ARTICLE IN PRESS Deep-Sea Research II 55 (2008) 2541–2554

Contents lists available at ScienceDirect

Deep-Sea Research II journal homepage: www.elsevier.com/locate/dsr2

Physical oceanographic conditions in the deepwater Gulf of Mexico in summer 2000–2002 Ann E. Jochens , Steven F. DiMarco Department of Oceanography, Texas A&M University, 3146 TAMU, College Station, TX 77843-3146, USA

a r t i c l e in fo

abstract

Article history: Accepted 9 July 2008 Available online 5 September 2008

The circulation and distribution of water properties in the water column of the Gulf of Mexico influence the flux of carbon to the benthic environment. The eddy field of the upper 1000 m creates environmental conditions that are favorable for biological productivity in an otherwise oligotrophic subtropical ocean. This eddy field results in the transport of nutrients and organic matter into the photic zone through cross-margin flow of shelf waters, upwelling in cyclones, and uplift from the interaction of anticyclones with bathymetry. These conditions then allow the productivity that becomes a possible source of carbon to the benthos. Data from four cruises during summers of 2000–2002 are used to describe the currents and water property distributions in the deepwater Gulf of Mexico, which consists of water depths greater than 400 m. Comparisons are made to historical data sets to provide an understanding of the persistence of the characteristics of the Gulf and the processes that occur there. The currents in the Gulf are surface intensified, have minimum in 800–1000 m depths, and also exhibit bottom intensification, especially near sloping topography. Historical time series records show current speeds near-bottom reach 50–100 cm s 1. At basin scales, these currents tend to flow cyclonically (counter-clockwise) along the bathymetry. These near-bottom, episodic, high-speed currents provide a mechanism for the transport of organic material, in both large and small particle sizes, from one benthic area to another. The distributions of temperature, salinity, nutrients, and dissolved oxygen during the study appear to be unchanged from historical findings. The source waters for the deep Gulf are the water masses brought into the Gulf by the Loop Current system. The properties in the upper 100–200 m are the most variable of the water column, consistent with their proximity to wind mixing, river discharge mixing, and atmospheric influences. Below 1500 m, there are no major horizontal variations in these water properties. & 2008 Elsevier Ltd. All rights reserved.

Keywords: Hydrography Water masses Ocean currents Gulf of Mexico

1. Introduction Throughout much of the world’s ocean, the supply of organic matter to the seabed depends substantially on the biological production in the overlying water column (e.g., Muller-Karger et al., 2005). This is true of the deepwater Gulf of Mexico (GOM), although there also are important inputs of carbon from the chemosynthetic communities associated with hydrocarbon seeps in the GOM (e.g., MacDonald, 2002). The circulation and the resulting distribution of nutrients and biomass within the water column ultimately influence the carbon flux to the underlying sediments. As part of the Deepwater Gulf of Mexico Benthic project (DGOMB), physical oceanographic data were collected within the

 Corresponding author. Tel.: +1 979 845 6714; fax: +1 979 847 7789.

E-mail address: [email protected] (A.E. Jochens). 0967-0645/$ - see front matter & 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.dsr2.2008.07.003

water column on each of the DGOMB cruises in 2000–2002. These data provide information about the physical environment above each benthic station. These data do not directly measure the influence of the circulation and water-column processes on the benthic communities because there is a temporal lag between the productivity of the upper waters and the flux of carbon into the benthos and there can be horizontal transport of carbon into or out of the water column above a benthic station that is not directly measured by the DGOMB station data. However, these data provide information on the physical environment of the water column that influences the benthic environment. An understanding of this physical system, then, can provide insights into the processes at work in the benthic habitat that are discussed in other papers of this volume. The GOM is a semi-closed basin with both broad and narrow continental shelves surrounding a deep abyss reaching 3800 m (Fig. 1). The earliest attempt at a comprehensive synthesis of the physical oceanography of the GOM is found in Capurro and Reid

ARTICLE IN PRESS 2542

A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

(1970). Since then, and particularly in the last 15 years, the study of the physical oceanography of the GOM has greatly accelerated. The present state of the knowledge of the circulation of the GOM is summarized in a recent comprehensive synthesis edited by Sturges and Lugo-Fernandez (2005). The major energy source for driving the circulation in the Gulf is the Loop Current system, consisting of the Loop Current itself and the Loop Current eddies (e.g., Elliot, 1982; Nowlin and McLellan, 1967; Forristall et al., 1992; Nowlin et al., 2000, 2001; DiMarco et al., 2005; DeHaan and Sturges, 2005). The Loop Current enters the Gulf through the Yucatan Channel, extends northward variable distances into the eastern Gulf, and then turns eastward to exit through the Straits of Florida. It is a surfaceintensified current reaching speeds of 200 cm s 1 (e.g., Cooper et al., 1990). The Loop Current and its associated LCEs are energetic, upper-layer current features and can be observed in sea surface height (SSH) fields (Plate 1). Note in Plate 1 that the Loop Current (SSH 450 cm) was a permanent feature present in all years; however, the extent of penetration of its northern boundary into the GOM was variable. The extension of the Loop Current far into the GOM is a precursor to the separation of a Loop Current Eddy (LCE; e.g., Vukovich, 1988; Sturges and Leben, 2000; Leben, 2005). An LCE is an anticyclonic (clockwise rotating) circulation feature, with diameters often of 200–400 km when newly shed from the Loop Current (Elliot, 1982). After separation, the LCEs move into the western GOM, break into parts, and eventually decay (e.g., Biggs et al., 1996; Hamilton et al., 1999; Lewis and Kirwan, 1985). Note in Plate 1A the decaying remnants of LCE Indigo centered near 261N 951W and LCE Juggernaut centered near 23.51N 921W. The Loop Current, LCEs, and the separation process is thought to drive the lower-level circulation through topographic Rossby wave forcing and/or other mechanisms (e.g., Hamilton, 1990; Hamilton and Lugo-Fernandez, 2001; DeHaan and Sturges, 2005; Welsh and Inoue, 2000). Of great interest to issues of productivity are the presence of both anticyclonic and cyclonic eddies over the slope near the shelf edge. These features can create conditions of cross-margin flow or

upwelling that enrich the biological productivity of the deep water regime (e.g., Biggs et al., 2005; Biggs and Mu¨ller-Karger, 1994; Mu¨ller-Karger et al., 1991; Jochens and Biggs, 2006; Sahl et al., 1997; Lewis and Kirwan, 1985). The cyclonic (counterclockwise rotating) eddies are seen as the closed circulation regions of low SSH (blues in Plate 1). Note in Plate 1 that the region near DeSoto Canyon (e.g., near 291N 87.51W) had cyclones or anticyclones near the shelf edge in all of the summers with DGOMB sampling. This condition was present in this region in the 2 years preceding the DGOMB study (Belabbassi et al., 2005), so it likely is a common occurrence. Off DeSoto Canyon, these features can move low salinity, biologically productive Mississippi River discharge off the shelf into deeper water (Biggs et al., 2008; Mu¨ller-Karger et al., 1991; Belabbassi et al., 2005; Walker et al., 2005), with consequent influence on the productivity of the deep water benthos. Similar conditions exist in the northwest GOM, where many of the anticyclonic LCEs decay near the shelf edge and spin up adjacent cyclonic eddies. Examination of a monthly time series of SSH fields throughout the years indicates these small slope eddies (cyclonic or anticyclonic) are typical of the slope area (Hamilton, 1992; Hamilton et al., 2002; Leben, 2005 and companion CD-ROM). Mu¨ller-Karger et al. (1991) observed a persistent cyclone near the US–Mexico border in multi-year data sets, including ocean color imagery. Time scales of these features can be on the order of months, persistent enough to have biological impacts. Cyclones also spin up in association with the Loop Current or LCEs in deep water (e.g., see the cyclone in Plate 1D centered near 251N 871W northwest of and adjacent to the Loop Current). The centers of cyclones are also regions of upwelling. This can increase productivity in the surface waters with consequent increase in delivery of organic matter to the benthos. Although the upper waters of the Loop Current, LCEs, and anticyclones are nutrient-poor, these features can bring nutrient-rich deep water up into the photic zone when they interact with the topography or cyclones (e.g., Biggs et al., 2005; Jochens and Biggs, 2006). Wind stress forcing is second only to the forcing by the Loop Current system in providing energy to the upper ocean circulation

