Natural Seep Inventory Final Report
back to Energy

The Natural Oil Seep Inventory was conducted from April, 2002, to March, 2004, jointly by:

Santa Barbara County, Planning and Development Department, Energy Division, and

U.S. Geological Survey, Western Coastal and Marine Geology Team, Menlo Park, CA

The study was funded in part by: State of California Resources Agency, Coastal Resources Grant Program, Project No. 42-A-01, Contract No. OCA13008

This report includes:

  1. Excerpts from the County’s summary final report to the Resources Agency (May 17, 2004);
  2. The USGS Final Report for the project (March 25, 2004), except for the following:

These data tables are not included here due to file size constraints and because the detailed data would be of little interest to most readers. The data can be provided on request.

For study background information, follow the link to Natural Seep Inventory. For further information, contact Dr. John Day.

 

Contents

Excerpts from County’s summary final report

Natural Seep Inventory and Identification [USGS Final Report]

Preliminary Conclusions

Acknowledgements

References

Appendix 1A. Geochemical Parameters

Appendix 2. Maps of each sampled beach with tarball groups

Appendix 3. Survey of tar collection and analyses during quarters 1 thru 4

Appendix 4. Tar mass histograms for each beach

Abstract

This project was conducted in conjunction with a U.S. Geological Service (USGS) project funded by the Minerals Management Service (MMS), utilizing the expertise of USGS research scientists and resources of their Menlo Park laboratory. We conducted monthly sampling at 10 Santa Barbara County beaches in order to establish a baseline estimate of the amount of tar present. Field crews composed of geology students from the University of California at Santa Barbara (UCSB) weighed tar and counted tar balls during a 12-month period. Sampling was extended an additional 4 months by USGS personnel. 346 tar samples were collected in 637 beach transects. More than 182 of the beach tar samples were analyzed for persistent hydrocarbons ("biomarkers") and isotopic composition. The samples analyzed include these beach tar samples, plus additional samples collected by the USGS from beaches, natural offshore oil seeps, and offshore production platforms. Biomarker ratios were computed from the analytical data. Clustering and principal component methods were applied to the biomarker data, resulting in a classification of the tar samples into 9 groups. The transect sampling data were used to estimate the amount of tar present per kilometer of shoreline of each beach during each month of the project. The locations of samples from the various tar groups were mapped using color coded dots, providing a graphical picture of the location of seep sources in relation to beaches where corresponding tar was found. Estimated beach tar amounts were presented in bar graphs.

The study has provided significant new information about the deposition of tar originating in offshore natural oil seeps onto Santa Barbara’s beaches. The tar "fingerprint" database and data on amounts of tar deposited on the beaches developed during the study contributes to the knowledge of baseline beach tar conditions. However, the study is not conclusive. Additional analysis of samples from natural seeps and production platforms will be needed to determine sources of some of the beach tar samples, and to refine the classification model and test its limitations in discriminating natural seep oil from platform-produced oil. The estimates of beach tar amounts indicate that the amount of tar present is highly variable. It may be possible to reanalyze the data set in the future, controlling for some of the factors responsible for the variability, such as tides and currents. An error analysis of the tar deposition data should be done.

Findings & Conclusions – In Brief

Patterns of Beach Tar Deposition

The tar and oil samples analyzed in this study could be classified into 9 groups, based on differences in chemical composition, as reflected in isotope analysis and ratios of various persistent hydrocarbons ("biomarkers"). Each beach tar sample falls into one of the groups. In some cases, oil sampled at a natural seep and tar collected at nearby depositional beaches fall into the same group. In other cases, nearby seeps that match the beach tar are not identified. It appears that some samples may originate in seeps near the Channel islands. The tar groups represented in the samples vary from beach to beach. At some beaches (e.g., Sacate Beach), almost all samples are associated with no more than two groups, while the tar collected at some other beaches is more diverse. The correspondence between tar sources and deposition is incomplete. Collection and analysis of additional oil from offshore seeps will be needed to fill in the gaps. The depositional patterns are discussed and mapped in the accompanying USGS report.

Discrimination of Seep Tar from Oil Produced at Platforms

Biomarker analysis is able to distinguish oil from distinct sources with confidence. Thus, beach tar believed to originate from natural seeps can be distinguished from oil produced at some platforms (e.g., Point Arguello platforms; Irene). However, samples of oil produced at Platform Holly (offshore of Coal Oil Point) and Platform "A" (in the eastern Santa Barbara Channel) are very similar to much of the oil found on Santa Barbara’s southern beaches. The biomarker method so far has been unable distinguish these oils. It is uncertain whether the inability to distinguish them represents a limitation of the method (which potentially could be overcome with refinements to the method), or whether the sources of the produced oil and beach tar are identical or nearly so.

Tar Deposition Rates and Variability

The amount of tar collected at each beach was converted into estimates of tar mass and number of tar balls per kilometer of beach. Comparisons of tar deposition at the 10 beaches are presented in graphs in the attached report. There are obvious differences in tar mass and numbers of tar balls among the beaches, with generally greater amounts occurring on northern beaches. Among the South Coast beaches sampled, Coal Oil Point receives the largest amount of tar, whereas Jalama and Surf beaches are most heavily tarred of the beaches north of Point Conception. Such differences are expected, due to the location of offshore seeps relative to the depositional beaches, the prevailing current patterns, the orientation of the shoreline, etc. On average, tar balls found on the four northern beaches were substantially larger than those found on the South Coast beaches. Many large tar blobs and ropes were found on the north beaches, whereas small tar balls and tiny flecks were more typical in the south. This finding is consistent with the fact that oil in the more northerly fields is generally heavier and more viscous than that found off the South Coast. Tar deposition appears to vary seasonally, with the greatest accumulations observed during summer and fall. The data shows large variability in tar deposition from month to month, and differences in this variability among beaches. Factors responsible for the variability are discussed below.

