- Email: [email protected]
Texas and Louisiana coastal vulnerability and shelf connectivity
ARTICLE IN PRESS MPB-08293; No of Pages 8 Marine Pollution Bulletin xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul
Texas and Louisiana coastal vulnerability and shelf connectivity Kristen M. Thyng* , Robert D. Hetland Department of Oceanography, Texas A&M University, United States
A R T I C L E
I N F O
Article history: Received 19 November 2015 Received in revised form 5 October 2016 Accepted 28 December 2016 Available online xxxx Keywords: Modeling Coastal Connectivity Pollution Transport Drifter
A B S T R A C T A numerical study of connectivity between the continental shelf and coast in the northwestern Gulf of Mexico using a circulation model and surface-limited numerical drifters shows that despite seasonal changes in winds, the overall connectivity of the shelf with the coastline is similar in the winter and summer, though it extends more offshore in Texas in summer. However, there is a spatial pattern to the connectivity: more of the inner shelf is connected with the coast in Texas as compared with Louisiana. Subsets of the coast do have seasonal variability: the coast near both Galveston and Port Aransas has more connectivity from upcoast in the winter and from offshore and downcoast in the summer. In both seasons, we find drifters reach the Port Aransas coast most frequently, with a stronger trend in the summer. These results are important for assessing likely pathways for spilled oil and other potentially hazardous material. © 2016 Elsevier Ltd. All rights reserved.
1. Introduction The Texas coastline is at risk of impact by material from many sources including those related to the oil and gas industry, which plays a significant role in the state economy. However, not all areas along the coast are equally at risk for oil spill impacts. We seek to understand which areas along the Texas and Louisiana coastlines are more likely to be hit by oil spills, under what conditions, and from where the material can originate. Analogously, we want to know what areas are not likely to be impacted. Numerical drifter tracking has often been used for understanding connectivity between regions of interest, and where released material will travel. Biological applications have included studies on larvae movement and regional connectivity for lobsters in the Gulf of Maine (Xue et al., 2008), coral in the Gulf of Mexico (Lugo-Fernández et al., 2001), larvae into Texas bays (Brown et al., 2000), and several species along the coast of Australia (Roughan et al., 2011); connectivity for several applications in Southern California (Mitarai et al., 2009) and especially harmful algal blooms in the Gulf of Maine (Li et al., 2014); and transport pathways for phytoplankton leading to harmful algal blooms in the Pacific Northwest (Giddings et al., 2014) and the Gulf of Mexico (Henrichs et al., 2015; Olascoaga et al., 2008; Thyng et al., 2013; Walsh et al., 2002). Transport of oil from spills
* Corresponding author.
is a major area of application as well. NOAA’s Oil Spill Response group runs a particle tracking system (GNOME) which is used for emergency response during a spill (NOAA, 2014). Additionally, they have the ability to run other scenarios in order to determine the threat level and sensitivity of various regions to likely spill locations (Barker, 1999). In the aftermath of the Deepwater Horizon oil spill in 2010, many groups ran numerical circulation models with drifter models to follow the trajectory of oil from the oil spill (e.g., Barker, 2011; Dietrich et al., 2012; Huntley et al., 2011; Liu et al., 2011; MacFadyen et al., 2011; North et al., 2011; Weisberg et al., 2011). The Deepwater Horizon oil spill led to a focus on transport in the north Gulf of Mexico, but less is known in the northwestern Gulf of Mexico. Several relatively large groups of drifters were released in the 1990s in the SCULP experiment to study the circulation of the Texas-Louisiana and Florida-Alabama shelves (LaCasce and Ohlmann, 2003; Ohlmann and Niiler, 2005). Drifter data has shown little connectivity between the Texas-Louisiana and west Florida shelf (LaCasce and Ohlmann, 2003; Ohlmann and Niiler, 2005). On the Texas-Louisiana shelf, flow tends to move along-shore – parallel to isobaths – but with cross-shore movement possible on the inner shelf due to convergent winds, and cross-shelf movement due to deeper Gulf eddies impinging at the shelf edge (Cho et al., 1998; Cochrane and Kelly, 1986; Ohlmann and Niiler, 2005). Oil spilled near Texas City in March 2014, traveled mostly downcoast (NOAA, 2015; Walpert et al., 2014). Smaller scale studies have examined larval transport and settlement in Texas bays and also noted the importance of along-coast
http://dx.doi.org/10.1016/j.marpolbul.2016.12.074 0025-326X/© 2016 Elsevier Ltd. All rights reserved.
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS 2
K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
movement (Brown et al., 2004, 2005, 2000). In a study of the physical mechanisms for harmful algal bloom initiation in the northwest Gulf of Mexico, researchers found evidence of connectivity between the southern and northwest Gulf of Mexico (Thyng et al., 2013). In this paper, we use numerical drifters in a high resolution circulation model of the Texas-Louisiana shelf to determine connectivity with the coast. We will show that while seasonal connectivity between the inner shelf and coast is largely uniform, there are seasonal variations for particular coast segments. Additionally, particular coastal areas, especially Port Aransas, have the potential to be impacted with drifters at a much higher rate than other areas in both the summer and winter.
The grid is curvilinear and has been stretched to have higher horizontal resolution near the Mississippi river delta (∼500 m) and lower near the open boundary (∼1–2 km). There are 30 vertical layers, which are stretched between 3 and 3000 m to capture the top and bottom of the water column. The ocean circulation model is oneway nested inside a HYCOM model of the Gulf of Mexico (HYCOM Consortium, 2013) for realistic boundary forcing, with surface forcing additionally of 2D wind, sea surface heat (short wave), and salt fluxes from the North American Regional Reanalysis (NARR) dataset. Data from eight rivers are input into the model domain as inflowing fresh water fluxes from the USGS (US Geological Survey) Real-Time Water Data for the Nation. More model setup details can be found in Zhang et al. (2012a).
2. Methodology 2.2. Lagrangian trajectory model 2.1. Numerical circulation model Velocity fields for this study are taken from a numerical ocean circulation model run using the Regional Ocean Modeling System (ROMS) (Shchepetkin and McWilliams, 2005). The model domain includes the Texas and Louisiana continental shelves (Fig. 1) (Hetland, 2015). This model has been used in several studies of the area (e.g., Fennel et al. (2013), Thyng et al. (2013), Zhang and Hetland (2012)). It has been previously validated with salinity data in order to understand how well the mixing processes are working (Zhang et al., 2012b), and with sea surface heights, currents, and temperature (Zhang et al., 2012a). The seasonal behavior of sea surface height, temperature, and velocities as well as the significant inertial oscillation signal are captured by the model. Additionally, the surface salinity field is spatially represented in the model output.
Transport and connectivity properties on the shelf are calculated using simulated drifters that represent passive parcels of water. Numerical drifters are run offline using velocity fields predicted by the ROMS model, which are saved every 4 h. The algorithm of the Lagrangian trajectory model is from TRACMASS, which steps numerical drifters in time natively on a staggered Arakawa C grid in such a way that allows for the maximum drifter trajectory accuracy possible given a numerical grid and the temporal frequency of circulation model output (Blanke and Raynaud, 1997; Döös, 1995; Döös et al., 2013). TRACMASS has been used in both oceanography and atmospheric studies for fundamental studies and applications. The TRACMASS algorithm, which is written in Fortran, has been wrapped in Python for running in batches of simulations; the Python system is named TracPy (Thyng and Hetland, 2014). This
Fig. 1. The region of interest. The Gulf of Mexico with the numerical domain indicated [turquoise] (A), a magnified view of the numerical domain grid with 20, 50, 100, and 500m isobaths shown (B), and a view of Galveston Bay with the actual grid resolution shown (C). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
is available online and continually in development (Thyng et al., 2014). Our model system has been shown to have skill in reproducing dispersive transport through the mean separation distance. We compared numerical drifters released within our model system to observations of in situ drifters released over the Texas shelf from the SCULP project (LaCasce and Ohlmann, 2003; Ohlmann and Niiler, 2005). The model drifters displayed a similar sharp increase in separation distance in time as the field drifters up to about 7 days, at which point drifters began exiting the numerical domain and could no longer be compared. Numerical surface-limited drifters were initialized uniformly in space with 1 km spacing, giving 267,622 drifters per experiment. Simulations of newly initialized sets of drifters were started daily for January 2004 through August 2014, and run forward in time for 30 days. Subgrid diffusion was not used in the experiments, and drifters positions were output every 48 min but calculated with fields updated in time every 9.6 min, maximum. (The time step is different for each drifter in space and time, but a maximum value was set.) Wind effects were not applied directly onto the drifters because the drifters as modeled are sitting just below the sea surface, outside of the wind’s direct effect; the wind does affect particle tracks by altering surface currents. This assumption is appropriate for representing oil slicks that have broken down into tarballs over a few days or weeks, and allows us to examine the fundamental transport pathways.