Fig. 1. Geographical setting of the deepwater Gulf of Mexico. Shown are locations of selected CTD/bottle stations on the DGOMB cruises in summers 2000–2002 (all except Sites 1 and 2) and of 300-kHz ADCP mooring stations on DGOMB cruises in June 2001 (MT3, S36, S42, and MT6) and June 2002 (Sites 1 and 2). The location of MMS mooring I1, deployed August 1999 through August 2000, is also shown. Bathymetric contours shown are 200, 1000, 2000, and 3000 m.

ARTICLE IN PRESS A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

2543

Plate 1. Locations of selected stations relative to sea-surface height on the DGOMB cruises in summers 2000–2002. CTD/bottle stations were taken at all locations except Sites 1 and 2. Moored 300-kHz ADCP stations were on cruises in June 2001 (MT3, S36, S42, and MT6) and June 2002 (Sites 1 and 2).

in the Gulf (e.g., Nowlin et al., 2001; Sturges, 1993). Wind stress produces currents in the upper waters of the ocean that transport nutrients, sediments, phytoplankton, and other biologically important matter from regions of generation or discharge, as in the case of river-borne material, to other regions in the GOM. Wind forcing occurs both as low-frequency regional circulation patterns forced by long-term low-frequency regional wind patterns and as episodic currents forced by high-frequency atmospheric events including tropical cyclones, extratropical cyclones, cold air outbreaks, and other frontal passages. Long-term patterns generated by wind stress include an anticyclonic circulation in the western GOM with a westward intensified boundary current (e.g., Sturges, 1993; DiMarco et al., 2005) and a persistent cyclonic circulation in the Bay of Campeche (Va´zquez de la Cerda et al., 2005). Nowlin et al. (2001) found the effects of tropical cyclones on oceanic currents were intense but localized and short-lived with diminishing effects observable about 7–10 days after passage (see also Nowlin et al., 2000 for discussion of literature). They also compared the deepwater current meter data with the occurrence of extratropical cyclones and frontal passages and found that most of these cyclones had no major effect on the currents in the deepwater GOM. Thus, frontal passages and extratropical cyclones had little or no effect on transport of materials, and when there was any effect, it was localized in time and space.

The world’s third largest river, the Mississippi River, and dozens of lesser rivers on average discharge approximately 20,000 m3 s 1 of fresh water into the northern GOM (Nowlin et al., 2001; Jochens et al., 2002). The Mississippi–Atchafalaya Rivers system discharges on the order of 1000 million m3 d 1 over the shelf. This affects stratification over large areas and enhances wind-driven coastal jets along frontal boundaries. Thus thermohaline forcing is important over the Gulf shelves. However, buoyancy forcing of the deepwater circulation from the Mississippi River is relatively minimal because such discharge is concentrated largely over the shelf. No water mass formation with its resulting thermohaline forcing of consequence is known to occur in the deepwater Gulf (Nowlin and Parker, 1974). However, off-shelf eddy processes can move riverine water off-shelf into deep water. There may be areas, such as the slope east of the Mississippi River Delta and associated with the western DeSoto Canyon, where such movement is frequent enough to have potential influence on the near-surface biology (Belabbassi et al., 2005; Walker et al., 2005; Morey et al., 2003) with possible consequent impacts to the benthic biology. It is in the context of the physical system that the transport and flux of organic materials, nutrients and dissolved oxygen throughout the water column in the GOM must be considered. The conditions of this system during the DGOMB cruises are examined in detail to provide insights into the biological productivity of the water column overlying the deepwater benthos.

ARTICLE IN PRESS 2544

A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

2. Data and methods Standard physical oceanographic instrumentation and sampling procedures were used to collect measurements on the DGOMB cruises. All such sampling was done to the same standards and using similar types of equipment as used on MMS-sponsored programs over the Texas–Louisiana shelf (Nowlin et al., 1998) and the northeast Gulf shelves (Jochens et al., 2002). Sixty total CTD stations were taken on the May/June 2000, June 2001, and June 2002 cruises. None were taken on the August 2002 cruise. Measurements of nitrate, phosphate, silicate, and dissolved oxygen concentrations were made from water samples on the cruises in 2000 and 2001. Shipboard ADCP measurements were made along track on DGOMB cruises in May/June 2000, June 2001, and June 2002. ADCP data also were collected on the August 2002 cruise as a complementary data set. Either a 150-kHz broadband ADCP or a narrow-band ADCP was operated simultaneously with an Ocean Surveyor 38-kHz (OS-38) ADCP. No data from the 38-kHz ADCP were collected on the June 2001 cruise. All ADCP instruments were manufactured by RD Instruments (RDI) of San Diego, CA. The OS-38 was on loan to Texas A&M University from the US Navy. A 300-kHz Workhorse ADCP (Model 300S Sentinal rated to 6000 m) was deployed during the June 2001 and June 2002 cruises at selected experimental stations. The instrument was lowered and left at the sea floor at the beginning of each experimental station. Locations, dates and times of deployment are given in Table 1 (see also Fig. 1). During the 2001 summer cruise, the ADCP was deployed at each of four process stations (MT3, S42, S36, and MT6). The instrument was placed 35 m above bottom in a downward-looking configuration. The deployments were short, ranging from 20 to 60 h. In summer 2002, the ADCP was deployed twice: at process stations in the Sigsbee Basin (site 1) and beneath the Loop Current (site 4). Both of these stations were in water depths greater than 3000 m. Deployment lengths were 36 and 48 h. Data quality was a concern for these deployments as quality generally decreased with distance from the instrument, such that data more than 20 m from the instrument, i.e., 40 m above bottom, were considered useless. The data loss likely is due to a lack of acoustic scatterers in the water column.

highest maximum and mean current speeds were near the sea surface with maxima reaching up to 200 cm s 1 in the eastern Gulf and 100 cm s 1 in the western Gulf. Current speeds decreased with depth tending toward a minimum at 800–1000 m. Current speeds increased somewhat with depth below that level, likely due to bottom intensification of currents. They further found flows tended to be anticyclonic in the upper waters and, consistent with findings of others (e.g., Hamilton and Lugo-Fernandez, 2001; DeHaan and Sturges, 2005), cyclonic in the lower waters. To gain an understanding of the vertical current structure, record-length mean and variance ellipses were determined for a mooring located in 2001 m of water in the north-central GOM at 89.7841W 27.2931N (Fig. 2; Nowlin et al., 2001). This mooring, called I1, was deployed from August 1999 through August 2000 (Hamilton and Lugo-Fernandez, 2001; Hamilton et al., 2003). It used 5–6 ADCPs and 7–8 in situ current meters to measure currents throughout the full water column. The ADCP records

3. Results 3.1. Current observations 3.1.1. Current observations throughout the water column The DGOMB program did not measure currents throughout the entire water column. So, other data sets were examined to provide background information on the current structure. DiMarco et al. (2001) examined the available measurements of currents in the deepwater GOM and determined the background currents by extracting the maximum, mean, and standard deviation for each instrument (see also Nowlin et al., 2001). They found that the

Fig. 2. Record-length current vectors and variance ellipses for records of duration greater than 100 d from MMS mooring I1 at 89.7841W 27.2931N.

