of the 50 m isobath, the local winds within the weather band dominated and drove currents from east to west during non-summer months influenced by the passage of frontal systems. Current fluctuations seaward of the 50 m isobath were primarily within the mesoscale band and predominantly oriented from west to east but with high variability. Along-shelf and across-shelf currents in the upper layer over the inner shelf, as reported by Nowlin et al. (2005), averaged about 10 cm/s and 1 cm/s, respectively. Over the outer shelf and near the seabed, flows were weaker.
|Board of Elections – Advisory Opinion 1398 – 1399||Panel Report on the draft Replacement London Plan|
|Draft: August 2July xxanuary x19, 20065 abstract [For presentation at the Advisory Group on apec financial System Capacity Building meeting at Seattle, February 28, 2007.]||Draft – do not quote or distribute|
|Ranch hand advisory committee||Medical Devices Advisory Committee|
|Positions, training & advisory roles||Veterinary medicine advisory committee|
|Advisory committee on immunization practices||National Vaccine Advisory Committee (nvac)|