CHAPTER 11. TEMPORAL AND SPATIAL PATTERNS FOR SURFZONE BACTERIA BEFORE AND AFTER DISINFECTION OF THE OCSD OUTFALL

Home

Chapter 1. Introduction and Background

Chapter 2. Methodology

Chapter 3. Surfzone Bacteria Patterns

Chapter 4. Subtidal Circulation Pathways

Chapter 5. Newport Canyon Transport Pathway

Chapter 6. Sea Breeze

Chapter 7. Tidal Transport Pathways

Chapter 8. Sediment Resuspension and Transport near the OCSD Outfall

Chapter 9. Nearshore Circulation and Transport Pathways

Chapter 10. Spatial and Temporal Patterns of Plume Tracers

Chapter 11. Temporal and Spatial Patterns for Surfzone Bacteria before and after Disinfection of the OCSD Outfall

Acknowledgements

More than 5 million people visit the beaches off Huntington Beach, California, each summer, helping to support a regional tourism industry of $80 million annually. However, large sections of these beaches were posted or closed for several months in the summer of 1999 and intermittently posted or closed in subsequent summers because bacteria levels in the surfzone exceeded beach sanitation standards in the California Health and Safety Code (Assembly Bill 411, or AB411) for extended periods. Because people stayed away from the beaches, local residents and tourists were seriously inconvenienced and the recreational and business communities in the region suffered serious economic losses. Local agencies conducted a wide variety of studies during Phase I (OCSD, 1999) and Phase II (Grant et al., 2000) of the Huntington Beach Shoreline Investigation to try to identify possible contamination pathways from upland sources, adjacent estuaries, and the coastal ocean. Unfortunately, these investigations could not determine specific sources for many of the contamination events. In particular, it was not clear whether coastal ocean processes occasionally brought bacteria-rich effluent from the Orange County Sanitation District’s (OCSD) ocean outfall to the beach.

In the summer of 2001, Phase III of the Huntington Beach Shoreline Investigation was instigated to study coastal-ocean processes that may allow the transport of OCSD’s outfall plume to the beach. Initial findings were reported to the OCSD board in May 2002. At that time, preliminary analysis indicated that the OCSD outfall was not responsible for the contamination events that caused beach closures. In April 2003, a full report on hypothetical coastal-ocean pathways that could transport a significant amount of effluent to the beach was presented to OCSD’s board and interested members of the public. Further analysis of this rich data set did not change the researchers’ initial conclusion that bacteria from the OCSD effluent plume were not responsible for beach closures. The data sets and analyses the scientists used to draw this conclusion are given in the first ten chapters of this Phase III final report.

Notwithstanding the above conclusion, OCSD chose not to renew its 301(h) waiver, and, further, began disinfecting their outfall in August 2002. In this chapter, we will examine the patterns in the extensive fecal indicator bacteria (FIB) data sets collected before and after disinfection of the outfall to determine whether the initial conclusion that the outfall was not the primary cause of beach closures is supported by the subsequent data sets. We will document the magnitude of reduction in FIB levels in the outfall itself and in the plume near the diffuser site (Figure 11-1). We will also look at the frequency characteristics of the FIB for the surfzone bacterial data collected at sites along the beach since 1998 (Figure 11-1). This data set includes the recent FIB measurements taken at the standard OCSD surfzone sites from stations 39N to 39S prior to disinfection of the outfall (see Chapter 3 for a discussion of the full data set) and the year of measurements that have been collected since that analysis. These additional data sets will help assess the usefulness of effluent disinfection. The information on the temporal and spatial patterns for contamination events may also suggest potential sources for the contamination that could be investigated in future studies and possibly identify other locations that could be subject to similar contamination events.

11.2. Pre- and Post-disinfection FIB Patterns at the Outfall Site

It took about two months to fully implement the procedures to disinfect the outfall. The process was finally stabilized in mid-October 2002, when the post-disinfection bacteria count measured in the last-stage bay before the effluent was piped to the ocean decreased dramatically (Figure 11-2). The median total coliform density was reduced by four orders of magnitude, from ~107 to ~103 MPN/100 mL. Reductions of similar magnitude were found in fecal coliform and enterococci. One can estimate a very conservative bacteria concentration in the coastal ocean near the diffuser by dividing the measured FIB density at the last-stage bay by 100 because the designed dilution rate at the diffuser is actually 1:200. It can be seen clearly in Figure 11-3 that the estimated FIB densities at the diffuser are usually smaller than those measured for surfzone bacteria densities at station 6N, which is located on the central portion of Huntington Beach. After October 2002, the bacterial levels near the diffuser were too low to cause contamination events at the beach, even if the probable further dilution of bacteria as they are transported in the coastal ocean is not taken into account.