Table 1 Locations and dates of deployment of the moored ADCP Name

Start date

End date

Latitude (1N)

Longitude (1W)

Depth (m)

Station name

2001-1 2001-2 2001-3 2001-4 2002-1 2002-2

03 June 2001 06 June 2001 09 June 2001 13 June 2001 03 June 2002 10 June 2002

04 June 2001 07 June 2001 12 June 2001 14 June 2001 05 June 2002 11 June 2002

28.2210 28.2501 28.9200 26.9999 25.0000 24.2500

89.4991 88.4168 87.6690 88.0000 90.0000 85.5000

935 755 1823 1740 3525 3410

MT3 S42 S36 MT6 Site 2 Site 4

ARTICLE IN PRESS A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

from the upper part of mooring I1 were decimated to approximate 20-m intervals to facilitate viewing in Fig. 2. The record shows strong surface-intensified currents and variability associated with the extended presence near the mooring of LCE Juggernaut, which was newly separated from the Loop Current during this period. The minimum currents and variability occur in the depth interval of 800–1000 m. There is intensification of currents and variance with depth below that level. Near-bottom currents and their variability were strongly oriented along the isobaths, due to the proximity of the mooring near the Sigsbee Escarpment. A mooring located closer to the escarpment showed stronger topographic steering while one located farther away showed weaker steering (Nowlin et al., 2001). Mean speeds at I1 for the deep currents below 1500 m were 10–14 cm s 1 with standard deviations of 8–9 cm s 1. Time series from I1 and two adjacent moorings (not shown) showed several

2545

large current events below 1500 m at the beginning of the record. Maximum speeds in the lower 500 m of I1 reached 55 cm s 1; an adjacent mooring during the same time period exceeded 90 cm s 1. These large-amplitude currents persisted for periods of 5–10 days and recurred after a 4–5 days period of weak currents in essentially the opposite direction (Nowlin et al., 2001). These vigorous currents are important factors in the transport of organic material in the lower water column. These large current events occurred when the extension of the Loop Current (soon to separate as Eddy Juggernaut) was adjacent to and approximately south of the mooring. The upper currents during this time at I1 show the strong northward flows associated with the northwest side of the anticyclonic eddy and the rotation of the flows to the east as the north side of the eddy passes the mooring and to the southeast as the northeast side of the eddy passes. During the last half of the records, when the eddy had

Plate 2. Near-surface velocity vectors from shipboard ADCP and sea surface height fields from two DGOMB cruises. Shown are ADCP currents at 14-m depth during (A) DGOMB Cruise 1 in May 2000 in the northwestern Gulf and (B) DGOMG Cruise 4 in August 2002 in the central and eastern Gulf. The date associated with the SSH field is given on the figure. B denotes the beginning location of the vertical section in Plate 4, and E denotes the end location.

ARTICLE IN PRESS 2546

A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

moved past the moorings, both the deep and upper currents were weaker. The currents at 1000 m or deeper are coherent. Hamilton and Lugo-Fernandez (2001) analyzed the data from I1 and the two adjacent moorings. Consistent with the results of Hamilton (1990), they found events such as energetic deep barotropic motions (e.g., topographic Rossby waves or deep eddies), sometimes with bottom intensification, in these deep current records. They concluded the topographic Rossby waves controlled the current dynamics of the lower water column. In their numerical modeling work, Oey and Lee (2002) provided more evidence of topographic Rossby waves along the northern deep basin and escarpment. 3.1.2. Shipboard current observations Shipboard ADCP observations of near-surface currents exhibit the influence of the eddy fields through which each cruise passed.

Two examples are discussed here. The physical oceanography of the GOM during the spring and summer of 2000 was dominated by the Loop Current in the eastern GOM (east of 901W) and by the presence and interaction of Loop Current Eddies Indigo and Juggernaut in the western GOM. LCE Indigo was centered near 26.251N 94.751W as evidenced by satellite altimeter observations (see Plate 1A and B). Plate 2A shows the SSH field of the northcentral GOM on 10 June 2000, which was during the second DGOMB cruise in 2000. Superimposed on the SSH field are nearsurface (14-m) current velocity vectors showing the magnitude and direction of currents along the cruise track. Both the May and June 2000 cruise tracks passed through several mesoscale circulation features that persisted over the spring and summer. The June 2000 cruise track passed through the center of LCE Indigo and north of a cyclonic eddy to the east of LCE Indigo. The strongest currents during the June 2000 cruise

Plate 3. Vertical section of current speed based on 150 kHz shipboard ADCP data collected along a section in (A) the western GOM during DGOMB Cruise 1 in June 2000 and (B) the southeastern GOM during DGOMB Cruise 4 in August 2002. Positions of circulation features are indicated above the contours.

ARTICLE IN PRESS A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

were associated with LCE Indigo. Comparison with May 2000 currents (not shown) suggests the eddy may have pushed closer to the continental shelf resulting in intensified currents along its northwestern limb (compare LCE Indigo in Plate 1A and B). Currents exceeded 120 cm s 1 near surface and 60 cm s 1 above 200 m. Plate 3A shows the vertical section of current velocity for the segment transiting LCE Indigo and the cyclone to its east (see Plate 2A). The cruise spent considerable time south and southeast of the Mississippi River Delta in the north-central GOM. Current speeds there were variable and weak as a result of the lack of strong circulation features. The mean currents for the length of the records for the May and June legs show surface intensification and exponential decay with depth, in agreement with previous studies (DiMarco et al., 2001) of moored current meter records of the deep GOM. For the May 2000 cruise, the record-length mean surface currents exceeded 30 cm s 1 in the upper 100 m and decayed to less than 15 cm s 1 at 250 m depth. Below 250 m mean currents smoothly decayed from 15 to 9 cm s 1 at 750 m. The standard deviation of currents shows a similar decay pattern with depth with values of 25 cm s 1 in the upper 100 m and about 7 cm s 1 below 600 m. This pattern has been identified as the first baroclinic dynamic mode (Charney and Flierl, 1981) based on hydrographic data of the deep GOM (Nowlin et al., 2001). Because of the longer time spent in the quiescent central GOM, mean speeds for the June 2000 cruise decayed from only 23 cm s 1 near surface to roughly 10 cm s 1 from 250 to 750 m depth. The vertical structure of currents during this cruise, however, was similar to the structure of the first dynamic mode for the GOM as seen on the May 2000 cruise. During the summer of 2002, there again were two cruises: in June and August. The tracks of both legs extended from the northern GOM to the Loop Current. During the June 2002 cruise (not shown), the track passed through or near several energetic circulation features, including the Loop Current, a detached LCE (241N 941W), a strong cyclone northwest of the Loop Current, and other weaker features. In general, currents were aligned with altimeter contours of SSH (see Plate 1D). The strongest observed currents during this cruise were in the Loop Current and the currents associated with anticyclonic features. In the Loop Current, currents exceeded 135 cm s 1 in the upper 100 m and 100 cm s 1 in the upper 150 m. Currents in the weak anticyclone, near 271N 901W in Plate 1D, peaked above 60 cm s 1 and averaged greater than 30 cm s 1 in the upper 200 m. Similar values were found on the eastern limb of the LCE. Mean currents in the upper 200 m of the water column ranged from 38 cm s 1 at 10 m depth to 13 cm s 1 at 200 m. The standard deviation similarly ranged from 25 cm s 1 at 10 m to 10 cm s 1 at 200 m. During the August 2002 cruise, the central GOM was nearly devoid of energetic circulation features, as evidenced by a nearly flat SSH field (Plate 2B). The dominant feature again was the Loop Current, which extended slightly further northward than in June 2002 (compare the Loop Current positions in Plates 1D and 2B). Currents in the Loop Current exceeded 180 cm s 1 to 50 m depth and 120 cm s 1 to 150 m (Plate 3B). The standard deviation of currents while in the Loop Current was nearly constant with depth at about 13 cm s 1. Record-length mean currents by depth level for this cruise decay from about 30 cm s 1 at the surface to about 10 cm s 1 at 200 m.