Complicating Factors

Part of the variability observed tar deposits may be attributable to under-sampling, both spatially and temporally. Part may be due to the fact that the accumulated tar deposit can change greatly over a single tidal cycle. A likely major cause of variability is that differences in wind, current, and surf energy affect how much tar can reach and strand on a beach; these effects vary among beaches and seasonally. Wind speed and surface chop affect natural dispersion of floating oil in the water column. Current direction affects the trajectory of floating oil and weathered tar balls. Surf and tidal action can wash the beach clean, or bury tar under the sand, or release previously sequestered tar. Variations in seep activity of different seeps may also be an important factor.

It may be possible, in a future analysis of the tar sample data, to explain some of the variability by taking into consideration wind, currents, and tides at the time of sampling, in relation to the relative locations of seeps and the beaches. However, at this point, the variability cannot be adequately explained.

Establishing a Baseline of Beach Tar Deposition

The tar "fingerprint" database contributes to the understanding of baseline conditions, and may aid in determining the cause and responsible party in the case of future unusual beach oiling events. The database constitutes a record of what types of oil, as classified by biomarker composition, are commonly found at points along the coastline (during the year 2002-2003).

The study has established the normal range of tar mass deposited during the year, and a qualitative description of its appearance. This information could aid in judging whether a future tarring event is unnatural and in determining whether or beach tar is likely to have originated from ongoing, natural seepage. However, the description of baseline conditions developed in the study is not definitive. In some cases it may not be possible to distinguish spilled oil from natural seep tar, based on the current database and tar "fingerprinting" methods. Further analysis of oil samples from platforms and natural seeps, and further refinement and validation of the tar classification model, should lead to improved differentiation of tar groups in the future.

 

 

 

Natural Seep Inventory and Identification for the County of Santa Barbara, California

Final Report

March 25, 2004

 

Prepared for the County of Santa Barbara by the U.S. Geological Survey,

Western Coastal and Marine Geology Team,

345 Middlefield Road Menlo Park, CA, 94025

Thomas D. Lorenson

Jennifer A. Dougherty

Frances D. Hostettler

Robert J. Rosenbauer

INTRODUCTION

The Santa Barbara County, California coastline contains long stretches of sandy beaches, rocky inlets, high cliffs hanging precipitously over crashing waves, and many other scenic wonders. This beautiful natural resource is, however, continually exposed to contamination from both natural and anthropogenic sources. In particular, the coastline is impacted by petroleum hydrocarbons that occur as tarballs washed up all along the shorelines and as onshore seepages from rocky outcrops and cliff faces. Natural sources for these petroleum hydrocarbons include prolific, frequently chronic, onshore and offshore shallow oil seeps, especially prominent along the southern California coast (State Lands Commission Staff Report, 1977). Anthropogenic sources include possible accidental oil spills from commercial vessel traffic, from offshore drilling rigs, and from ships involved in the processing and transport of oil along the coastal shipping lanes.

Differentiating between natural and anthropogenic petroleum sources and determining specific sources of coastal contamination is essential to evaluate threats to the ecosystems and to limit contaminant impact. Although crude oils and source rocks in the California borderland oil fields have been extensively characterized (e.g., Curiale et al., 1985), published geochemical work on the substantial (i.e., approximately 20,000 tonnes/year discharged into the ocean, as estimated by a U.S. Academy of Sciences report, NAS, 2002) hydrocarbon beach tar accumulations along the California coast is limited. Reed and Kaplan (1977) used stable isotopic ratios of sulfur, nitrogen, and carbon to distinguish seep oils, beach tars, and crude oils from the southern California Borderland. Another early study utilized stable isotopic ratios of carbon and sulfur and total sulfur content of asphaltene fractions to correlate beach tars deposited near Los Angeles with their probable sources, to distinguish natural seep oils from imported tanker crude oils and local production wells, and to evaluate seasonal distribution patterns and transport (Hartman and Hammond, 1981). A more recent study used various molecular parameters of tar residues on beaches within the Monterey Bay National Marine Sanctuary to try to ascertain sources (Kvenvolden et al., 2000). Finally, a preliminary report on coastal tar and oil seeps considers the geologic framework and some potential tarball correlations related to this study (Kvenvolden and Hostettler, 2003).

These works all conclude that much of the tar accumulation originates from the Miocene Monterey Formation. Source rock in the Monterey Formation shares several chemical characteristics with local tars, including: 1) unusually "heavy" d13C (around –23‰); 2) aliphatic biomarker parameters 28,30-bisnorhopane indicating an anoxic marine depositional environment, (Curiale et al., 1985), high C35 ab-hopane 22S and 22R epimers compared to C34, and the presence of gammacerane (Peters and Moldowan, 1993); 3) a characteristic value (>3) for the biomarker parameter called "the triplet" (Kvenvolden et al., 1995), defined in Appendix 1; 4) a small but consistent presence of oleanane; 5) sterane parameters indicating low maturity versus fully mature hopane parameters; 6) very low diasteranes relative to regular steranes, indicating a clastic-poor marine source rock; and, 7) abundant aromatized steranes, especially monoaromatics relative to triaromatics, indicating low thermal maturity (Curiale et al., 1985); and 8) sulfur-rich PAH, such as dibenzothiophenes.

Although the above chemical components are common to all the tarballs, their relative proportions vary. A fingerprinting technique utilizing ratios of these constituents, and other biomarker parameters from both the aliphatic and aromatic hydrocarbon suites, allow discrimination among different tar samples. Tars can be correlated with each other and with distant sources.

The chemical composition of the tarballs is linked to its geochemical history. Despite the large number of offshore shallow hydrocarbon seeps, and the constant impingement of tar onto the shoreline, little is known about the mechanics of hydrocarbon formation in shallow seeps, specific sources of tarballs, or their transport from the marine environment onto the shore. There is at present no irrefutable data linking tar on beaches to specific offshore natural seeps (Leifer et al., 2002).