2.3. Metrics We seek regions of the shelf that are strongly connected to or disconnected from the coast, and regions of the coast that are most
3
commonly reached by water parcels from offshore. To this end, we consider two metrics: connectivity and vulnerability. Connectivity to the coastline is presented as the probability of drifter transport from a particular region to near the coastline during a set period of time – 30 days in this study. A drifter is counted as near the coastline if it reaches within 5 km, representing a range at which near-shore processes may start to dominate the transport dynamics. In other words, of the total number of drifters starting in a given box in the numerical domain, we find what percent travel into the coastal region; each drifter is counted only once no matter how many coastal boxes it enters. This metric was inspired by NOAA’s Office of Response and Restoration’s Trajectory Analysis Planner (Barker, 1999). Vulnerability of the coastline is calculated as the number of drifters that enter a coast box in a set time period (drifters can be counted more than once for this case; once for each coastal box entered). The coast box size used in this analysis is approximately 5 km by 5 km, following along the coastline outside the barrier islands and across bay mouths. For all results, seasonal metrics are calculated using drifters started in January and February for winter and in July and August for summer. Note that these metrics are for drifters reaching near the coast — a more detailed coastal model that includes surf-zone dynamics would be needed for further study of where exactly on the beach drifters would land. The two metrics used in this study are illustrated in Fig. 2. Of the sample drifters shown, those that reach the coastline region (highlighted as a transparent red band along the coast) are drawn in red, as compared with the drifters that do not reach in coastline, shown in gray. Connectivity is calculated as histograms of the starting locations (red circles) of drifters that reach the coastline. This demonstrates what areas on the shelf are connected with the coast, and therefore where we would be particularly concerned about hazardous material entering the system since there would be a direct
Fig. 2. Sample drifter trajectories are shown from a winter simulation (starting January 1st, 2008). The area along the coastline and approximately 5 km away is highlighted with a light red band. Drifters that are inside this region at some point during the 30 day simulation are colored in red; others are gray. Circles mark the drifter starting location. Labels along the coastline in the left subplot show along-coast distance [km], as well as notable locations for reference, both of which are used in coast vulnerability analysis (Fig. 4). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS 4
K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
transport path to the coast. Vulnerability measures the number of drifters that reach different areas along the coastline. Every coastal box entered by each of the sample red drifters includes a count of that drifter, as well as every other drifter that enters the box during the experiment advection time. This tells us which areas along the coast are most likely to be hit by material that enters the system and ultimately reaches the coastline.
3. Results The overall probability of drifters reaching the coast from the shelf for the winter and summer seasons is shown in Fig. 3. Darker shades of green indicate increased probability. Results are similar in winter and summer; however, the off-shore connectivity is more limited in winter as compared with summer. Connectivity near Galveston Bay is similar in both seasons, but there is a stronger offshore connection with Port Aransas in the summer. While the results are similar seasonally, we do see significant spatial differences in connectivity: the Texas coastline is more strongly connected with the nearby water whereas Louisiana has only a limited area of coastal connectivity with the rest of the numerical domain. Coastal vulnerability is also shown in Fig. 3; shades of pink show how many drifters reach each analysis box along the coastline. In the winter, the coastline from Mexico (near 24◦ N) up to around Galveston Bay is vulnerable to impact. The vulnerable region is more limited in the summer: from around Port Aransas to around Galveston Bay. In both seasons, the Port Aransas region is the most vulnerable and Galveston Bay has limited vulnerability. Conversely, the Louisiana coast shows low levels of impact of drifters relative to the Texas coast. This trend of increased coastal vulnerability at Port Aransas changes with advection time (Fig. 4). For shorter advection times (3 and 5 days, shown as more transparent lines) in both summer (orange) and winter (blue), Port Aransas is reached by drifters, but so is Laguna Madre in Mexico, Atchafalaya and Terrebone Bays, and
the San Bernard and Rockefeller State Wildlife Refuges, at about the same rate. As the advection time increases (more opaque lines), however, Port Aransas is hit much more frequently than any other location along the coastline — this again holds true in both summer and winter, but is more amplified in summer. Also notable in evaluating the coastal vulnerability for different advection times (Fig. 4) are the areas on the coastline that are either consistently hit (peaks in figure) or missed (valleys). While Port Aransas area and Matagorda Bay are hit to varying degrees with advection time, the south end of Matagorda Bay is consistently missed. Similarly, while parts of Atchafalaya and Terrebone Bays are consistently hit, an area between them is routinely missed. Results show the south end of Matagorda Island as vulnerable to impact and the south end of Matagorda Bay as typically missed — this matches the pattern of shoreline oiling from the Texas City “Y” oil spill (from Galveston Bay) in March 2014 (MacFadyen et al., 2015). The difference in connectivity for a particular region can be very different as compared with the average over the entire model domain. Fig. 5 shows the coastal connectivity and vulnerability for the area just around Galveston Bay. In the winter, drifters reaching Galveston Bay originate east in the along-shore direction, which is consistent with the drifters being advected due to the winter downcoast mean winds (Morey et al., 2005). In the summer, drifters that reach Galveston Bay tend to originate offshore, which is consistent with the summer upcoast mean wind trend. The vulnerability of the Galveston Bay coast is similar in winter and summer: more impact is typical farther south along the coast and lesser farther north. The coastal connectivity and vulnerability are shown for Port Aransas in Fig. 6. The behavior seen in this case is similar to that of Galveston: drifters originate upcoast of Port Aransas in the winter and more offshore in the summer, probably largely due to the seasonal wind patterns. The vulnerability changes from being higher on the southern end of Port Aransas in the winter to being stronger on the northern end in the summer. The area of connectivity with Port Aransas is larger than with Galveston; the reasons for this are discussed in the next section.
Fig. 3. Texas coastal connectivity and vulnerability is shown for the winter and summer seasons. The probability [%] of drifters starting in a region and traveling within 5 km of the coastline in 30 days is shown as shades of green. Shades of pink indicate the number of drifters that reach a given analysis box along the coast over the same time period. Lighter gray lines indicate the 10, 20, 50, 100, 150, 200, 250, 300, 350, 400, and 450 m isobaths. Magnified subplots show Galveston Bay (upper left) and Port Aransas (bottom right). Overlaid blue lines show U.S. shipping lanes (BOEM, 2016). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
5
Fig. 4. Mean coastal vulnerability for summer (orange solid line) and winter (blue dashed line) as a function of along-coast distance from Mexico to Louisiana; along-coast distance is shown on the map in Fig. 2 and particular locations names are indicated for reference and interpretation. Six different advection times are shown with increasing opacity: 3, 5, 10, 15, 20, and 30 days. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
4. Discussion and conclusions We find that the connectivity and early vulnerability patterns are similar in winter and summer over the full domain. There are largescale differences between Texas and Louisiana, with Texas having a broader region of connectivity – most likely due to the fresh water inputs from Louisiana pushing the flow seaward (see Bianchi et al. (2013) for a similar example in Barataria Bay). The coastal vulnerability evolution in time (Fig. 4) indicates that at first particles tend interact with the part of the coast nearest them, if at all. Later, alongshore currents can cause regions of convergence to create hotspots of vulnerability. The patterns that develop in time are robust, since
Fig. 4 shows a lot of structure. Any structure due to individual forcing events or ephemeral flow features is averaged away in this aggregate figure, which combines information from 11 years. There are three main differences that may contribute to the Port Aransas coast region having a larger area of connectivity with the shelf compared with Galveston, despite having the same main forcing mechanism of wind. First, the geometry of the coastline at Port Aransas is such that the mean wind direction in neither the winter nor summer is blocked by land in the same way that Galveston is blocked in the winter. Because of this, alongshore currents are closer to being aligned with the mean wind direction at Port Aransas, leading to more connection. Second, the shelf is narrower near Port
Fig. 5. Coastal connectivity and vulnerability for Galveston Bay for the winter and summer seasons. Overlaid gray arrows indicate the approximate seasonal mean wind direction. See Fig. 3 for more plot details.
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS 6
K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
Fig. 6. Coastal connectivity and vulnerability for Port Aransas is shown for the winter and summer seasons. Overlaid gray arrows indicate the approximate seasonal mean wind direction. See Fig. 3 for more plot details.