3.1.3. Moored current observations The deployments of the 300-kHz ADCP at the six process stations in summers of 2001 and 2002 were of short duration. Because of the short deployment period, these observations should be considered as a representative snapshot of the physical

2547

conditions that existed while each process station was occupied. They should not be considered as the climatology of currents possible at each location. The station locations relative to the upper eddy field are shown in Plate 1C for 2001 and Plate 1D for 2002. The near-bottom currents measured by the moored 300-kHz ADCP at the four process stations during summer of 2001 were generally weak with record-length mean speeds of about 2.5, 7.5, 2.5, and 5 cm s 1 at stations MT3, S42, S36, and MT6, respectively (Fig. 3). Notably there was no strong surface eddy feature associated with any of these stations (see Plate 1C). Mean directions were generally aligned along local bathymetry. Vertically, current speeds decreased slightly with closeness to the bottom. Standard deviations of speed were generally consistent for all four deployments at 1–2 cm s 1. The sea temperature at the ADCP transducer head was nearly constant throughout each deployment at 5.0, 6.0, 4.1, and 4.2 1C. Vertical and error velocities (an indicator of the amount of horizontal inhomogeneity of currents in the water column) were small (1–2 cm s 1) for all deployments. For three of four deployments, the bottom boundary layer appeared to be just less than 30 m thick. At S42, there was more vertical structure as the record-length mean cross-slope velocity changed sign at about 20 m above bottom; down-slope below 20 m and upslope above 20 m. This structure may be related to the close proximity of the station to the steep bathymetry of the Sigsbee escarpment. The record-length mean along-slope component was nearly zero from 30 m above bottom to the bottom. Examination of the time series of both velocity components showed a clear near-diurnal oscillation with amplitude of about 7 cm s 1 and about 6 h out of phase. Because of the short deployment it is not possible to determine whether this oscillation is due to tidal forcing, which is believed to be small in the deep Gulf, or is of inertial origin. Diurnal oscillations also were present at stations MT3, S36, and MT6. With amplitudes of 2–3 cm s 1, the oscillations at these locations were significantly weaker than those at S42. In summer 2002, record-length mean currents were much stronger than the previous year’s observations increasing from 20 to 40 cm s 1 between 55 and 40 m above bottom (Fig. 3). The standard deviation of current also increased considerably in this range from 10 to 30 cm s 1 as the number of usable data decreased. The stations were not located in regions with strong eddies or near topography. The increase in speed with depth, then, likely is an artifact of the decrease in data quality with depth. The record-length mean of individual velocity components show a mean southwest direction at both stations. Neither deployment shows a detectable tidal signal. Mean temperatures are about 4.3 1C. Vertical and error velocity components are disturbingly large (at times greater than 10 cm s 1).

3.2. Water properties 3.2.1. Water properties during DGOMB cruises The Loop Current is the source water for the deep basin of the Gulf. Water properties in both the eastern and western deep Gulf have characteristics similar to those within the Loop Current (Morrison and Nowlin, 1977; Morrison et al., 1983). Only one DGOMB station, S4 in June 2002, sampled in the Loop Current. The vertical structure of water properties in this station can be compared with those in other features in the interior of the Gulf. We show salinity as an example. The general salinity pattern for all stations consists of four major characteristics (Plate 4). Most notable is lack of horizontal variability in salinity below depths of approximately 800–1000 m.

ARTICLE IN PRESS 2548

A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

Fig. 3. Record-length mean speed (solid) and 71 standard deviation (dashed) of currents measured using a bottom-moored, downward-looking 300-kHz ADCP during four deployments on Cruise 3 in 2001 and two deployments during Cruise 4 in 2002. The instrument was deployed at 35 and 60 m above bottom in 2001 and 2002, respectively. Note the scale differences between the 2001 and 2002 deployments.

This is present in all CTD data and is a characteristic feature of the GOM deep waters (Nowlin, 1972; Nowlin and McLellan, 1967). Second is a salinity maximum, which is a characteristic of the Subtropical Underwater that enters the Gulf with the Loop Current (Morrison and Nowlin, 1977). This maximum is reduced with increasing distance from the source waters. However, this decrease is not only a function of distance from the Loop Current, but also of mixing processes. This can be seen by comparing the maxima of stations S36 and RW6. Station S36 is located geographically closer to the Loop Current than station RW6, yet has a very reduced salinity maximum. Station RW6 has a predominant salinity maximum pattern comparable to that of the Loop Current station S4, yet it is in the northwest corner of the Gulf. Station S36 sampled in a small slope anticyclone, where waters were subject to mixing due to interaction of the eddy with the bathymetry and to entraining of shelf waters. Both the mixing and entrainment processes contributed to the consequence erosion of the salinity maximum. In contrast, station RW6 sampled near the middle of LCE Indigo, which had carried Loop Current water into the western GOM. This shows the importance of the eddy fields on the deep water property patterns. The third characteristic of the salinity profiles is the salinity minimum at about 500–1000 m or sy of 27.3–27.5 kg m 3. This is associated with the Antarctic Intermediate Water that is brought into the Gulf by the Loop Current (Morrison and Nowlin, 1977). The depth at which this minimum occurs is variable. It is related

to the type of eddy feature the station sampled or the proximity to the bathymetric effects on the slope. The fourth characteristic is the variability in the salinity values of the upper 50–100 m. These upper waters have lower salinity concentrations than at depth. Part of the variability is due to the mixing of low-salinity water discharged from the Mississippi River with the near-surface oceanic waters. Of the selected stations shown in Plate 4, the lowest near-surface salinity values occur at S36, which is located southeast of the Mississippi River Delta (see Fig. 1). Other stations, west as well as east of the Mississippi River Delta, exhibited low salinity in these upper waters. In these cases, there were circulation features adjacent to the Mississippi River that transported shelf water, which had been diluted by mixing with river water, to the slope regions adjacent to the river. This situation has been observed by others (e.g., Belabbassi et al., 2005; Mu¨ller-Karger et al., 1991; Walker et al., 2005). Vertical sections of properties were examined where the station locations allowed. Spacing of stations was large, being generally 35 km or greater. This spacing allows consideration of large-scale features only. In general, the few transects showed similar results. Transects taken across strong anticyclonic eddies showed the characteristic pattern of depressed isolines near the eddy center. Examples are discussed below. There were no transects with reasonably spaced stations taken across strong cyclonic eddies. However, the transects that crossed weak parts of