Because many of the tarballs from offshore seeps are transported significant distances from their sources by ocean currents, geochemical assignment of their origin provides insight into the circulation patterns of the coastal currents. The circulation patterns within the Santa Barbara Channel are well studied (Hickey, 1998; Winant et al., 1999). Persistent cyclonic circulation, upwelling conditions, and wind-relaxing drive the currents in a season dependent pattern. The net result of drifter studies is a combination of in-channel deposition, both on the mainland coast and on the Channel Islands, with flow predominantly toward the south and east in the spring and summer (California Current) and to the west and north in the late fall and winter (Davidson Current and the Southern California Countercurrent). Mapping depositional sites of tarballs that also drift with these ocean currents, complement these drifter studies, as well as provide information on the fate of these petrogenic contaminants in the coastal environment.

This study examines the possible origins of tars and provides qualitative rates of deposition measured over a three-year period. Another aim of this work is to build a library of coastal tar fingerprints as a database for future investigations. This work has been complimentary to a study funded by the Mineral Management Service with similar goals that has been ongoing from 2001 to 2004. These works will be combined and reported in a peer-reviewed scientific journal after completion of both studies.

Figure 1. Locations of beaches, sampled oil platforms and natural seep samples.

 

Figure 2. Viscous tar sampled offshore Point. Conception. This tar differs in both morphology and chemical composition from oil and tar found offshore the southern coast of Santa Barbara.

Figure 3. Sampling a natural oil slick. Inset shows the evolution of oil and mousse into tar patties that are often found on the southern Santa Barbara County coastline.


Figure 4. Fresh tar residue at Coal Oil Point. Photograph was taken on June 10, 2003.

 

Figure 5. Tarball morphology typically found on Santa Barbara Counties’ northern beaches.

METHODS

A total of 346 tarballs and tar residues were collected from coastal locations over 16 months. Tar samples are listed with descriptions and geochemical parameters in Appendix 1B, summarized in Table 1 and Appendix 3, and mapped for each beach in Appendix 2. Each tarball was collected from rocks or sand with a clean knife and placed in pre-cleaned glass jars for transport to the laboratory. Tars floating in the salt water were placed in a clean glass jar, and any water was poured off before analysis. Tar samples were dissolved in dichloromethane (DCM), filtered through glass wool to remove particulates, and air-dried under a hood to remove the DCM. After filtration and removal of DCM, an aliquot was taken of the clean extract for bulk stable carbon isotope analysis. The results are reported in the d notation in parts per thousand (‰) relative to the Pee Dee Belemnite (PDB) standard.

A second portion of the extract (~ 25 mg) was dissolved as completely as possible by sonication and mechanical agitation in 5 ml of hexane. This solution was then loaded onto a liquid chromatography column for compound class separation. Columns were prepared with the addition of about 5 mm of activated copper at the bottom (to remove elemental sulfur), 2.5 g of 5% deactivated neutral alumina and 2.5 g and 5.0 g of 62 and 923 silica gels, respectively. Two separate fractions were collected—saturate (hexane eluent) and aromatic (30% DCM eluent) and analyzed by gas chromatography/mass spectrometry (GC/MS). Compounds were identified by comparison with known standards or with published reference spectra. Isoprenoids and n-Alkanes, and a suite of b-carotenoid-related compounds were profiled with extracted ion (EI) chromatograms (m/z 57 and 125, respectively). Selected biomarker ratios, listed below, were calculated from GC/MS/select ion monitoring (SIM) chromatograms of m/z 191 (terpanes/hopanes) and 217 (steranes) using peak heights. 25,28,30-Trisnorhopane (T177) and the presence or absence of a 25-norhopane series was monitored by m/z 177. Extracted ion profiles from Total ion counts (TIC) of the aliphatic and aromatic hydrocarbon fractions were used for the following ions: m/z 253 for monoaromatic steroids (M, summed from contributions in both the aromatic and aliphatic fractions), m/z 231 for triaromatic steroids (T), m/z 242 for monomethyl chrysenes, and m/z 212 and 206 for dimethyl-, and m/z 226 and 220 for trimethyl-dibenzothiophenes and phenanthrenes, respectively. Either summed areas or peak heights (see Appendix I) of these compounds were used to determine other parameter ratios. The biomarker and isotope values were used to correlate the samples and group them according to their probable source locations.

Beach sampling protocols

The selection of beaches for sampling was based on several factors; geographic distribution over the coastal length of Santa Barbara County, a variety of lengths ranging from 0.2 to 8 km, access, and proximity to known oil seeps.

The beaches were sampled monthly by teams of 2 to 4 people. Typically a total of five perpendicular transects per beach were randomly chosen Transects were located by walking an arbitrary number of minutes down the beach determined by a digital counter. Each transect length was measured using a metric tape. The tape was left on the beach while team members picked up, counted and weighed tarballs found within 2 meters of the transect center line. A random tarball was weighed and stored in a chemically clean sample jar from each transect for chemical fingerprint analyses. The relative freshness of the sample was noted on a scale of 1 to 3, 1 being fresh. A global positioning receiver was used to record the location of each sample, and that of each transect. The spatial record of the samples and transects is stored on a GIS (geographic information system) database located at the US Geological Survey and will be made available. The four northern county beaches were previously monitored in a similar fashion on a quarterly basis from July 2001 through April, 2002, thus the record for these beaches extends from July 2001 through August 2003.

North Beach Team

Beach name: Jalama

 Date: 12/13/02

 Time: 12:00 pm

 Tide: +1.5

    

Personnel: Jamie Jones, Kris Broderick

          
 

Transect 1

Transect 2

Transect 3

Transect4

Transect5

Number of Tarballs

5

14

9

10

16

Total Tarball weight (gm)

518

38

22

66

44

Tarball sample weight (gm)

8

11

  

12

20

Tarball sample freshness 1-3 (1 fresh, 3 old)

2

1

  

2

1

Latitude (ddd, WGS84)

34.51671

34.51860

  

34.50759

34.50647

Longitude (ddd, WGS84)

120.50638

120.50809

  

120.50072

120.50025

Length of Transect (m, from water to cliff)

31

19

23

29

21

Azimuth of Transect (magnetic, 0-360°)

210

220

221

124

133

Time (min) to next Transect

4

1

1

2

1

Distance from cliff to sample (m)

9

11

  