Aransas than at Galveston, so that deep off-shelf eddies can have more of an impact. Third, the river plume could be an important factor for the coast connectivity and is not necessarily found at both locations. The river plume commonly reaches Galveston Bay in the winter and summer, but only sometimes reaches Port Aransas in the winter (and not in the summer). These variations in the coastline geometry and forcing were also found to be important for convergent flows along the coast in Zhang and Hetland (2012). The importance of the spatial connectivity patterns shown in Fig. 3 is closely linked with the locations of hydrocarbon activity in the region. In the coastal zone (within 3 nautical miles of the shore), the input of petroleum hydrocarbons into the northwest
Gulf of Mexico is dominated by land-based consumption input to the water by rivers and runoff: 12,000 tons average annual output between 1990 and 1999 (Natural Resource Council, 2002). However, other sources include from platforms (90 tons/year), spills (870 tons/year), and recreational vehicles (770). Offshore input of petroleum hydrocarbons in the area is dominated by natural seeps (70,000 tons/year), but there are also inputs due to platforms (50 tons/year) and spills and operational discharges (1645 tons/year). Extraction of oil and gas is widespread on the northwest shelf — existing platforms are less prevalent on the Texas shelf but fill the inner Louisiana shelf (Fig. 7). There are many more platforms in Louisiana waters than Texas waters, but most of them do not
Fig. 7. Oil and gas platforms in this region in 2006 (NOAA, 2016b).
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
7
Fig. 8. Karenia brevis cell presence and absence data (NOAA, 2016a). Spatial patterns qualitatively correlate reasonably well with vulnerability pattern shown in Figs. 3 and 4.
overlap with the limited region of connectivity (Fig. 3). However, the fewer platforms in Texas mostly do overlap with the connectivity regions in both seasons. Accidental spills and operational discharges are likely to occur in shipping lanes, which in Texas overlap significantly with the areas that are connected with the coast, particularly in the summer (Fig. 3). The shelf areas that are connected with Galveston Bay (Fig. 5) and Port Aransas (Fig. 6) overlap with shipping lanes upcoast (east) of the location in the winter, but also offshore and downcoast (west) in the summer — generally we see greater connectivity with the coastline in the summer, increasing the area with potential for bringing hazardous material ashore. Additional context for the spatial connectivity patterns (Figs. 3 and 4) is the sensitivity to oil of the coastline regions. Evaluations of sensitivity are available for Texas (Gundlach et al., 1981), and the Mississippi river delta (Mendelssohn et al., 2012). In a comparison of the environmental sensitivity index (ESI) along the Texas coast (not shown, data from The Texas General Land Office (2016)) with where water parcels (representing oil) is most likely to hit, we found that the outermost coast – which mostly consists of barrier islands – is not acutely sensitive to oil. In contrast, the inland bay waters tend to be highly sensitive, with ESI values of 10 (NOAA Office of Response and Restoration, 2016a). Thus, while the impact of oil spills to the barrier islands would not be environmentally disastrous, the oil may still enter the channels to reach very sensitive bays. The coastline in Louisiana is very sensitive to oiling, but has lower vulnerability in relation to this numerical domain (not shown, data from NOAA Office of Response and Restoration (2016c)). It is worth noting that while barrier islands are not as impacted environmentally, they are popular for tourism and thus are economically important. Material transport patterns are important for issues beyond oil spills. Harmful algal blooms occur in the northwest Gulf of Mexico
and have been shown to be initiated by the physics of the flow patterns (Hetland and Campbell, 2007; Pitcher et al., 1998; Stumpf et al., 2008; Thyng et al., 2013). Indeed, we see qualitative spatial correlations between results from this work and patterns in Karenia brevis cell count detection (Fig. 8). Using NOAA’s Harmful Algal Blooms Observing System (HABSOS) mapping tool showing the presence and absence of cells from field data collections (NOAA, 2016a), we see an increase in cell presence between Brownsville and Corpus Christi. Few measurements between Corpus Christi and Galveston show the presence of cells, despite data having been taken. Our results suggest, from the perspective of the coastline, why cells can tend to aggregate and initiate blooms at Port Aransas and not in Louisiana — this is particularly visible in Fig. 4. After bloom initiation in the late summer/early fall, along-coast connectivity determines how blooms subsequently travel. On-going work by the authors shows confirmations of what is seen in the HABSOS tool: blooms tend to travel downcoast after initiation as the season shifts toward winter. Beyond harmful algal blooms, a group of researchers are currently funded to regularly collect garbage that has washed ashore along the Texas beaches and create a database to track spatial and temporal locations. Once such a database exists, it will provide a useful comparison with this work (NOAA Office of Response and Restoration, 2016b).
Acknowledgments This study was funded jointly by a grant from the BP/The Gulf of Mexico Research Initiative (SA12-09/GoMRI-006), the Texas General Land Office through TGLO Improving Hydrodynamic Predictions (Award No. 10-096-000-3927), and through Improving Oil Spill Predictions Near Shore and Across The Bay/Coastal Interface (Award No. 16-098-000-9290). The model output of the Texas-Louisiana shelf is
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS 8
K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
freely available (Hetland, 2015). Files used to run the drifter tracks and analysis files used in this work are available (https://github.com/ kthyng/shelf_transport), as is the drifter model https://github.com/ kthyng/tracpy.
References Barker, C.H., 2011. A statistical outlook for the deepwater horizon oil spill. Geophys. Monogr. Ser. 195, 237–244. Barker, C.H., 1999. The NOAA trajectory analysis planner: TAP 11. OCEANS’99 MTS/IEEE. Riding the Crest into the 21st Century. vol. 3. IEEE., pp. 1256–1261. Bianchi, T.S., Allison, M.A., Cai, W.-J., 2013. Biogeochemical Dynamics at Major River– Coastal Interfaces: Linkages With Global Change. Cambridge University Press. Blanke, B., Raynaud, S., 1997. Kinematics of the pacific equatorial undercurrent: an Eulerian and Lagrangian approach from gcm results. J. Phys. Oceanogr. 27 (6), 1038–1053. BOEM, 2016. GOMR Fairways NAD 27. July Brown, C.A., Holt, S.A., Jackson, G.A., Brooks, D.A., Holt, G.J., 2004. Simulating larval supply to estuarine nursery areas: how important are physical processes to the supply of larvae to the Aransas pass inlet? Fish. Oceanogr. 