ARTICLE IN PRESS A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

2549

Plate 4. Selected station data showing salinity versus density (right) and depth (left). Shown are stations RW6 (2000) in a Loop Current Eddy in the western Gulf, S36 (2000) in a cyclonic eddy in the eastern Gulf, S39 (2000) in a slope anticyclonic eddy in the northeastern Gulf, GK-Furrows (GKF; 2001) at the edge of a Loop Current Eddy in the central Gulf, S4 in the Loop Current (2002), S2 at the edge of a Loop Current Eddy in the south central Gulf (2002), and S5 at the edge of a cyclone in the eastern Gulf (2002).

cyclones showed a tendency for the characteristic doming of isolines. No examples are presented here. The transect selected for discussion here was the western-most line on the first DGOMB cruise in May–June 2000. It extended from the shelf edge at about 27.51N 961W to approximately the middle of the anticyclonic LCE Indigo at about 261N 94.51W (see Fig. 1 for locations and Plate 1A for locations relative to LCE Indigo). Fig. 4A shows the contours of density anomaly (sy) along this transect. The station taken at the shelf edge was RW1 at the left side of the figure, while the station taken near the middle of the LCE was RW6 at the right side. Isopycnals show the stability of the water column by their monotonic increase with depth. They exhibit a pattern typical of anticyclonic eddies with the downward sloping or deepening of the isopycnals from the eddy edge to the eddy center The vertical pattern of potential temperature (not shown) is similar, although the temperature monotonically decreases with depth. Salinity contours, shown in Fig. 4B, also exhibit the downward slope from the eddy edge to the eddy center that was seen in density. There is a salinity maximum between 100 and 200 m depth, which is near the 25.4 kg m 3 sy line (shown as a dashed line), that corresponds to the signature of the Subtropical Underwater mass that the LCE has transported into the western GOM from the Loop Current. The Antarctic Intermediate Water mass, with its salinity minimum, is evident in water depths of about 900 m and sy of 27.5 kg m 3 (also shown as a dashed line). This water mass also was transported from the source waters of the Loop Current to the western GOM by the LCE. Vertical contours of nitrate and dissolved oxygen along the transect are shown in Figs 4C and D, respectively. Both exhibit the downward trend in isolines indicative of an anticyclonic circulation feature.

Historical data show that nutrients are low in the photic zone (approximately the upper 60–100 m) of the Loop Current and LCEs (Morrison and Nowlin, 1977). The Loop Current and LCEs, however, also carry the high nutrient concentrations of the Antarctic Intermediate Water and modified North Atlantic Deep Water into the Gulf at depth. The nitrate concentrations are below detection limits in the upper 100–150 m (Fig. 4C). The phosphate and silicate (not shown) are also very low, being o0.5 and o1 mM, respectively, in the upper water column. This reflects the consumption of the nutrients by phytoplankton in the presence of sunlight. The complete utilization of nitrate, with some phosphate and silicate still being detectable, suggests the biological system is nitrate limited. All the three nutrients increase with depth, reflecting both transport of nutrient-rich deep waters into the GOM and nutrient regeneration as detritus falls and decays through the water column. The Antarctic Intermediate Water mass is evident in the nitrate maximum at about 900 m and sy of 27.5 kg m 3 (shown as a dashed line). The oxygen section (Fig. 4D) shows the presence of the Tropical Atlantic Central Water, which is characterized by the oxygen minimum located at about 600 m depth and sy of 27.15 kg m 3 (shown as a dashed line). The effects of atmospheric exchanges with the surface are seen in the high values of oxygen (44.5 mL L 1) in the upper 100 m. The deep waters below about 1000 m also are relatively high in oxygen. The sources of this deep oxygen are the highly oxygenated deep water masses entering into the GOM through the Yucatan Channel (Morrison and Nowlin, 1977; Jochens et al., 2005; see also Rivas et al., 2005). Fig. 5 shows profiles versus density of dissolved oxygen, nitrate, phosphate, and silicate from all DGOMB 2000 and 2001 data. The profiles distinguish between data taken in the eastern GOM (triangles) and those taken in the western GOM (pluses). Here west is taken to be west of 901W. Overall patterns are the

ARTICLE IN PRESS 2550

A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

Fig. 4. Water properties along the western-most transect of DGOMB cruise 1 in May/June 2000. Shown are the density anomaly (as sq), salinity, nitrate, and dissolved oxygen. Dashed lines show the core sq. of the Gulf’s major water masses: 24.5 for Subtropical Underwater, 26.5 for Sargasso Sea Water, 27.15 for Tropical Atlantic Central Water, 27.3–27.5 for Antarctic Intermediate Water, and 27.7 for North Atlantic Deep Water.

same in the east and west. The differences in distributions in the upper 100–200 m of the water column reflect the different types of features (cyclones and anticyclones) and the bathymetric regions (slope and abyssal plain) sampled, as well as differences caused by surface processes, such as localized photosynthesis and respiration, winds and associated mixing, heating and cooling, and precipitation and evaporation, which impact rates of oxygen input and consumption. The oxygen profile (Fig. 5A) shows high concentrations in the upper water column, reflecting photosynthesis and exchanges with the atmosphere. Values decrease with depth in the upper 500 m, as expected from the vertical oxygen distribution of the source waters, particularly the dissolved oxygen minimum that is characteristic of the Tropical Atlantic Central Water. Local consumption of oxygen in the near-surface waters during the decay of detritus falling through the water column also contributes to this dissolved oxygen structure. Below about 500 m, the oxygen concentration then increases with depth. Compare the high variability of oxygen concentrations in the upper water column with the small variability in the deep water column. This

difference reflects the variability in processes that input or consume oxygen. The nutrient profiles (Figs. 5B–D) also reflect the processes that increase or decrease the concentrations. The upper water column, where the nutrients have largely been consumed by phytoplankton, has low concentrations. Concentrations then increase with depth to a maximum near the depth or density surface of the Antarctic Intermediate Water mass on the sy density surfaces of 27.3–27.5 kg m3. This reflects the distribution of the nutrients in the source water masses, although the local decay of detritus regenerates the nutrients and contributes to this distribution. Deeper in the water column the concentrations decrease or remain fairly constant, again due mainly to the distributions found in the source waters.