2

3

South Beach Team

Beach name: Sacate

Date: 3/8/03

Time: 7:00a

Tide: +1

Personnel: Sara & Karen

Transect 1

Transect 2

Transect 3

Transect4

Transect5

Number of Tarballs

39

6

2

4

3

Total Tarball weight (gm)

32

45

5

32

9

Tarball sample weight (gm)

-

32

-

30

-

Tarball sample freshness 1-3 (1 fresh, 3 old)

-

2

-

3

-

Latitude (ddd, WGS84) N

34.46970

34.47020

34.47061

34.47082

34.47050

Longitude (ddd, WGS84) W

120.29898

120.29829

120.29532

120.29452

120.29221

Length of Transect (m, from water to cliff)

27

27

29

46

24

Azimuth of Transect (magnetic, 0-360°)

124

138

170

167

170

Time (min) to Transect

3

2

4

2

5

Random Sample

Tarball sample weight (gm)

8

Tarball sample freshness 1-3 (1 fr, 3 old)

2

Latitude (ddd, WGS84) N

34.46975

Longitude (ddd, WGS84) W

120.29900

Figure 6. Examples of field worksheets on which were recorded field data at each beach for each month. Two separate teams worked on the northern and southern beaches during the program. The entire set of field data is provided in monthly spreadsheet files on compact disk.

RESULTS

Tarball mass and tarball number on beaches

Table 1 shows the diagnostic average parameters for each beach sampled during the program from July 2001 through August 2003. The north county beaches (Casmalia, Surf, Boathouse and Jalama) have a noticeably larger tarball mass and size as compared to south county beaches (Figure 7). This mass and size difference likely reflects oil and tar differences at their sources. The suspect seeps impacting the north county beaches likely have tarwhip-like sources producing large, viscous tar masses. Figure 2 shows a tar whip seep sample collected at the sea surface near Point Conception.

In contrast the southern beaches are more often impacted by proximal oil seeps that occur frequently from about offshore Gaviota east to beyond the study area into Santa Monica Bay. Here oil rises to the surface and undergoes weathering for some period, coalescing into smaller, more fragile tar patties seen in Figure 3, Areas near known oil seeps offshore Coal Oil Point impact Coal Oil Point with millions of small tar patties seen in Figure 4. Table 1 demonstrates that Coal Oil Point receives the greatest number of tarballs (about 930,000 at any one time, whereas the average size of one of these tarballs is about 1gm. In contrast, data from Jalama Beach has an estimated 4200 tarballs on the beach, however the average tarball mass is about 67 gm. In order to normalize the accumulation of tar mass on each beach, a simple calculation relating the weight of tar found each beach in relation to the beach length, where; Average Estimated Mass on Beach (gm) x Beach Length (km) = Average Tar Mass on Beach (gm/km) was made with the results shown in Table 1. More tar mass was found on the four northern beaches relative to the six southern beaches. Notable exceptions to this are Casmalia beach in the north where less tar accumulates and Arroyo Burro beach in the south where more tar accumulates. It is assumed the amount of tar mass accumulation on a beach is related to the distance and flux of a contributing seep or seeps. If this relationship is correct then we can speculate that Casmalia beach is further from the Point Conception seep sources.

Timing of tarball deposition on beaches

A goal of the sampling program was to document seasonal changes of tar deposition. Figures 9 and 10 show histograms of the average tar mass accumulation for both the northern and southern beaches. In general tarballs accumulate at a faster rate or remain longer on all beaches during the summer and fall months. The reasons for this are unclear based on our observations, however we speculate that factors such as prevailing winds and currents combined with more quiescent wave conditions favors the accumulation and preservation of tarballs on the beach during the summer and fall months. In contrast, winter storms remove beach sand and other materials.

Specifically for the northern beaches, Surf and Jalama beaches receive the largest quantity of tar residue peaking in summer and fall months, however from April through August, 2003, Surf beach had higher accumulation rates than Jalama beach, a trend in reverse of that established from July 2001 through March 2003.

The south county beach most impacted by tar deposition is Coal Oil Point (maximum estimated 88,000 gm/km beach in October 2002) followed Arroyo Burro (maximum estimated 20,000 gm/km beach in June 2003). Other beaches had spikes of increased tar deposition over just one sampling period during the 15 month sampling program for the the southern beaches. Single month tar deposition highs for these beaches are perhaps one-time events that focus tar on that beach, or perhaps reflect an increase in nearby-seep activity. Histograms showing the estimated mass of tar for each beach are presented in Appendix 4.

Table 1. Average values of tarballs parameters showing average values for each beach sampled by counting and weighing tarballs along 4 m wide, 3 to 5 beach perpendicular transects and scaled up to the length of each beach. Data from each beach represents measurements made on one day for each month sampled. Tarballs less than about 3mm in diameter were not counted, however the mass of these accumulations was estimated when noted by beach survey personnel. Calculated numbers are rounded to 2 significant figures.

Beach

Ave. No. Tarballs on Beach

Ave Est. Mass on Beach (gm)

Ave wt. Tarball (gm)

Ave Tar Mass on Beach (gm/km)

Times

Sampled

Casmalia

460

2800

11.0

2400

19

Surf

8700

400000

36.0

50000

17

Boathouse

110

3900

11.0

20000

18

Jalama

4200

220000

67.0

56000

19

Secate

5900

3000

3.3

1800

13

Gaviota

7900

3400

0.7

2100

14

Tajiguas

9400

1000

0.6

1000

14

Coal Oil Point

930000

88000

1.0

22000

14

Arroyo Burro

64000

63000

10.0

7900

13

Loon Point

3500

3600

2.0

1120

14

 

Figure 7. Average tarball weight for each beach measured over the program.

The northern beaches collect the largest and heaviest tarballs.


Figure 8. Average estimated number of tarball observed on each beach. Coal Oil Point beach receives the highest number of tarballs. The upper figure displays the same data on a logarithmic scale.

Figure 9. Composite histogram showing the estimated tar mass accumulation for six southern beaches during the sampling period. Tar masses are estimated as grams tar per kilometer of beach. Coal Oil Point receives the largest quantity of tar residue, and the tar tends to accumulate more in the summer and fall months.