13 (3), 181–196. Brown, C.A., Jackson, G.A., Holt, S.A., Holt, G.J., 2005. Spatial and temporal patterns in modeled particle transport to estuarine habitat with comparisons to larval fish settlement patterns. Estuar. Coast. Shelf Sci. 64 (1), 33–46. Brown, C.A., Jackson, G.A., Brooks, D.A., 2000. Particle transport through a narrow tidal inlet due to tidal forcing and implications for larval transport. J. Geophys. Res. 105 (C10), 24141–24156. Cho, K., Reid, R.O., Nowlin, W.D., 1998. Objectively mapped stream function fields on the Texas-Louisiana shelf based on 32 months of moored current meter data. J. Geophys. Res. 103 (C5), 10,377–10,390. Cochrane, J.D., Kelly, F.J., 1986. Low-frequency circulation on the Texas-Louisiana continental shelf. J. Geophys. Res. 91 (C9), 10,645–10,659. Dietrich, J., Trahan, C., Howard, M., Fleming, J., Weaver, R., Tanaka, S., Yu, L., Luettich, R., Jr, Dawson, C., Westerink, J., et al. 2012. Surface trajectories of oil transport along the northern coastline of the Gulf of Mexico. Cont. Shelf Res. 41, 17–47. Döös, K., 1995. Interocean exchange of water masses. J. Geophys. Res. 100 (C7), 13–13499. Döös, K., Kjellsson, J., Jönsson, B., 2013. TRACMASS – a Lagrangian trajectory model. Preventive Methods for Coastal Protection. Springer., pp. 225–249. Fennel, K., Hu, J., Laurent, A., Marta-Almeida, M., Hetland, R., 2013. Sensitivity of hypoxia predictions for the northern Gulf of Mexico to sediment oxygen consumption and model nesting. J. Geophys. Res. 118 (2), 990–1002. Giddings, S.D., MacCready, P., Hickey, B.M., Banas, N.S., Davis, K.A., Siedlecki, S.A., Trainer, V.L., Kudela, R.M., Pelland, N.A., Connolly, T.P., 2014. Hindcasts of potential harmful algal bloom transport pathways on the Pacific northwest coast. J. Geophys. Res. 119 (4), 2439–2461. Gundlach, E.R., Finkelstein, K.J., Sadd, J.L., 1981. Impact and persistence of ixtoc i oil on the south Texas coast. International Oil Spill Conference. vol. 1981. American Petroleum Institute., pp. 477–485. 1. Henrichs, D.W., Hetland, R.D., Campbell, L., 2015. Identifying bloom origins of the toxic dinoflagellate Karenia brevis in the western Gulf of Mexico using a spatially explicit individual-based model. Ecol. Model. 313, 251–258. Hetland, R., 2015. TXLA shelf Regional Ocean Modeling System 2003-2011 hindcast for the Texas-Louisiana shelf. Distributed by: Gulf of Mexico Research Initiative Information and Data Cooperative (GRIIDC), Harte Research Institute, Texas A&M University, Corpus Christi. Hetland, R.D., Campbell, L., 2007, Oct. Convergent blooms of Karenia brevis along the Texas coast. Geophys. Res. Lett. 34 (19), L19604. October. Huntley, H.S., Lipphardt, B., Kirwan, A., 2011. Surface Drift Predictions of the Deepwater Horizon Spill: The Lagrangian Perspective. Monitoring and Modeling the Deepwater Horizon Oil Spill: A Record-Breaking Enterprise, , pp. 179–195. HYCOM Consortium, 2013. HYCOM. February. http://hycom.org/dataserver/ goml0pt04/expt-31pt0. LaCasce, J., Ohlmann, C., 2003. Relative dispersion at the surface of the Gulf of Mexico. J. Mar. Res. 61 (3), 285–312. Li, Y., He, R., Manning, J.P., 2014. Coastal connectivity in the Gulf of Maine in spring and summer of 2004–2009. Deep-Sea Res. II Top. Stud. Oceanogr. 103, 199–209. Liu, Y.Y., Weisberg, R.H., Hu, C.C., Zheng, L.L., 2011. Trajectory Forecast as a Rapid Response to the Deepwater Horizon Oil Spill. Monitoring and Modeling the Deepwater Horizon Oil Spill: A Record-Breaking Enterprise, , pp. 153–165. Lugo-Fernández, A., Deslarzes, K.J., Price, J.M., Boland, G.S., Morin, M.V., 2001. Inferring probable dispersal of flower garden banks coral larvae (Gulf of Mexico) using observed and simulated drifter trajectories. Cont. Shelf Res. 21 (1), 47–67. MacFadyen, A., Watabayashi, G., Barker, C., Beegle-Krause, C., 2011. Tactical Modeling of Surface Oil Transport During the Deepwater Horizon Spill Response. Monitoring and Modeling the Deepwater Horizon Oil Spill: A Record-Breaking Enterprise, pp. 167–178
MacFadyen, A., Watabayashi, G., Simecek-Beatty, D., Lankford, J., 2015. Trajectory modeling in support of the Texas City Y oil spill response. Gulf of Mexico Research Institute Conference 2015, Houston, TX. Mendelssohn, I.A., Andersen, G.L., Baltz, D.M., Caffey, R.H., Carman, K.R., Fleeger, J.W., Joye, S.B., Lin, Q., Maltby, E., Overton, E.B., et al. 2012. Oil impacts on coastal wetlands: implications for the Mississippi river delta ecosystem after the deepwater horizon oil spill. BioScience 62 (6), 562–574. Mitarai, S., Siegel, D., Watson, J., Dong, C., McWilliams, J., 2009. Quantifying connectivity in the coastal ocean with application to the southern California bight. J. Geophys. Res 114 (C10). Morey, S.L., Zavala-Hidalgo, J., O’Brien, J.J., 2005. The seasonal variability of continental shelf circulation in the northern and western Gulf of Mexico from a high-resolution numerical model. Geophys. Monogr. Ser. 161, 203–218. Natural Resource Council, 2002. Oil in the Sea III: Inputs, Fates and Effects. NOAA, 2014. GNOME. November NOAA, 2015. Texas City “Y” Incident. September NOAA, 2016a. Harmful Algal BloomS Observing System. September NOAA, 2016b. Ocean Explorer. July NOAA Office of Response and Restoration, 2016a. An Example Shoreline Ranking. July NOAA Office of Response and Restoration, 2016b. Marine Debris Monitoring and Assessment Project. July NOAA Office of Response and Restoration, 2016c. Download ESI Maps and GIS Data. July North, E.W., Adams, E.E., Schlag, Z., Sherwood, C.R., He, R., Hyun, K.H., Socolofsky, S.A., 2011. Simulating oil droplet dispersal from the deepwater horizon spill with a Lagrangian approach. Geophys. Monogr. Ser. 195, 217–226. Ohlmann, J.C., Niiler, P.P., 2005. Circulation over the continental shelf in the northern Gulf of Mexico. Prog. Oceanogr. 64 (1), 45–81. Olascoaga, M., Beron-Vera, F., Brand, L., Kocak, H., 2008. Tracing the early development of harmful algal blooms on the west Florida shelf with the aid of Lagrangian coherent structures. J. Geophys. Res. 113 (C12). Pitcher, G., Boyd, A., Horstman, D., Mitchell-Innes, B., 1998. Subsurface dinoflagellate populations, frontal blooms and the formation of red tide in the southern Benguela upwelling system. Mar. Ecol. Prog. Ser. 172, 253–264. Roughan, M., Macdonald, H.S., Baird, M.E., Glasby, T.M., 2011. Modelling coastal connectivity in a western boundary current: seasonal and inter-annual variability. Deep-Sea Res. II Top. Stud. Oceanogr. 58 (5), 628–644. Shchepetkin, A.F., McWilliams, J.C., 2005. The regional oceanic modeling system (ROMS): a split-explicit, free-surface, topography-following-coordinate oceanic model. Ocean Model. 9 (4), 347–404. January. Stumpf, R.P., Litaker, R.W., Lanerolle, L., Tester, P.A., 2008. Hydrodynamic accumulation of Karenia off the west coast of Florida. Cont. Shelf Res. 28 (1), 189–213. January. The Texas General Land Office, 2016. GIS Maps & Data: Environmental Sensitivity Index Shoreline. July Thyng, K.M., Barker, C.H., Jordahl, K., Cherian, D., 2014. TracPy: Wrapping the Fortran Lagrangian trajectory model TRACMASS. Proc. of the 13th Python in Science Conf. (SciPy 2014). Thyng, K.M., Hetland, R.D., 2014. TracPy: Wrapping the Fortran Lagrangian trajectory model TRACMASS. Thyng, K.M., Hetland, R.D., Ogle, M.T., Zhang, X., Chen, F., Campbell, L., 2013. Origins of Karenia brevis harmful algal blooms along the Texas coast. Limnol. Oceanogr. 3, 269–278. Walpert, J.N., Guinasso, N.L., Lee, W.L., Liu, D., Buschang, S.G., 2014. Tabs respondera quick response buoy to supplement the tabs network. Oceans-St. John’s, 2014. pp. 1–6. Walsh, J.J., Haddad, K.D., Dieterle, D.A., Weisberg, R.H., Li, Z., Yang, H., Muller-Karger, F.E., Heil, C.A., Bissett, W.P., 2002. A numerical analysis of landfall of the 1979 red tide of Karenia brevis along the west coast of Florida. Cont. Shelf Res. 22 (1), 15–38. Weisberg, R.H., Zheng, L., Liu, Y., 2011. Tracking subsurface oil in the aftermath of the deepwater horizon well blowout. Geophys. Monogr. Ser. 195, 205–215. Xue, H., Incze, L., Xu, D., Wolff, N., Pettigrew, N., 2008. Connectivity of lobster populations in the coastal gulf of Maine: Part i: circulation and larval transport potential. Ecol. Model. 210 (1), 193–211. Zhang, X., Marta-Almeida, M., Hetland, R.D., 2012a. A high-resolution pre-operational forecast model of circulation on the Texas-Louisiana continental shelf and slope. J. Oper. Oceanogr. 5 (1), 19–34. Zhang, X., Hetland, R.D., Marta-Almeida, M., DiMarco, S.F., 2012b. A numerical investigation of the Mississippi and Atchafalaya freshwater transport, filling and flushing times on the Texas-Louisiana Shelf. J. Geophys. Res. 117 (C11), C11009. November. Zhang, Z., Hetland, R.D., 2012. A numerical study on convergence of alongshore flows over the Texas-Louisiana shelf. J. Geophys. Res. 117 (C11010), 18. http:// dx.doi.org/10.1029/2012JC008145.