3.2.2. Comparison of DGOMB with historical water properties The temperature–salinity diagrams for the DGOMB cruises are shown in Fig. 6. The stations with the low salinity in the upper waters were located adjacent to the Mississippi River, which is the

ARTICLE IN PRESS A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

2551

Fig. 5. Water properties versus density for DGOMB cruises in 2000 and 2001. Pluses denote data taken west of 901W, triangles at and east of 901W. Horizontal lines show the density surface of the core of the water masses for, top to bottom, Subtropical Underwater (24.5), 18 1C Sargasso Sea Water (26.5), Tropical Atlantic Central Water (27.15), Antarctic Intermediate Water (27.3–27.5), and North Atlantic Deep Water (27.7).

source of the freshening of oceanic waters. The salinity maximum associated with the Subtropical Underwater and its various states of erosion can be seen at temperatures of 21–23 1C. The most prominent profile showing this feature was taken in the Loop Current at station S4. The salinity minimum of the Antarctic Intermediate Water is present at temperatures of 6 1C. Most notable is the horizontal uniformity of the y–S diagram in waters with potential temperatures of about 17 1C or less. This basin-wide uniformity was noted by Nowlin and McLellan (1967) as indicating the Gulf waters ‘‘constitute essentially a single system.’’ The profiles shown in Fig. 6 are in good agreement with the profiles in Nowlin and McLellan using data collected in 1962, indicating the system has not changed significantly in the 40 years between the sets of cruises. The dissolved oxygen, nitrate, phosphate, and silicate data from the DGOMB cruises were compared with published historical data sets from the eastern and western GOM basins. Morrison and Nowlin (1977) published results from a May 1972

cruise of the R/V Alaminos conducted in the eastern GOM with emphasis on the region occupied by the Loop Current. Morrison et al. (1983) published results from an April 1978 cruise of the R/V Gyre in the western GOM. The data from the 1972 cruise were available digitally, while the data from the 1978 cruise were not. So comparisons with the 1978 data are estimates based on scanned figures from the Morrison et al. (1983) paper. The comparisons indicate the deep waters have not undergone major changes over the 30 years between the cruises. Fig. 7 shows the dissolved oxygen comparison. The comparisons indicate that the basic patterns seen in the DGOMB data described above were present in the historical data. There is less variability (both deep and shallow) in the DGOBM data than in the historical data. In general, this likely is due to improvements in sampling and analysis techniques and skills, rather than a climatic change in the patterns. The oxygen pattern in the east (Fig. 7A) suggests that the oxygen concentrations for syp26.5 kg m 3 might be greater in the

ARTICLE IN PRESS 2552

A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

Fig. 6. Composite potential temperature salinity diagram for CTD stations from DGOMB cruises in May–June 2000, June 2001, and June 2002. Lines of density in sigma-theta (kg m 3) also are shown (dashed).

1972 data than in the DGOMB data. However, this actually reflects the fact that the sampling on the 1972 cruise was mainly in the region of the Loop Current, while the DGOMB sampling was throughout the eastern GOM. The older data also sampled in the area where the oxygen maximum associated with the 18 1C Sargasso Sea Water is present, whereas most DGOMB stations were located in regions where this water mass had eroded away. Thus, this difference in DGOMB versus 1972 oxygen concentrations does not reflect a climatic change, but rather the fact that the 1972 sampling occurred directly in the source waters for oxygen in the eastern basin and so had not been diluted by mixing from other processes that modify oxygen content. Nowlin et al. (1969) undertook a careful and detailed examination of dissolved oxygen data sets from the deep waters of the GOM. They were motivated by findings from three cruises in 1966 and 1967 that seemingly contradicted the study of McLellan and Nowlin (1963). That earlier study, based on data from the 1962 cruise 62-H-3 of the Hidalgo, had suggested there were large gradients in dissolved oxygen below approximately 1500-m depth. Nowlin et al. (1969) compared the dissolved oxygen data from 62-H-3 and the 1964 cruises, 64-A-2 and 64-A-3, of the Alaminos with those from the 1935 cruise of the Atlantis, the 1958 and 1959 cruises of Hidalgo (58-H-4, 58-H-1, 59-H-2), and the 1966 and 1967 cruises of the Alaminos (66-A-8, 67-A-4, 67-A-8). They concluded that the data from the 1962 and 1964 cruises were based on faulty sampling or poor analytical techniques. They determined the data sets from the 1935, 1958, 1959, 1966, and 1967 cruises were good. Results of their analysis of these good data showed there was no clearly discernable horizontal variation in dissolved oxygen in the deep waters of the Gulf. Nowlin et al. (1969) argued that this was consistent with the horizontal uniformity of salinity and potential temperature in the data from the 1966 and 1967 cruises and that McLellan and

Fig. 7. Dissolved oxygen versus density for the eastern (A) and western (B) Gulf of Mexico. Pluses are from DGOMB cruises in May/June 2000 and June 2001. In the east, dots show data from the R/V Alaminos 72A09 cruise in May 1972. In the west, dots give estimated data from Fig. 3 in Morrison et al. (1983), showing data from 78G03 cruise in April 1978. Horizontal lines show the density surface of the core of the water masses for, top to bottom, Subtropical Underwater (24.5), 18 1C Sargasso Sea Water (26.5), Tropical Atlantic Central Water (27.15), Antarctic Intermediate Water (27.3–27.5), and North Atlantic Deep Water (27.7).

ARTICLE IN PRESS A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

2553

Table 2 DGOMB cruises and historical statistics for dissolved oxygen at or below 1500 m depth in the Gulf of Mexico (historical statistics are after Nowlin et al., 1969; Wu¨st data are in the core of the NADW, based on data from 1932 to 1937; DGOMB data are from this study) 1

1

1

Data set identifier

Mean (mL L

Atlantis 1935 58-H-1 58-H-4 59-H-2 66-A-8 67-A-4 Gulf Basin 67-A-8

4.96 5.08 4.87 5.01 4.99 5.03 5.01

0.134 0.124 0.054 0.055 0.048 0.077 0.111

0.028 0.028 0.020 0.013 0.011 0.010 0.024

23 19 7 17 19 55 21

67-A-4 Yucatan Basin Wu¨st (1964) Yucatan Basin

5.61 5.64

0.099 0.33

0.030 0.076

11 19

62-H-3a 64-A-2 and 3a

4.62 4.58

0.280 0.482

0.026 0.059

114 67

DGOMB 2000/2001 DGOMB 2000 DGOMB 2001

4.998 4.976 5.051

0.127 0.134 0.085

0.013 0.016 0.016

102 73 29

a

)

Standard deviation (mL L

)

Standard error (mL L

)

Number of data points

Determined to be bad data by Nowlin et al. (1969).

Nowlin (1963) had found in the 62-H-3 data. They also found that there were only slight vertical gradients below the depth of the Yucatan sill. Table 2 lists the mean, standard deviation, and standard error of the mean for each of the cruises examined by Nowlin et al. (1969). They found no evidence that oxygen concentrations had changed during the 30-year time period covered. Table 2 also includes the statistics for the DGOMB cruises. The mean dissolved oxygen at or below 1500 m for the DGOMB cruises is 4.998 mL L 1, which is comparable to the mean dissolved oxygen concentration of 4.99 mL L 1 found for all the ‘‘good’’ data by Nowlin et al. (1969). Moreover, the statistics shown in Table 2 for the Gulf basin indicate the dissolved oxygen concentrations from 2000 and 2001 are comparable to the concentrations from the historical cruises spanning the period 1935–1967. These comparisons indicate that dissolved oxygen concentrations in waters at or below 1500 m have not changed during the 70-year time period covered.