Figure 10. Composite histogram showing the estimated tar mass accumulation for four northern beaches during the sampling period. Tar masses are estimated as grams tar per kilometer of beach. Surf and Jalama beaches receive the largest quantity of tar residue, and the tar tends to accumulate more in the summer and fall months.

Biomarker analyses of tarballs and oils

A comprehensive suite of biomarker analyses and whole-oil carbon isotopic composition analyses were completed on 182 selected tarballs and 34 oil samples. The oil samples represent three offshore oil producing areas and ROV or diver collected seep samples are included as reference materials for this project. Sixteen samples were collectively taken from four platforms offshore of Point Arguello, (Harvest, Hermosa, Hildalgo, and Irene) two from platform Holly, and two from Platform A. Each sample was treated as described in the method section. After analyses, biomarker parameters were calculated, often given as ratios of relative compound concentrations to another compound were compiled (Appendix1B).

The parameter ratios used in this study are described in Appendix 1A, and summarized according to multivariate groupings in Appendix 1B. The various parameter ratios were chosen to include as many as possible of the chemical families and constituents common to these tars because all the tars originate from a common Miocene Monterey source, and thus many of the differences between groups of tarballs are small. The ubiquitous triterpane, C30 -hopane, was used to normalize seven of the parameters, thus serving as a pseudo conserved-internal-standard (Wang et al., 1998). Ease of measurement was also a factor in choosing parameters in order for this study to be more broadly utilized.

A chemometric multivariate statistical approach was applied to the data to sort out the differences in the biomarker ratios, to test for correlations between and within the sample groups, and to attempt to relate the tar residues to possible local sources. We used an interactive statistical and data analysis software package called JMP Statistical Discovery, commercially available from the SAS Institute, Inc., Cary, N.C. (Brand names are used for identification only and do not imply endorsement by the U.S. Geological Survey). Applicable data from Appendix 1B were subjected to hierarchical cluster analyses (HCA) and principal component analyses (PCA). We chose an incremental hierarchical clustering technique. The data were first standardized by the variable mean and standard deviation. PCA is a tool to reduce the dimensionality of a set of data by depicting relationships among variables and to assess each variable’s contribution to the overall variance of the data. PCA uses a separate algorithm and is a visual validation of the clustering based on HCA. The statistical procedure involves a plot of the data in multi-dimensional space, followed by a standardized PCA.

The results of the multivariate statistical analyses are presented in Figures 11 and 12. Figure 11 shows a HCA for the data set. The best fit of the data results in a total of nine tar and oil families, of which, four groups have only seven samples (Groups 3, 5, 6, and 9). The remaining five groups comprise the bulk of the sample set. The largest group (1, red-orange) contains 114 samples comprised of tarballs from all ten beaches; however, the majority of the samples come from Surf, Boathouse, and Jalama beaches. Three seep samples, viscous tar whips, collected near Point Conception fall into this group. The correspondence of beach proximity to seep samples indicates that these and likely many more seeps in the Point Conception area contribute the majority of tarballs to these areas. Characteristic of this group are average biomarker ratio values with no one obvious compound or group of compounds that make the group distinctive. This group corresponds nicely to the largest group of tar samples found on the three largest Channel Islands beaches (Hostettler et al., in press) and suggested to be sourced in a fairly broad area in the Santa Barbara Channel.

A significant exception to these generalizations are noted for Loon Point beach, the most southern and eastern beach sampled, where each sample falls into group 1 (Appendix 4, last map). This result implies two disparate suppositions; 1, assuming the source for this tar group is centered around Point Conception, then the tarballs must be transported here by well directed ocean currents. Such a supposition seems unlikely given the lack of group 1 samples on nearby Arroyo Burro beach. A more likely rationale for the origin of these samples may be that a separate source(s) are to the east of Loon Point beach. Many possible sources are implied by the the number or possible seep sources within group 1 (Figure 11). Loon Point samples are spread throughout group 1 and thus into many smaller subgroups in group 1, if one source were suspected the samples would likely fall into 1 to 2 subgroups. Sampling beach tarballs on beaches east of Loon Point can test this hypothesis.

A relatively small group 2 (green) contains ten samples, five of which are seep or oil mousse samples collected in the Santa Barbara Channel. The two Platform A oil samples are within this group. Only four tarball samples are within the group and they are geographically confined to the most western three beaches of the six southern beaches (Gaviota, Secate, and Tajiguas). The relationship between where the oil mousse samples were taken at sea relative to the beaches suggest that similar oil seepages impact these beaches from seeps that are in the general vicinity of these beaches. This group is not widely distributed and does not correlate well with any Channel Island tars (Hostettler et al., in press). One anomaly in this group is the relatively high T/(T+M) (avg. 0.45) and the low Tm/Ts and BI (avg. 1.6 and 0.26 respectively) signaling higher thermal maturity associated with a deep-sourced oil. Finding Platform A oil samples in this group is not surprising, but the other oil mousse samples found in this group are somewhat unexpected perhaps indicating that the seep source(s) for this group come from deeper oil sources.

Group 4, orange, contains 18 samples, and includes those mainly collected on Casmalia beach and five from Surf beach and one from Boathouse beach. These three beaches are the most northerly of the beaches and are likely impacted by an unknown seep source that likely lies to the north of Casmalia beach. These samples are unique in that they have a high refractory index (RI). This group roughly corresponds to a small group of 3 samples found on San Miguel Island by Hostettler et al. (in press).

A relatively large group, group 8, labeled blue, is populated by tarballs predominantly found on the southern beaches, mainly Coal Oil Point and Arroyo Burro. Only 3 tarballs from this group were found on the northern beaches. This group also contains four seep samples, three of which were collected over actively bubbling seeps at sea. One of these seeps was offshore Gaviota, and the others located west of Coal Oil Point. Platform Holly oil samples are also included in this group. In August, 2004 civic-minded people reported unusual amounts of oil washing ashore on Arroyo Burro beach. Two of these samples (03-187 and 03-188) fall within this group. In an earlier memo submitted to the County, it was concluded that it was impossible to conclusively differentiate samples from natural seepage from Platform Holly oils. Platform Holly oils have biomarker parameters that are similar to seep oils.