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
Contents lists available at ScienceDirect
Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul
Texas and Louisiana coastal vulnerability and shelf connectivity Kristen M. Thyng* , Robert D. Hetland Department of Oceanography, Texas A&M University, United States
A R T I C L E
I N F O
Article history: Received 19 November 2015 Received in revised form 5 October 2016 Accepted 28 December 2016 Available online xxxx Keywords: Modeling Coastal Connectivity Pollution Transport Drifter
A B S T R A C T A numerical study of connectivity between the continental shelf and coast in the northwestern Gulf of Mexico using a circulation model and surface-limited numerical drifters shows that despite seasonal changes in winds, the overall connectivity of the shelf with the coastline is similar in the winter and summer, though it extends more offshore in Texas in summer. However, there is a spatial pattern to the connectivity: more of the inner shelf is connected with the coast in Texas as compared with Louisiana. Subsets of the coast do have seasonal variability: the coast near both Galveston and Port Aransas has more connectivity from upcoast in the winter and from offshore and downcoast in the summer. In both seasons, we find drifters reach the Port Aransas coast most frequently, with a stronger trend in the summer. These results are important for assessing likely pathways for spilled oil and other potentially hazardous material. © 2016 Elsevier Ltd. All rights reserved.
1. Introduction The Texas coastline is at risk of impact by material from many sources including those related to the oil and gas industry, which plays a significant role in the state economy. However, not all areas along the coast are equally at risk for oil spill impacts. We seek to understand which areas along the Texas and Louisiana coastlines are more likely to be hit by oil spills, under what conditions, and from where the material can originate. Analogously, we want to know what areas are not likely to be impacted. Numerical drifter tracking has often been used for understanding connectivity between regions of interest, and where released material will travel. Biological applications have included studies on larvae movement and regional connectivity for lobsters in the Gulf of Maine (Xue et al., 2008), coral in the Gulf of Mexico (Lugo-Fernández et al., 2001), larvae into Texas bays (Brown et al., 2000), and several species along the coast of Australia (Roughan et al., 2011); connectivity for several applications in Southern California (Mitarai et al., 2009) and especially harmful algal blooms in the Gulf of Maine (Li et al., 2014); and transport pathways for phytoplankton leading to harmful algal blooms in the Pacific Northwest (Giddings et al., 2014) and the Gulf of Mexico (Henrichs et al., 2015; Olascoaga et al., 2008; Thyng et al., 2013; Walsh et al., 2002). Transport of oil from spills
* Corresponding author.
is a major area of application as well. NOAA’s Oil Spill Response group runs a particle tracking system (GNOME) which is used for emergency response during a spill (NOAA, 2014). Additionally, they have the ability to run other scenarios in order to determine the threat level and sensitivity of various regions to likely spill locations (Barker, 1999). In the aftermath of the Deepwater Horizon oil spill in 2010, many groups ran numerical circulation models with drifter models to follow the trajectory of oil from the oil spill (e.g., Barker, 2011; Dietrich et al., 2012; Huntley et al., 2011; Liu et al., 2011; MacFadyen et al., 2011; North et al., 2011; Weisberg et al., 2011). The Deepwater Horizon oil spill led to a focus on transport in the north Gulf of Mexico, but less is known in the northwestern Gulf of Mexico. Several relatively large groups of drifters were released in the 1990s in the SCULP experiment to study the circulation of the Texas-Louisiana and Florida-Alabama shelves (LaCasce and Ohlmann, 2003; Ohlmann and Niiler, 2005). Drifter data has shown little connectivity between the Texas-Louisiana and west Florida shelf (LaCasce and Ohlmann, 2003; Ohlmann and Niiler, 2005). On the Texas-Louisiana shelf, flow tends to move along-shore – parallel to isobaths – but with cross-shore movement possible on the inner shelf due to convergent winds, and cross-shelf movement due to deeper Gulf eddies impinging at the shelf edge (Cho et al., 1998; Cochrane and Kelly, 1986; Ohlmann and Niiler, 2005). Oil spilled near Texas City in March 2014, traveled mostly downcoast (NOAA, 2015; Walpert et al., 2014). Smaller scale studies have examined larval transport and settlement in Texas bays and also noted the importance of along-coast
http://dx.doi.org/10.1016/j.marpolbul.2016.12.074 0025-326X/© 2016 Elsevier Ltd. All rights reserved.
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS 2
K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
movement (Brown et al., 2004, 2005, 2000). In a study of the physical mechanisms for harmful algal bloom initiation in the northwest Gulf of Mexico, researchers found evidence of connectivity between the southern and northwest Gulf of Mexico (Thyng et al., 2013). In this paper, we use numerical drifters in a high resolution circulation model of the Texas-Louisiana shelf to determine connectivity with the coast. We will show that while seasonal connectivity between the inner shelf and coast is largely uniform, there are seasonal variations for particular coast segments. Additionally, particular coastal areas, especially Port Aransas, have the potential to be impacted with drifters at a much higher rate than other areas in both the summer and winter.
The grid is curvilinear and has been stretched to have higher horizontal resolution near the Mississippi river delta (∼500 m) and lower near the open boundary (∼1–2 km). There are 30 vertical layers, which are stretched between 3 and 3000 m to capture the top and bottom of the water column. The ocean circulation model is oneway nested inside a HYCOM model of the Gulf of Mexico (HYCOM Consortium, 2013) for realistic boundary forcing, with surface forcing additionally of 2D wind, sea surface heat (short wave), and salt fluxes from the North American Regional Reanalysis (NARR) dataset. Data from eight rivers are input into the model domain as inflowing fresh water fluxes from the USGS (US Geological Survey) Real-Time Water Data for the Nation. More model setup details can be found in Zhang et al. (2012a).
2. Methodology 2.2. Lagrangian trajectory model 2.1. Numerical circulation model Velocity fields for this study are taken from a numerical ocean circulation model run using the Regional Ocean Modeling System (ROMS) (Shchepetkin and McWilliams, 2005). The model domain includes the Texas and Louisiana continental shelves (Fig. 1) (Hetland, 2015). This model has been used in several studies of the area (e.g., Fennel et al. (2013), Thyng et al. (2013), Zhang and Hetland (2012)). It has been previously validated with salinity data in order to understand how well the mixing processes are working (Zhang et al., 2012b), and with sea surface heights, currents, and temperature (Zhang et al., 2012a). The seasonal behavior of sea surface height, temperature, and velocities as well as the significant inertial oscillation signal are captured by the model. Additionally, the surface salinity field is spatially represented in the model output.
Transport and connectivity properties on the shelf are calculated using simulated drifters that represent passive parcels of water. Numerical drifters are run offline using velocity fields predicted by the ROMS model, which are saved every 4 h. The algorithm of the Lagrangian trajectory model is from TRACMASS, which steps numerical drifters in time natively on a staggered Arakawa C grid in such a way that allows for the maximum drifter trajectory accuracy possible given a numerical grid and the temporal frequency of circulation model output (Blanke and Raynaud, 1997; Döös, 1995; Döös et al., 2013). TRACMASS has been used in both oceanography and atmospheric studies for fundamental studies and applications. The TRACMASS algorithm, which is written in Fortran, has been wrapped in Python for running in batches of simulations; the Python system is named TracPy (Thyng and Hetland, 2014). This
Fig. 1. The region of interest. The Gulf of Mexico with the numerical domain indicated [turquoise] (A), a magnified view of the numerical domain grid with 20, 50, 100, and 500m isobaths shown (B), and a view of Galveston Bay with the actual grid resolution shown (C). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
is available online and continually in development (Thyng et al., 2014). Our model system has been shown to have skill in reproducing dispersive transport through the mean separation distance. We compared numerical drifters released within our model system to observations of in situ drifters released over the Texas shelf from the SCULP project (LaCasce and Ohlmann, 2003; Ohlmann and Niiler, 2005). The model drifters displayed a similar sharp increase in separation distance in time as the field drifters up to about 7 days, at which point drifters began exiting the numerical domain and could no longer be compared. Numerical surface-limited drifters were initialized uniformly in space with 1 km spacing, giving 267,622 drifters per experiment. Simulations of newly initialized sets of drifters were started daily for January 2004 through August 2014, and run forward in time for 30 days. Subgrid diffusion was not used in the experiments, and drifters positions were output every 48 min but calculated with fields updated in time every 9.6 min, maximum. (The time step is different for each drifter in space and time, but a maximum value was set.) Wind effects were not applied directly onto the drifters because the drifters as modeled are sitting just below the sea surface, outside of the wind’s direct effect; the wind does affect particle tracks by altering surface currents. This assumption is appropriate for representing oil slicks that have broken down into tarballs over a few days or weeks, and allows us to examine the fundamental transport pathways.