4. Conclusions The results of the DGOMB field study are consistent with previous studies of the circulation and distribution of water properties in the deep water Gulf of Mexico (GOM). The eddy field of the upper 1000 m creates environmental conditions that are favorable for biological productivity in an otherwise oligotrophic subtropical ocean. The eddy field during DGOMB provided three mechanisms for creating biologically favorable conditions over the slope: cross-margin flow, upwelling in cyclones, and uplift from the interaction of anticyclones with bathymetry. The Loop Current, Loop Current Eddies, and the smaller anticyclonic and cyclonic eddies can generate cross-margin flow that pulls shelf waters onto the slope of the Gulf. This is evident in the lowsalinity, near-surface waters sampled at stations on the slope adjacent to the Mississippi River Delta. The cross-margin flow brings nutrients and/or biologically productive waters to the slope environment. In the vicinity of the Mississippi River Delta, this physical system results in the persistent input of nutrients and chlorophyll. The circulation of the cyclonic eddies itself results in the uplift of relatively nutrient-rich mid-depth waters into the photic zone of the water column. From the SSH fields, it is clear these mesoscale features also can persist for several months or

more, allowing the growth of phytoplankton over the slope regimes affected. The interaction of the anticyclones with the bathymetry or with cyclones also can result in the uplift of nutrients to the photic zone creating conditions favorable for phytoplankton growth. The near-surface currents during DGOMB included high-speed, energetic flows that sea surface height fields show lasted for time scales of months. When adjacent to the shelf, these flows can result in the persistent availability of relatively high nutrient concentrations to the photic zone of the affected areas. One such region is the slope to the east of the Mississippi River Delta, where slope eddies can transport low-salinity, relatively nutrient-rich waters associated with the Mississippi River disharge off-shelf into the slope and deep water regime of DeSoto Canyon. Historical data show the deep near-bottom currents also can be high-speed and energetic, although the few, short-term DGOMB samples did not have such currents. These high-speed currents provide a mechanism for the transport of organic material, in both larger and small particle sizes, from one benthic area to another. The poorer current data quality of the abyssal plain stations, S2 and S4, was likely due to lack of acoustic scatterers. The slope stations had sufficient scatterers to produce good-quality data. This suggests there was a difference in the particulate content of the water near-bottom between the abyssal plain and slope, and this might be indicative of weaker currents over the abyssal plain than over the slope. The distributions of water properties in the deep water GOM as determined by the DGOMB data sets appear to be unchanged from those of historical data. The waters of the upper 100 m, which are subject to the influence of the winds and atmospheric conditions, have variable concentrations of salinity, nutrients and dissolved oxygen that depend on the eddy field, the proximity to the Mississippi River Delta and the shelf, and presumably the season, although DGOMB sampled only in summer. Below this, the water properties quickly take on the characteristics of the source waters brought into the Gulf by the Loop Current. Below temperatures of approximately 17 1C or depths of 800 m, the variability of the properties becomes greatly reduced. Below about 1500 m depth, the horizontal distributions of temperature, salinity, nitrate, phosphate, silicate, and dissolved oxygen are approximately uniform. The water masses of the deep Gulf are those with sources in the Atlantic that are brought in by the Loop Current.

ARTICLE IN PRESS 2554

A.E. Jochens, S.F. DiMarco / Deep-Sea Research II 55 (2008) 2541–2554

Acknowledgements This work was sponsored by the US Minerals Management Service under Contract nos. 1435-01-99-CT-30991 (DGOMB Project) and 1435-01-02-CT-85080 (Oxygen Project) and 143501-98-CT-30910 (Deepwater Reanalysis Project). The I1 current data were provided by Science Applications International Corporation, which collected the data under MMS Contract no. 143501-96-CT-30825. The SSH fields were provided courtesy of Dr. Robert Leben, Colorado Center for Astrodynamics Research. References Belabbassi, L., Nowlin Jr., W.D., Jochens, A.E., Biggs, D.C., Chapman, P., 2005. Summertime nutrient supply to near-surface waters of the northeastern Gulf of Mexico: 1998, 1999, and 2000. Gulf of Mexico Science 23 (2), 137–160. Biggs, D.C., Mu¨ller-Karger, F.E., 1994. Ship and satellite observations of chlorophyll stocks in interacting cyclone-anticyclone eddy pairs in the western Gulf of Mexico. Journal of Geophysical Research 99 (C4), 7371–7384. Biggs, D.C., Fargion, G.S., Hamilton, P., Leben, R.R., 1996. Cleavage of a Gulf of Mexico Loop Current eddy by a deep water cyclone. Journal of Geophysical Research 101 (C9), 20,629–20,641. Biggs, D., Jochens, A., Howard, M., DiMarco, S., Mullin, K., Leben, R., Muller-Karger, F., Hu, C., 2005. Eddy forced variations in on-margin and off-margin summertime circulation along the 1000-m isobath of the northern Gulf of Mexico, 2000–2003. In: Sturges, W., Lugo-Fernandez, A. (Eds.), Circulation in the Gulf of Mexico: Observations and Models, Geophysical Monograph Series, vol. 161. American Geophysical Union, 360pp. Biggs, D.C., Hu, C., Mu¨ller-Karger, F.E., 2008. Remotely sensed sea surface chlorophyll and POC flux at deep Gulf of Mexico benthos sampling stations. Deep Sea Research II, this issue [doi:10.1016/j.dsr2.2008.07.013]. Capurro, L.R.A., Reid, J.L. (Eds.), 1970. Contributions on the Physical Oceanography of the Gulf of Mexico. Gulf Publishing Co., Houston, 288pp. Charney, J.G., Flierl, G.R., 1981. Oceanic analogues of atmospheric motions. In: Warren, B.A., Wunsch, C. (Eds.), Evolution of Physical Oceanography. MIT Press, Cambridge, MA, pp. 504–548. Cooper, C., Forristall, G.Z., Joyce, T.M., 1990. Velocity and hydrographic structure of two Gulf of Mexico warm-core rings. Journal of Geophysical Research 95 (C2), 1663–1679. DeHaan, C.J., Sturges, W., 2005. Deep cyclonic circulation in the Gulf of Mexico. Journal of Physical Oceanography 35 (10), 1801–1812. DiMarco, S.F., Reid, R.O., Jochens, A.E., Nowlin Jr., W.D., Howard, M.K., 2001. General characteristics of currents in the deepwater Gulf of Mexico. OTC 12993. In: Proceedings of Offshore Technology Conference 2001, Houston, TX. DiMarco, S.F., Nowlin Jr., W.D., Reid, R.O., 2005. A statistical description of the velocity fields from upper ocean drifters in the Gulf of Mexico. In: Sturges, W., Lugo-Fernandez, A. (Eds.), Circulation in the Gulf of Mexico: Observations and Models, Geophysical Monograph Series, vol. 161. American Geophysical Union, 360pp. Elliot, B.A., 1982. Anticyclonic rings in the Gulf of Mexico. Journal of Physical Oceanography 12, 1292–1309. Forristall, G.Z., Schaudt, K.J., Cooper, C.K., 1992. Evolution and kinematics of a Loop Current Eddy in the Gulf of Mexico during 1985. Journal of Geophysical Research 97 (C2), 2173–2184. Hamilton, P., 1990. Deep currents in the Gulf of Mexico. Journal of Physical Oceanography 20, 1087–1104. Hamilton, P., 1992. Lower continental slope cyclonic eddies in the central Gulf of Mexico. Journal of Geophysical Research 97 (C2), 2185–2200. Hamilton, P., Lugo-Fernandez, A., 2001. Observation of high speed deep currents in the northern Gulf of Mexico. Geophysical Research Letters 28 (14), 2867–2870. Hamilton, P., Fargion, G.S., Biggs, D.C., 1999. Loop Current eddy paths in the western Gulf of Mexico. Journal of Physical Oceanography 29, 1180–1207. Hamilton, P., Berger, T.J., Johnson, W., 2002. On the structure and motions of cyclones in the northern Gulf of Mexico. Journal of Geophysical Research 107 (C12), 3208, 18pp. Hamilton, P., Singer, J.J., Waddell, E., Donohue, K., 2003. Deepwater observations in the northern Gulf of Mexico from in-situ current meters and PIES, Final Report, Volume II: Technical Report. US Department of the Interior, Minerals Management Service, Gulf of Mexico OCS Region, New Orleans, LA. OCS Study MMS 2003-049, 95pp. Jochens, A.E., Biggs, D.C., 2006. Spatial variability in sperm whale encounters in summer 2005 in an upwelling regime in the northern Gulf of Mexico. In revision. Jochens, A.E., DiMarco, S.F., Nowlin Jr., W.D., Reid, R.O., Kennicutt II, M.C., 2002. Northeastern Gulf of Mexico Chemical Oceanography and Hydrography Study: Synthesis Report. OCS Study MMS 2002-055, US Department of the Interior, Minerals Management Service, Gulf of Mexico OCS Region, New Orleans, LA, 586pp. Jochens, A.E., Bender, L.C., DiMarco, S.F., Morse, J.W., Kennicutt II, M.C., Howard, M.K., Nowlin Jr., W.D., 2005. Understanding the Processes that Maintain the