All of the platform oil samples from offshore of Point Arguello, (Harvest, Hermosa, Hildalgo, and Irene) group into the last major group (group 7, yellow). Two oil slick samples collected off of Point Conception (offshore Cojo anchorage) correlate with this group. Only one beach tarball sample is found within this group, and it is unique in that it was collected on Surf beach right after a known oil spill, the Torch Oil Spill of September 8, 1997. The oil spill sample is correlated with three samples taken from platform Irene, where the oil was originally pumped. These samples show high thermal maturity (high T/T+M, low Tm/Ts, and BI) as would be expected from production oils which are pumped from deeper levels and have experienced more thermal maturation.

Four samples taken by divers at Shane seep, in the western part of the many seeps that occur near Coal Oil Point are assembled in two groups with little relation to any of the major tarball and oil groups. These samples are unique in that they were collected on the seafloor near active gas venting areas. These samples were oil-residue impregnated sand, likely representing biodegraded oil associated with the seep. Oil samples skimmed from the surface in a related study (samples 03-172, 173, and 174) are found in groups 1 and 8 highlighting the differences in oil likely coming from the seep to tar residue at the seep vent.

Figure 11. Preliminary hierarchical cluster diagram of all samples analyzed to date. In this version 17 source biomarker parameters are used. Biomarker parameters sensitive to thermal maturity of oil or those susceptible to biodegradation are not included.

Figure 12. Principal components analyses results for the study samples. Approximately 79% of the sample variability is incorporated in the three principal components and plotted. Color groups correspond to those in Figure 11.

PRELIMINARY CONCLUSIONS

This study provides new information on the geochemical nature of tarballs common to the the Santa Barbara County coastline and beaches:

All beached tarballs in this sample set share geochemical source characteristics typical of source rock in the Monterey Miocene formation. Differences in relative amounts of constituents reflecting different inputs, levels of thermal maturity, degrees of biodegradation, and probably slightly different depositional facies allow fingerprinting and correlation by chemometric analysis. The range of individual fingerprint parameters within tarball groups, however, is somewhat broad and correlations are not as tight as might be expected from, for example, a spill of a specific crude oil such as the Exxon Valdez. Apparently, seep oil and related shoreline tarballs, even if from the same source, have small local variations in constituent concentrations, giving broader ranges within the chemical signature.

A few generalizations can be made from the chemical fingerprint library. Three main tarball groups correspond to three main seepage areas; one likely north of Casmalia Beach, one centered around Pt. Conception, and the third area extending from about offshore Gaviota to offshore Coal Oil Point. Oil from Platforms Holly and "A" are grouped with natural seep samples and tarballs, thus oil coming directly from either of these platforms cannot be irrefutably distinguished from natural seep sources and their resultant tarballs using the methods employed in this study. In contrast oils from the four platforms offshore of Point Arguello are grouped together, and can be distinguished from natural tarballs. We found no tarball that could be traced to any of these platform oils in our study, however if there were a spill in the future this chemical fingerprint library would be useful in tracking deposition.

The timing of tar accumulation on the beaches occurs mainly during the summer and fall months. Beaches closest to known seep sources such as Coal Oil Point tend accumulate more tarball mass than those further from these sources. The average size of tarballs is much larger for the north county beaches that those of the south indicating a different tarball source area that corroborates geochemical signature of the tarballs.

ACKNOWLEGEDEMENTS

We gratefully acknowledge the help of the beach sampling parties who made monthly surveys of the beaches, recorded data and collected samples:

NORTH TEAM

Kris Broderick

Jamie Jones

Katie Hickling

SOUTH TEAM

Sara Benjamin

Christy Till

Karen Vasko

REFERENCES

Bence, A.E., Kvenvolden, K.A., Kennicutt II, M.C., 1996. Organic geochemistry applied to environmental assessments of Prince William Sound, Alaska, after the Exxon Valdez oil spill—a review. Organic Geochemistry 24, 7-42.

Curiale, J.A., Cameron, D., Davis, D.V., 1985. Biological marker distribution and significance in oils and rocks of the Monterey Formation, California. Geochimica et Cosmochimica Acta 49, 271-288.

Ensminger, A. van Dorsselaer, A., Spykerelle, C., Albrecht, P., Ourisson, G., 1974. Pentacyclic triterpanes of the hopane type as ubiquitous geochemical markers—origin and significance. In: Tissot, B. and Brenner, F., (Eds.), Advances in Organic Geochemistry 1973, Editions Technip, Paris, pp. 245-260.

Grantham, P.J., Wakefield, L.L., 1988. Variations in the sterane carbon number distribution of marine source derived crude oils through geologic time. Organic Geochemistry 12, 61-73.

Hartman, B.A., Hammond, D.E., 1981. The use of carbon and sulfur isotopes as correlation parameters for the source identification of beach tar in the southern California borderland. Geochimica et Cosmochimica Acta 45, 309-319.

Hickey, B.H., 1998. Coastal oceanography of western North America from the tip of Baja California to Vancouver Island, Coastal segment (8,E). In: Robinson, A.R. and Brink, K.H. (Eds.), The Sea, Volume 11, John Wiley & Sons, Inc., p. 345-393.

Hostettler, F.D., Rosenbauer, R.J., Kvenvolden, K.A., 1999. PAH refractory index as a source discriminant of hydrocarbon input from crude oil and coal in Prince William Sound, Alaska. Organic Geochemistry 30, 873-879.

Hostettler, F.D., Rosenbauer, R.J., Lorenson, T.D., Dougherty, J.A., in press, Geochemical characterization of tarballs on beaches along the California coast. Part I—Shallow seepage impacting the Santa Barbara Channel Islands, Santa Cruz, Santa Rosa and San Miguel.