2.3. Metrics We seek regions of the shelf that are strongly connected to or disconnected from the coast, and regions of the coast that are most
3
commonly reached by water parcels from offshore. To this end, we consider two metrics: connectivity and vulnerability. Connectivity to the coastline is presented as the probability of drifter transport from a particular region to near the coastline during a set period of time – 30 days in this study. A drifter is counted as near the coastline if it reaches within 5 km, representing a range at which near-shore processes may start to dominate the transport dynamics. In other words, of the total number of drifters starting in a given box in the numerical domain, we find what percent travel into the coastal region; each drifter is counted only once no matter how many coastal boxes it enters. This metric was inspired by NOAA’s Office of Response and Restoration’s Trajectory Analysis Planner (Barker, 1999). Vulnerability of the coastline is calculated as the number of drifters that enter a coast box in a set time period (drifters can be counted more than once for this case; once for each coastal box entered). The coast box size used in this analysis is approximately 5 km by 5 km, following along the coastline outside the barrier islands and across bay mouths. For all results, seasonal metrics are calculated using drifters started in January and February for winter and in July and August for summer. Note that these metrics are for drifters reaching near the coast — a more detailed coastal model that includes surf-zone dynamics would be needed for further study of where exactly on the beach drifters would land. The two metrics used in this study are illustrated in Fig. 2. Of the sample drifters shown, those that reach the coastline region (highlighted as a transparent red band along the coast) are drawn in red, as compared with the drifters that do not reach in coastline, shown in gray. Connectivity is calculated as histograms of the starting locations (red circles) of drifters that reach the coastline. This demonstrates what areas on the shelf are connected with the coast, and therefore where we would be particularly concerned about hazardous material entering the system since there would be a direct
Fig. 2. Sample drifter trajectories are shown from a winter simulation (starting January 1st, 2008). The area along the coastline and approximately 5 km away is highlighted with a light red band. Drifters that are inside this region at some point during the 30 day simulation are colored in red; others are gray. Circles mark the drifter starting location. Labels along the coastline in the left subplot show along-coast distance [km], as well as notable locations for reference, both of which are used in coast vulnerability analysis (Fig. 4). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS 4
K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
transport path to the coast. Vulnerability measures the number of drifters that reach different areas along the coastline. Every coastal box entered by each of the sample red drifters includes a count of that drifter, as well as every other drifter that enters the box during the experiment advection time. This tells us which areas along the coast are most likely to be hit by material that enters the system and ultimately reaches the coastline.
3. Results The overall probability of drifters reaching the coast from the shelf for the winter and summer seasons is shown in Fig. 3. Darker shades of green indicate increased probability. Results are similar in winter and summer; however, the off-shore connectivity is more limited in winter as compared with summer. Connectivity near Galveston Bay is similar in both seasons, but there is a stronger offshore connection with Port Aransas in the summer. While the results are similar seasonally, we do see significant spatial differences in connectivity: the Texas coastline is more strongly connected with the nearby water whereas Louisiana has only a limited area of coastal connectivity with the rest of the numerical domain. Coastal vulnerability is also shown in Fig. 3; shades of pink show how many drifters reach each analysis box along the coastline. In the winter, the coastline from Mexico (near 24◦ N) up to around Galveston Bay is vulnerable to impact. The vulnerable region is more limited in the summer: from around Port Aransas to around Galveston Bay. In both seasons, the Port Aransas region is the most vulnerable and Galveston Bay has limited vulnerability. Conversely, the Louisiana coast shows low levels of impact of drifters relative to the Texas coast. This trend of increased coastal vulnerability at Port Aransas changes with advection time (Fig. 4). For shorter advection times (3 and 5 days, shown as more transparent lines) in both summer (orange) and winter (blue), Port Aransas is reached by drifters, but so is Laguna Madre in Mexico, Atchafalaya and Terrebone Bays, and
the San Bernard and Rockefeller State Wildlife Refuges, at about the same rate. As the advection time increases (more opaque lines), however, Port Aransas is hit much more frequently than any other location along the coastline — this again holds true in both summer and winter, but is more amplified in summer. Also notable in evaluating the coastal vulnerability for different advection times (Fig. 4) are the areas on the coastline that are either consistently hit (peaks in figure) or missed (valleys). While Port Aransas area and Matagorda Bay are hit to varying degrees with advection time, the south end of Matagorda Bay is consistently missed. Similarly, while parts of Atchafalaya and Terrebone Bays are consistently hit, an area between them is routinely missed. Results show the south end of Matagorda Island as vulnerable to impact and the south end of Matagorda Bay as typically missed — this matches the pattern of shoreline oiling from the Texas City “Y” oil spill (from Galveston Bay) in March 2014 (MacFadyen et al., 2015). The difference in connectivity for a particular region can be very different as compared with the average over the entire model domain. Fig. 5 shows the coastal connectivity and vulnerability for the area just around Galveston Bay. In the winter, drifters reaching Galveston Bay originate east in the along-shore direction, which is consistent with the drifters being advected due to the winter downcoast mean winds (Morey et al., 2005). In the summer, drifters that reach Galveston Bay tend to originate offshore, which is consistent with the summer upcoast mean wind trend. The vulnerability of the Galveston Bay coast is similar in winter and summer: more impact is typical farther south along the coast and lesser farther north. The coastal connectivity and vulnerability are shown for Port Aransas in Fig. 6. The behavior seen in this case is similar to that of Galveston: drifters originate upcoast of Port Aransas in the winter and more offshore in the summer, probably largely due to the seasonal wind patterns. The vulnerability changes from being higher on the southern end of Port Aransas in the winter to being stronger on the northern end in the summer. The area of connectivity with Port Aransas is larger than with Galveston; the reasons for this are discussed in the next section.
Fig. 3. Texas coastal connectivity and vulnerability is shown for the winter and summer seasons. The probability [%] of drifters starting in a region and traveling within 5 km of the coastline in 30 days is shown as shades of green. Shades of pink indicate the number of drifters that reach a given analysis box along the coast over the same time period. Lighter gray lines indicate the 10, 20, 50, 100, 150, 200, 250, 300, 350, 400, and 450 m isobaths. Magnified subplots show Galveston Bay (upper left) and Port Aransas (bottom right). Overlaid blue lines show U.S. shipping lanes (BOEM, 2016). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
5
Fig. 4. Mean coastal vulnerability for summer (orange solid line) and winter (blue dashed line) as a function of along-coast distance from Mexico to Louisiana; along-coast distance is shown on the map in Fig. 2 and particular locations names are indicated for reference and interpretation. Six different advection times are shown with increasing opacity: 3, 5, 10, 15, 20, and 30 days. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
4. Discussion and conclusions We find that the connectivity and early vulnerability patterns are similar in winter and summer over the full domain. There are largescale differences between Texas and Louisiana, with Texas having a broader region of connectivity – most likely due to the fresh water inputs from Louisiana pushing the flow seaward (see Bianchi et al. (2013) for a similar example in Barataria Bay). The coastal vulnerability evolution in time (Fig. 4) indicates that at first particles tend interact with the part of the coast nearest them, if at all. Later, alongshore currents can cause regions of convergence to create hotspots of vulnerability. The patterns that develop in time are robust, since
Fig. 4 shows a lot of structure. Any structure due to individual forcing events or ephemeral flow features is averaged away in this aggregate figure, which combines information from 11 years. There are three main differences that may contribute to the Port Aransas coast region having a larger area of connectivity with the shelf compared with Galveston, despite having the same main forcing mechanism of wind. First, the geometry of the coastline at Port Aransas is such that the mean wind direction in neither the winter nor summer is blocked by land in the same way that Galveston is blocked in the winter. Because of this, alongshore currents are closer to being aligned with the mean wind direction at Port Aransas, leading to more connection. Second, the shelf is narrower near Port
Fig. 5. Coastal connectivity and vulnerability for Galveston Bay for the winter and summer seasons. Overlaid gray arrows indicate the approximate seasonal mean wind direction. See Fig. 3 for more plot details.
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS 6
K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
Fig. 6. Coastal connectivity and vulnerability for Port Aransas is shown for the winter and summer seasons. Overlaid gray arrows indicate the approximate seasonal mean wind direction. See Fig. 3 for more plot details.