Oxygen levels in the Deep Gulf of Mexico: Synthesis Report. US Dept. of the Interior, Minerals Management Service, Gulf of Mexico OCS Region, New Orleans, LA. OCS Study MMS 2005-032, 142pp. Leben, R.R., 2005. Altimeter-derived Loop Current metrics. In: Sturges, W., LugoFernandez, A. (Eds.), Circulation in the Gulf of Mexico: Observations and Models, Geophysical Monograph Series, vol. 161. American Geophysical Union, 360pp. Lewis, J.K., Kirwan Jr., A.D., 1985. Some observations of ring topography and ring–ring interactions in the Gulf of Mexico. Journal of Geophysical Research 90 (C5), 9017–9028. MacDonald, I.R. (Ed.), 2002. Stability and Change in Gulf of Mexico Chemosynthetic Communities. Volume II: Technical Report. OCS Study MMS 2002-036. Minerals Management Service, Gulf of Mexico OCS Region, US Department of the Interior, New Orleans, LA, 456pp. McLellan, H.J., Nowlin Jr., W.D., 1963. Some features of the deep water in the Gulf of Mexico. Journal of Marine Research 21 (3), 233–245. Morey, S.L., Schroeder, W.W., O’Brien, J.J., Zavala-Hidalgo, J., 2003. The annual cycle of riverine influence in the eastern Gulf of Mexico basin. Geophysical Research Letters 30 (16), 1867. Morrison, J.M., Nowlin Jr., W.D., 1977. Repeated nutrient, oxygen, and density sections through the Loop Current. Journal of Marine Research 35 (1), 105–128. Morrison, J.M., Merrell Jr., W.J., Key, R.M., Key, T.C., 1983. Property distributions and deep chemical measurements within the western Gulf of Mexico. Journal of Geophysical Research 88 (C4), 2601–2608. Mu¨ller-Karger, F.E., Walsh, J.J., Evans, R.H., Meyers, M.B., 1991. On the seasonal phytoplankton concentration and sea surface temperature cycles of the Gulf of Mexico as determined by satellites. Journal of Geophysical Research 96 (C7), 12,645–12,665. Muller-Karger, F.E., Varela, R., Thunell, R., Luerssen, R., Hu, C., Walsh, J.J., 2005. The importance of continental margins in the global carbon cycle. Geophysical Research Letters 32 (L01602), 4. Nowlin Jr., W.D., 1972. Winter Circulation Patterns and Property Distributions. In: Capurro, L.R.A., Reid, J.L. (Eds.), Contributions on the Physical Oceanography of the Gulf of Mexico. Texas A&M University Oceanographic Studies, vol. 2. Gulf Publishing Co., Houston, pp. 3–51, 288pp. Nowlin Jr., W.D., McLellan, H., 1967. A characterization of the Gulf of Mexico waters in winter. Journal of Marine Research 25 (1), 29–59. Nowlin Jr., W.D., Parker, C.A., 1974. Effects of a cold-air outbreak on shelf waters of the Gulf of Mexico. Journal of Physical Oceanography 4 (3), 467–486. Nowlin Jr., W.D., Paskausky, D.F., McLellan, H.J., 1969. Recent dissolved-oxygen measurements in the Gulf of Mexico deep waters. Journal of Marine Research 27 (1), 39–44. Nowlin Jr., W.D., Jochens, A.E., Reid, R.O., DiMarco, S.F., 1998. Texas–Louisiana Shelf Circulation and Transport Processes Study: Synthesis Report, Volume I Technical Report. OCS Study MMS 98-0035. US Department of the Interior, Minerals Management Service, Gulf of Mexico OCS Region, New Orleans, LA. 502pp. Reviewed by a Science Review Board. Nowlin Jr., W.D., Jochens, A.E., DiMarco, S.F., Reid, R.O., 2000. Chapter 4: Physical Oceanography, in Continental Shelf Associates Inc., Deepwater Gulf of Mexico Environmental and Socioeconomic Data Search and Literature Synthesis. Volume I: Narrative Report, pp. 61–121. US Department of the Interior, Minerals Management Service, Gulf of Mexico OCS Region, New Orleans, LA, OCS Study MMS 2000-049, 340pp. Nowlin Jr., W.D., Jochens, A.E., DiMarco, S.F., Reid, R.O., Howard, M.K., 2001. Deepwater Physical Oceanography Reanalysis and Synthesis of Historical Data: Synthesis Report. OCS Study MMS 2001-064, US Dept. of the Interior, Minerals Management Service, Gulf of Mexico OCS Region, New Orleans, LA, 528pp. Oey, L.-Y., Lee, H.-C., 2002. Deep eddy energy and topographic Rossby waves in the Gulf of Mexico. Journal of Physical Oceanography 32 (12), 3499–3527. Rivas, D., Badan, A., Ochoa, J., 2005. The ventilation of the deep Gulf of Mexico. Journal of Physical Oceanography 35 (10), 1763–1781. Sahl, L.E., Wiesenburg, D.A., Merrell, W.J., 1997. Interactions of mesoscale features with Texas shelf and slope waters. Continental Shelf Research 17 (2), 117–136. Sturges, W., 1993. The annual cycle of the western boundary current in the Gulf of Mexico. Journal of Geophysical Research 98 (C10), 18,053–18,068. Sturges, W., Leben, R., 2000. Frequency of ring separations from the Loop Current in the Gulf of Mexico: a revised estimate. Journal of Physical Oceanography 30, 1814–1819. Sturges, W., Lugo-Fernandez, A. (Eds.), 2005. Circulation in the Gulf of Mexico: Observations and Models, Geophysical Monograph Series, vol. 161. American Geophysical Union, 360pp. Va´zquez de la Cerda, A.M., Reid, R.O., DiMarco, S.F., Jochens, A.E., 2005. Bay of Campeche circulation: An update. In: Sturges, W., Lugo-Fernandez, A. (Eds.), Circulation in the Gulf of Mexico: Observations and Models, Geophysical Monograph Series, vol. 161. American Geophysical Union, 360pp. Vukovich, F.M., 1988. Loop Current boundary variations. Journal of Geophysical Research 93 (C12), 15,585–15,591. Walker, N.D., Wiseman, W.J., Rouse Jr., L.J., Babin, A., 2005. Effects of river discharge, wind stress, and slope eddies on circulation and the satelliteobserved structure of the Mississippi River Plume. Journal of Coastal Research 21 (6), 1228–1244. Welsh, S.E., Inoue, M., 2000. Loop Current rings and the deep circulation in the Gulf of Mexico. Journal of Geophysical Research 105, 16,951–16,959.