Kaplan, I.R., Galperin, Y., Lu, S.-T., Lee, R.-P., 1997. Forensic environmental geochemistry: differentiation of fuel-types, their sources and release times. Organic Geochemistry 27, 289-317.

Kvenvolden, K.A., Hostettler, F.D., Carlson, P.R., Rapp, J.B., Threlkeld, C.N., Warden, A., 1995. Ubiquitous tar balls with a California-source signature on the shorelines of Prince William Sound, Alaska. Environmental Science and Technology 29, 2684-2694.

Kvenvolden, K.A., Rosenbauer, R.J., Hostettler, F.D., Lorenson, T.D., 2000. Application of organic geochemistry to coastal tar residues from Central California. International Geology Review 22, 1-14.

Kvenvolden, K.A., Hostettler, F.D., 2003. Coastal tar and natural oil seeps in California—A tribute to I.R. Kaplan. In press

Leifer, I., Luyendyk, B., Broderick, K., 2002. Tracking seep oil from seabed to sea surface and beyond at Coal Oil Point, California. Proceedings of the Coastal World Oceans 2002 Conference, Santa Barbara, CA, October 24-27, 2002.

Mackenzie, A.S., 1984. Applications of biological markers in petroleum geochemistry. In: Brooks, J. and Welte, D., (Eds.), Advances in Petroleum Geochemistry, v. 1, Academic Press, London, pp. 115-214.

Mackenzie, A.S., Patience, R.L., Maxwell, J.R., Vandenbroucke, M., and Durand, B., 1980. Molecular parameters of maturation in the Toarcian shales, Paris Basin, France—1. Changes in the configuration of acyclic isoprenoid alkanes, steranes, and triterpanes. Geochimica et Cosmochimica Acta 44, 1709-1721.

Moldowan, J.M., Lee, C.Y., Watt, D.S., Jeganathan, A., Slougui, N.-E., Gallegos, E.J., 1991. Analysis and occurrence of C26-steranes in petroleum and source rocks. Geochimica et Cosmochimica Acta 55, 1065-1081.

NAS, U.S. National Academy of Sciences (2002) Oil in the Sea III: Inputs, Fates, and Effects. National Academy Press, Washington, D.C.

Palacas, J.G., Anders, D.E., King, J.D., 1984. South Florida Basin—a prime example of carbonate source rocks of petroleum. In: Palacas, J.G., (Ed.), Petroleum Geochemistry and Source Rock Potential of Carbonate Rocks. American Association of Petroleum Geologists, Studies in Geology No. 18, pp. 71-96.

Peters, K.E., Moldowan, J.M., 1993. The Biomarker Guide. Prentice Hall, Englewood Cliffs, New Jersey, 363 pp.

Reed, W.E. and Kaplan, I.R., 1977, in State Lands Commission Staff Report, California offshore gas, oil, and tar seeps, 155-225.

State Lands Commission Staff Report, 1977. California offshore gas, oil, and tar seeps. 449 p.

Venkatesan, M.I., 1988. Occurrence and possible sources of perylene in marine systems—A review. Marine Chemistry 25, 1-27.

Wang, Z., Fingas, M., Blenkinsopp, S., Sergy, G., Landriault, M., Sigouin, L., Foght, J., Semple, K. Westlake, D.W.S., 1998. Comparison of oil composition changes due to biodegradation and physical weathering in different oils. Journal of Chromatography A 809, 89-107.

Waples, D.W., Machihara, T., 1991. Biomarkers for geologists—a practical guide to the application of steranes and triterpanes in petroleum geology. American Association of Petroleum Geologists, Methods in Exploration No. 9, 91 pp.

Winant, C.D., Alden, D.J., Dever, E.P., Edwards, K.A., Hendershott, M.C., 1999. Near surface trajectories off central and southern California. Journal of Geophysical Research 104, 15713-15726.

APPENDIX 1. Geochemical Parameters(1A) and Calculated biomarker data(1B)

[NOTE- Appendix 1B is not included here. See comments before Table of Contents.]

APPENDIX 1A. Geochemical Parameters

Whole oil:

1. d13C, the carbon isotopic composition of whole tar residues. These compositions are useful for oil and source rock correlations (Peters and Moldowan, 1993).

Saturate fraction:

2. Pr/Ph, pristane/phytane. This is a widely used source parameter (Peters and Moldowan, 1993) based on two of the most common isoprenoids in crude oils. These compounds are readily lost by degradation, and are not present in most of the tarballs in this study, although they are prominent in unweathered production or crude oils.

Triterpanes (hopanes), m/z 191 SIM chromatograms:

3. Tm/Ts, 17a-22,29,30-trisnorhopane/18a-22,29,30-trisnorneohopane. This ratio is used as both a source and maturity parameter (Seifert and Moldowan, 1978).

4. Triplet, [C26-tricyclic terpane (S?) + C26-tricyclic terpane (R?)/C24-tetracyclic terpane]. This source parameter was used to distinguish coastal tar residues in Prince William Sound (Kvenvolden et al., 1995).

5. 23T/C30, C23 tricyclic terpane/17a,21b(H)-hopane. This ratio is a source parameter adapted from Peters and Moldowan (1993).

6. 23Tri/24Tri, C23 tricyclic terpane/C24 tricyclic terpane. Source parameter.

7. C30/C29, 17a,21b(H)-hopane/17a,21b(H)-30-norhopane. This ratio was used by Palacas et al. (1984) as a source parameter.

8. C31S/(S+R), 17a,21b(H)-homohopane (22S)/17a,21b(H)-homohopane (22S+22R). This epimer ratio is a hopane maturity parameter used extensively in petroleum geochemistry; the equilibrium ratio at full maturity is ~0.6 (Ensminger et al., 1974; Mackenzie, 1984).

9. BI, Bisnorhopane Index, 28,30-bisnorhopane/17a,21b(H)-hopane. This source ratio has been used to characterize oils from the Monterey Formation. The presence of 28,30-bisnorhopane, in addition to indicating a marine, highly reducing depositional environment (Curiale et al., 1985), is reported to be passed on from bitumen rather than generated from kerogen and therefore decreases with thermal maturity (Peters and Moldowan, 1993). Therefore, this ratio would be higher in oils sourced from near-surface facies.