Aransas than at Galveston, so that deep off-shelf eddies can have more of an impact. Third, the river plume could be an important factor for the coast connectivity and is not necessarily found at both locations. The river plume commonly reaches Galveston Bay in the winter and summer, but only sometimes reaches Port Aransas in the winter (and not in the summer). These variations in the coastline geometry and forcing were also found to be important for convergent flows along the coast in Zhang and Hetland (2012). The importance of the spatial connectivity patterns shown in Fig. 3 is closely linked with the locations of hydrocarbon activity in the region. In the coastal zone (within 3 nautical miles of the shore), the input of petroleum hydrocarbons into the northwest
Gulf of Mexico is dominated by land-based consumption input to the water by rivers and runoff: 12,000 tons average annual output between 1990 and 1999 (Natural Resource Council, 2002). However, other sources include from platforms (90 tons/year), spills (870 tons/year), and recreational vehicles (770). Offshore input of petroleum hydrocarbons in the area is dominated by natural seeps (70,000 tons/year), but there are also inputs due to platforms (50 tons/year) and spills and operational discharges (1645 tons/year). Extraction of oil and gas is widespread on the northwest shelf — existing platforms are less prevalent on the Texas shelf but fill the inner Louisiana shelf (Fig. 7). There are many more platforms in Louisiana waters than Texas waters, but most of them do not
Fig. 7. Oil and gas platforms in this region in 2006 (NOAA, 2016b).
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
7
Fig. 8. Karenia brevis cell presence and absence data (NOAA, 2016a). Spatial patterns qualitatively correlate reasonably well with vulnerability pattern shown in Figs. 3 and 4.
overlap with the limited region of connectivity (Fig. 3). However, the fewer platforms in Texas mostly do overlap with the connectivity regions in both seasons. Accidental spills and operational discharges are likely to occur in shipping lanes, which in Texas overlap significantly with the areas that are connected with the coast, particularly in the summer (Fig. 3). The shelf areas that are connected with Galveston Bay (Fig. 5) and Port Aransas (Fig. 6) overlap with shipping lanes upcoast (east) of the location in the winter, but also offshore and downcoast (west) in the summer — generally we see greater connectivity with the coastline in the summer, increasing the area with potential for bringing hazardous material ashore. Additional context for the spatial connectivity patterns (Figs. 3 and 4) is the sensitivity to oil of the coastline regions. Evaluations of sensitivity are available for Texas (Gundlach et al., 1981), and the Mississippi river delta (Mendelssohn et al., 2012). In a comparison of the environmental sensitivity index (ESI) along the Texas coast (not shown, data from The Texas General Land Office (2016)) with where water parcels (representing oil) is most likely to hit, we found that the outermost coast – which mostly consists of barrier islands – is not acutely sensitive to oil. In contrast, the inland bay waters tend to be highly sensitive, with ESI values of 10 (NOAA Office of Response and Restoration, 2016a). Thus, while the impact of oil spills to the barrier islands would not be environmentally disastrous, the oil may still enter the channels to reach very sensitive bays. The coastline in Louisiana is very sensitive to oiling, but has lower vulnerability in relation to this numerical domain (not shown, data from NOAA Office of Response and Restoration (2016c)). It is worth noting that while barrier islands are not as impacted environmentally, they are popular for tourism and thus are economically important. Material transport patterns are important for issues beyond oil spills. Harmful algal blooms occur in the northwest Gulf of Mexico
and have been shown to be initiated by the physics of the flow patterns (Hetland and Campbell, 2007; Pitcher et al., 1998; Stumpf et al., 2008; Thyng et al., 2013). Indeed, we see qualitative spatial correlations between results from this work and patterns in Karenia brevis cell count detection (Fig. 8). Using NOAA’s Harmful Algal Blooms Observing System (HABSOS) mapping tool showing the presence and absence of cells from field data collections (NOAA, 2016a), we see an increase in cell presence between Brownsville and Corpus Christi. Few measurements between Corpus Christi and Galveston show the presence of cells, despite data having been taken. Our results suggest, from the perspective of the coastline, why cells can tend to aggregate and initiate blooms at Port Aransas and not in Louisiana — this is particularly visible in Fig. 4. After bloom initiation in the late summer/early fall, along-coast connectivity determines how blooms subsequently travel. On-going work by the authors shows confirmations of what is seen in the HABSOS tool: blooms tend to travel downcoast after initiation as the season shifts toward winter. Beyond harmful algal blooms, a group of researchers are currently funded to regularly collect garbage that has washed ashore along the Texas beaches and create a database to track spatial and temporal locations. Once such a database exists, it will provide a useful comparison with this work (NOAA Office of Response and Restoration, 2016b).
Acknowledgments This study was funded jointly by a grant from the BP/The Gulf of Mexico Research Initiative (SA12-09/GoMRI-006), the Texas General Land Office through TGLO Improving Hydrodynamic Predictions (Award No. 10-096-000-3927), and through Improving Oil Spill Predictions Near Shore and Across The Bay/Coastal Interface (Award No. 16-098-000-9290). The model output of the Texas-Louisiana shelf is
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074
ARTICLE IN PRESS 8
K. Thyng, R. Hetland / Marine Pollution Bulletin xxx (2016) xxx–xxx
freely available (Hetland, 2015). Files used to run the drifter tracks and analysis files used in this work are available (https://github.com/ kthyng/shelf_transport), as is the drifter model https://github.com/ kthyng/tracpy.
References Barker, C.H., 2011. A statistical outlook for the deepwater horizon oil spill. Geophys. Monogr. Ser. 195, 237–244. Barker, C.H., 1999. The NOAA trajectory analysis planner: TAP 11. OCEANS’99 MTS/IEEE. Riding the Crest into the 21st Century. vol. 3. IEEE., pp. 1256–1261. Bianchi, T.S., Allison, M.A., Cai, W.-J., 2013. Biogeochemical Dynamics at Major River– Coastal Interfaces: Linkages With Global Change. Cambridge University Press. Blanke, B., Raynaud, S., 1997. Kinematics of the pacific equatorial undercurrent: an Eulerian and Lagrangian approach from gcm results. J. Phys. Oceanogr. 27 (6), 1038–1053. BOEM, 2016. GOMR Fairways NAD 27. July Brown, C.A., Holt, S.A., Jackson, G.A., Brooks, D.A., Holt, G.J., 2004. Simulating larval supply to estuarine nursery areas: how important are physical processes to the supply of larvae to the Aransas pass inlet? Fish. Oceanogr. 