10. OI, Oleanane Index, 18a+b(H)-oleanane/17a,21b(H)-hopane. This commonly used source parameter indicates a contribution from Cretaceous and younger plant material (Peters and Moldowan, 1993). In the California coastal tars, oleanane is generally present, but in low amounts.

11. GI, Gammacerane Index, gammacerane/17a,21b(H)-hopane. This ratio is used as a source parameter; abundant gammacerane is a carbonate/evaporite facies indicator and a marker for highly reducing, hypersaline depositional environments (Peters and Moldowan, 1993).

12. T177/Hop, 25,28,30-trisnorhopane/17a,21b(H)-hopane. (Nor24/C30) 25,28,30-trisnorhopane is found in some oils from the Monterey Formation (Curiale et al., 1985) so this index is a possible source indicator.

Steranes, m/z 217 SIM chromatograms:

13. C29S/(S+R), 24-ethyl-5a,14a,17a(H)-cholestane (20S)/ 24-ethyl-5a,14a,17a(H)-cholestane (20S+20R). This sterane epimer ratio is commonly used as a maturity parameter; the equilibrium value at full maturity is ~0.5 (Mackenzie et al., 1980).

14. C28R/C29R, 24-methyl-5a,14a,17a(H)-cholestane (20R)/ 24-ethyl-5a,14a,17a(H)-cholestane (20R). This source parameter has been modified from discussions in Grantham and Wakefield (1988) and Waples and Machihara (1991).

15. Dominant sterane(s). This descriptor indicates the sterane(s) that are most prominent in the m/z 217 chromatogram. The m/z 217 chromatogram may also include a fragment of bisnorhopane (BN), which is noted if it is one of the most prominent peaks.

16. a27R/Hop, a Sterane Index, 5a,14a,17a(H)-cholestane/17a,21b(H)-hopane. This parameter gives an indication of relative proportions of a common regular sterane to hopane. In this study it helps track sterane biodegradation.

17. nor26&27/Hop, another Sterane Index, two tentatively identified steranes, C26 24-nor-5a-cholestane (Moldowan et al., 1991) and C27 27-nor-24-methyl-5a-cholestane (Schouten et al., 1994), indexed to hopane. This is a source parameter and may serve as a maturity parameter, particularly in subsequent studies when deeper production oils are considered.

18. a27R/nor27, 5a,14a,17a(H)-cholestane/C27 27-nor-24-methyl-5a-cholestane. A sterane parameter which also tracks sterane biodegradation.

Aromatic fraction

19. PAH-RI, Polycyclic Aromatic Hydrocarbon-Refractory Index. This index is a source parameter, the ratio of the second, usually major, peak in the highly refractory C26 to C28 triaromatic sterane suite (m/z 231) to that of the first, usually dominant, peak in the monomethyl chrysenes (m/z 242) (Hostettler et al., 1999).

20. T/(T+M). T = ∑triaromatic steranes (areas), C26 to C28, m/z 231; M = ∑monoaromatic steranes (areas), C26 to C28, m/z 253. Aromatic steroid parameter. This is a thermal maturity and source parameter, widely used, modified from that described in Peters and Moldowan (1993). Low values, reflecting relatively higher levels of the monoaromatic steroids, indicate low thermal maturity.

21. ∑C2D/∑C2P, ∑dimethyl dibenzothiophenes (m/z 212)/ ∑dimethyl phenanthrenes (m/z 206). Source parameter indicating relative levels of sulfur-containing PAH to regular PAH (Kaplan et al., 1997; Bence et al., 1996).

22. C3D/C3P, ∑trimethyl dibenzothiophenes (m/z 226)/ ∑trimethyl phenanthrenes (m/z 220). Source parameter as #21.

23. Pery/Chr, a PAH parameter, perylene normalized to chrysene. Perylene helps distinguish shallow-seeping oils from deeper oils. Perylene has a biogenic origin and is associated with near-surface bitumens (Ventakesan, 1988)

24. Tri28/Tri29, areas, C28 tricyclic terpanes (doublet)/C29 tricyclic terpanes (doublet). Source parameter.

25. abC31S/C30, 17a, 21b(H)-homohopane(22S)/17a, 21b(H)-hopane. Source parameter.

26. aaaC27R (∑27,28,29) 5a,14a,17a(H)-cholestane (20R) and 24-methyl-5a,14a/ (∑5a,14a,17a(H)-cholestane (20R) and 24-methyl-5a,14a,17a(H)-cholestane (20R) and 24-ethyl-5a,14a,17a(H)-cholestane (20R)). Source parameter.

27. aaaC28R (∑27,28,29) 5a,14a,17a(H)-cholestane (20R) and 24-ethyl-5a,14a,17a(H)-cholestane (20R)/∑5a,14a,17a(H)-cholestane (20R) and 24-methyl-5a,14a,17a(H)-cholestane (20R) and 24-ethyl-5a,14a,17a(H)-cholestane (20R)). Source parameter.

28. aaaC29R (∑27,28,29) , 24-ethyl-5a,14a,17a(H)-cholestane (20R ) / (∑5a,14a,17a(H)-cholestane (20R) and 24-methyl-5a,14a,17a(H)-cholestane (20R) and 24-ethyl-5a,14a,17a(H)-cholestane (20R)). Source parameter.

29. Sum C3BT/Sum C3N, ∑ C3 benzothiophenes (m/z 176)/∑ C3 naphthalenes (m/z 170). Source parameter.

APPENDIX 2. Maps of each sampled beach with tarball groups

APPENDIX 3. Summary of tarball collection and analyses

Table 1. Summary of tar collection and analyses during quarters 1 thru 4.

Beaches

Transects

Total Samples

Archived samples

Run samples

South

6

387

171

69

103

North

4

250

175

96

79

Totals

10

637

346

165

182

 

[NOTE – The remaining detailed data tables of Appendix 3 are omitted here due to file size limitations.]

APPENDIX 4. Tar mass histograms for each beach