13 (3), 181–196. Brown, C.A., Jackson, G.A., Holt, S.A., Holt, G.J., 2005. Spatial and temporal patterns in modeled particle transport to estuarine habitat with comparisons to larval fish settlement patterns. Estuar. Coast. Shelf Sci. 64 (1), 33–46. Brown, C.A., Jackson, G.A., Brooks, D.A., 2000. Particle transport through a narrow tidal inlet due to tidal forcing and implications for larval transport. J. Geophys. Res. 105 (C10), 24141–24156. Cho, K., Reid, R.O., Nowlin, W.D., 1998. Objectively mapped stream function fields on the Texas-Louisiana shelf based on 32 months of moored current meter data. J. Geophys. Res. 103 (C5), 10,377–10,390. Cochrane, J.D., Kelly, F.J., 1986. Low-frequency circulation on the Texas-Louisiana continental shelf. J. Geophys. Res. 91 (C9), 10,645–10,659. Dietrich, J., Trahan, C., Howard, M., Fleming, J., Weaver, R., Tanaka, S., Yu, L., Luettich, R., Jr, Dawson, C., Westerink, J., et al. 2012. Surface trajectories of oil transport along the northern coastline of the Gulf of Mexico. Cont. Shelf Res. 41, 17–47. Döös, K., 1995. Interocean exchange of water masses. J. Geophys. Res. 100 (C7), 13–13499. Döös, K., Kjellsson, J., Jönsson, B., 2013. TRACMASS – a Lagrangian trajectory model. Preventive Methods for Coastal Protection. Springer., pp. 225–249. Fennel, K., Hu, J., Laurent, A., Marta-Almeida, M., Hetland, R., 2013. Sensitivity of hypoxia predictions for the northern Gulf of Mexico to sediment oxygen consumption and model nesting. J. Geophys. Res. 118 (2), 990–1002. Giddings, S.D., MacCready, P., Hickey, B.M., Banas, N.S., Davis, K.A., Siedlecki, S.A., Trainer, V.L., Kudela, R.M., Pelland, N.A., Connolly, T.P., 2014. Hindcasts of potential harmful algal bloom transport pathways on the Pacific northwest coast. J. Geophys. Res. 119 (4), 2439–2461. Gundlach, E.R., Finkelstein, K.J., Sadd, J.L., 1981. Impact and persistence of ixtoc i oil on the south Texas coast. International Oil Spill Conference. vol. 1981. American Petroleum Institute., pp. 477–485. 1. Henrichs, D.W., Hetland, R.D., Campbell, L., 2015. Identifying bloom origins of the toxic dinoflagellate Karenia brevis in the western Gulf of Mexico using a spatially explicit individual-based model. Ecol. Model. 313, 251–258. Hetland, R., 2015. TXLA shelf Regional Ocean Modeling System 2003-2011 hindcast for the Texas-Louisiana shelf. Distributed by: Gulf of Mexico Research Initiative Information and Data Cooperative (GRIIDC), Harte Research Institute, Texas A&M University, Corpus Christi. Hetland, R.D., Campbell, L., 2007, Oct. Convergent blooms of Karenia brevis along the Texas coast. Geophys. Res. Lett. 34 (19), L19604. October. Huntley, H.S., Lipphardt, B., Kirwan, A., 2011. Surface Drift Predictions of the Deepwater Horizon Spill: The Lagrangian Perspective. Monitoring and Modeling the Deepwater Horizon Oil Spill: A Record-Breaking Enterprise, , pp. 179–195. HYCOM Consortium, 2013. HYCOM. February. http://hycom.org/dataserver/ goml0pt04/expt-31pt0. LaCasce, J., Ohlmann, C., 2003. Relative dispersion at the surface of the Gulf of Mexico. J. Mar. Res. 61 (3), 285–312. Li, Y., He, R., Manning, J.P., 2014. Coastal connectivity in the Gulf of Maine in spring and summer of 2004–2009. Deep-Sea Res. II Top. Stud. Oceanogr. 103, 199–209. Liu, Y.Y., Weisberg, R.H., Hu, C.C., Zheng, L.L., 2011. Trajectory Forecast as a Rapid Response to the Deepwater Horizon Oil Spill. Monitoring and Modeling the Deepwater Horizon Oil Spill: A Record-Breaking Enterprise, , pp. 153–165. Lugo-Fernández, A., Deslarzes, K.J., Price, J.M., Boland, G.S., Morin, M.V., 2001. Inferring probable dispersal of flower garden banks coral larvae (Gulf of Mexico) using observed and simulated drifter trajectories. Cont. Shelf Res. 21 (1), 47–67. MacFadyen, A., Watabayashi, G., Barker, C., Beegle-Krause, C., 2011. Tactical Modeling of Surface Oil Transport During the Deepwater Horizon Spill Response. Monitoring and Modeling the Deepwater Horizon Oil Spill: A Record-Breaking Enterprise, pp. 167–178
MacFadyen, A., Watabayashi, G., Simecek-Beatty, D., Lankford, J., 2015. Trajectory modeling in support of the Texas City Y oil spill response. Gulf of Mexico Research Institute Conference 2015, Houston, TX. Mendelssohn, I.A., Andersen, G.L., Baltz, D.M., Caffey, R.H., Carman, K.R., Fleeger, J.W., Joye, S.B., Lin, Q., Maltby, E., Overton, E.B., et al. 2012. Oil impacts on coastal wetlands: implications for the Mississippi river delta ecosystem after the deepwater horizon oil spill. BioScience 62 (6), 562–574. Mitarai, S., Siegel, D., Watson, J., Dong, C., McWilliams, J., 2009. Quantifying connectivity in the coastal ocean with application to the southern California bight. J. Geophys. Res 114 (C10). Morey, S.L., Zavala-Hidalgo, J., O’Brien, J.J., 2005. The seasonal variability of continental shelf circulation in the northern and western Gulf of Mexico from a high-resolution numerical model. Geophys. Monogr. Ser. 161, 203–218. Natural Resource Council, 2002. Oil in the Sea III: Inputs, Fates and Effects. NOAA, 2014. GNOME. November NOAA, 2015. Texas City “Y” Incident. September NOAA, 2016a. Harmful Algal BloomS Observing System. September NOAA, 2016b. Ocean Explorer. July NOAA Office of Response and Restoration, 2016a. An Example Shoreline Ranking. July NOAA Office of Response and Restoration, 2016b. Marine Debris Monitoring and Assessment Project. July NOAA Office of Response and Restoration, 2016c. Download ESI Maps and GIS Data. July North, E.W., Adams, E.E., Schlag, Z., Sherwood, C.R., He, R., Hyun, K.H., Socolofsky, S.A., 2011. Simulating oil droplet dispersal from the deepwater horizon spill with a Lagrangian approach. Geophys. Monogr. Ser. 195, 217–226. Ohlmann, J.C., Niiler, P.P., 2005. Circulation over the continental shelf in the northern Gulf of Mexico. Prog. Oceanogr. 64 (1), 45–81. Olascoaga, M., Beron-Vera, F., Brand, L., Kocak, H., 2008. Tracing the early development of harmful algal blooms on the west Florida shelf with the aid of Lagrangian coherent structures. J. Geophys. Res. 113 (C12). Pitcher, G., Boyd, A., Horstman, D., Mitchell-Innes, B., 1998. Subsurface dinoflagellate populations, frontal blooms and the formation of red tide in the southern Benguela upwelling system. Mar. Ecol. Prog. Ser. 172, 253–264. Roughan, M., Macdonald, H.S., Baird, M.E., Glasby, T.M., 2011. Modelling coastal connectivity in a western boundary current: seasonal and inter-annual variability. Deep-Sea Res. II Top. Stud. Oceanogr. 58 (5), 628–644. Shchepetkin, A.F., McWilliams, J.C., 2005. The regional oceanic modeling system (ROMS): a split-explicit, free-surface, topography-following-coordinate oceanic model. Ocean Model. 9 (4), 347–404. January. Stumpf, R.P., Litaker, R.W., Lanerolle, L., Tester, P.A., 2008. Hydrodynamic accumulation of Karenia off the west coast of Florida. Cont. Shelf Res. 28 (1), 189–213. January. The Texas General Land Office, 2016. GIS Maps & Data: Environmental Sensitivity Index Shoreline. July Thyng, K.M., Barker, C.H., Jordahl, K., Cherian, D., 2014. TracPy: Wrapping the Fortran Lagrangian trajectory model TRACMASS. Proc. of the 13th Python in Science Conf. (SciPy 2014). Thyng, K.M., Hetland, R.D., 2014. TracPy: Wrapping the Fortran Lagrangian trajectory model TRACMASS. Thyng, K.M., Hetland, R.D., Ogle, M.T., Zhang, X., Chen, F., Campbell, L., 2013. Origins of Karenia brevis harmful algal blooms along the Texas coast. Limnol. Oceanogr. 3, 269–278. Walpert, J.N., Guinasso, N.L., Lee, W.L., Liu, D., Buschang, S.G., 2014. Tabs respondera quick response buoy to supplement the tabs network. Oceans-St. John’s, 2014. pp. 1–6. Walsh, J.J., Haddad, K.D., Dieterle, D.A., Weisberg, R.H., Li, Z., Yang, H., Muller-Karger, F.E., Heil, C.A., Bissett, W.P., 2002. A numerical analysis of landfall of the 1979 red tide of Karenia brevis along the west coast of Florida. Cont. Shelf Res. 22 (1), 15–38. Weisberg, R.H., Zheng, L., Liu, Y., 2011. Tracking subsurface oil in the aftermath of the deepwater horizon well blowout. Geophys. Monogr. Ser. 195, 205–215. Xue, H., Incze, L., Xu, D., Wolff, N., Pettigrew, N., 2008. Connectivity of lobster populations in the coastal gulf of Maine: Part i: circulation and larval transport potential. Ecol. Model. 210 (1), 193–211. Zhang, X., Marta-Almeida, M., Hetland, R.D., 2012a. A high-resolution pre-operational forecast model of circulation on the Texas-Louisiana continental shelf and slope. J. Oper. Oceanogr. 5 (1), 19–34. Zhang, X., Hetland, R.D., Marta-Almeida, M., DiMarco, S.F., 2012b. A numerical investigation of the Mississippi and Atchafalaya freshwater transport, filling and flushing times on the Texas-Louisiana Shelf. J. Geophys. Res. 117 (C11), C11009. November. Zhang, Z., Hetland, R.D., 2012. A numerical study on convergence of alongshore flows over the Texas-Louisiana shelf. J. Geophys. Res. 117 (C11010), 18. http:// dx.doi.org/10.1029/2012JC008145.
Please cite this article as: K. Thyng, R. Hetland, Texas and Louisiana coastal vulnerability and shelf connectivity, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.12.074