The actual densities of bacteria at the outfall site before and after disinfection of the final effluent were measured during several cruises. These measurements confirm the above estimate that, after disinfection, much lower densities of bacteria are being released into the coastal ocean (Figure 11-4). Before disinfection, the measured levels of bacteria at OCSD’s outfall site at the normal depths the outfall plume rises to in the coastal ocean (centered on 30-45 m) (Jones, 2004) were above AB411 standards. After disinfection, the bacterial levels were reduced considerably. Even at the core of the effluent plume, the box plots show that most bacteria samples were below AB411 standards for the periods between August 2002 and June 2003. Additional FIB measurements from a 1-km grid of stations around the outfall confirm that very small densities of bacteria are being discharged from the disinfected outfall (OCSD, 2004).

Yet, measurements of bacteria in the surfzone during this same period of time show that contamination events remained a problem (Figure 11-4). Surfzone bacteria densities still occasionally exceeded AB411 standards, especially for enterococci. Postings and closures continued to occur at Huntington and the adjacent beaches.

11.3. Frequency-Domain Characteristics and Distribution Patterns for Surfzone FIB

The surfzone FIB are usually measured five days per week, with one of the measurement days occurring on a weekend. Here, we use Lomb periodograms (Lomb, 1976) to characterize the frequency patterns in FIB time series at the 18 surfzone sampling stations because Lomb periodograms calculate the dominant frequencies in an unequally spaced data set. Lomb periodograms (hereafter periodograms) essentially use the least-squares method to fit a set of frequencies to an unevenly spaced time series. The set of frequencies fit to the data set do not necessarily have to be a complete set of basis functions or satisfy the Raleigh criteria. Here, we normalized the periodograms by the maximum value in each periodogram. Hence, the amplitudes for spectral peaks can only be compared within individual periodograms. The amplitude of a spectral peak must exceed the 95% confidence limits in order to distinguish it from random noise. The further the amplitude of a spectral peak exceeds the confidence limit, the more confident we are that the fit is not the result of random noise. The confidence limits for the normalized periodograms were calculated according to Scargle (1982), using the modified normalization proposed by Horne and Baliunas (1986).

Periodograms of the log of the bacterial counts were computed for three time periods:

1) Five years of pre- and post-disinfection total coliform, fecal coliform, and enterococci data from March 1998-December 2003;

2) One year of pre-disinfection total coliform, fecal coliform, and enterococci data from September 2001-September 2002; and

3) One year of post-disinfection total coliform, fecal coliform, and enterococci data from October 2002-November 2003.

11.3.1. Patterns Using FIB Data from 1998-2003

The most prominent feature in the periodograms for the recent five years of total coliform data is the annual cycle (Figure 11-5a p.1, 11-5a p.2). The annual peaks overwhelm other spectral peaks at most of the surfzone sampling sites. The most dominant annual peaks are found near the Santa Ana River (Station 0) and at stations south of the river. It is interesting to note that the relative amplitude of the annual cycle in total coliform decreases sharply at stations 3N-15N and increases again at stations further to the north. This is the only region where the fortnightly cycle, which is related to the spring-neap cycle of tides, dominates the spectrum. The dominance of the fortnightly cycle is particularly strong between stations 3N and 12N, a region where contamination events on Huntington Beach have recently been centered.

The frequency patterns in the periodograms of the recent five-year fecal coliform data reflect those for total coliform data (Figure 11-5b p.1, 11-5b p.2). The patterns for the pre- and post-disinfection of fecal coliform also reflect those for total coliform. The figures are included in this chapter for completeness, but individual patterns for fecal coliform will not be discussed further.

The spatial patterns in the recent five-year enterococci data set are markedly different from those for total coliform and fecal coliform data (Figure 11-5c p.1, 11-5c p.2). The fortnightly cycle, instead of being roughly confined to a small region between stations 3N and 12N, is present at most surfzone sampling stations. It is particularly strong between stations 3N and 21N, a spatial region that encompasses the fortnightly cycle seen for total coliform, and in stations located at and south of the mouth of Newport Harbor (stations 29S, 39S). The fortnightly cycle is relatively weaker, but significant at other southern stations that bracket Newport Pier and the head of Newport Canyon. The only stations that do not have significant fortnightly cycles in enterococci are located near the mouth and south of the Santa Ana River.

The annual cycle in enterococci is dominant primarily near the mouth of the Santa Ana River and at stations to the south, similar to that seen in total coliform. However, near the mouth of Newport Harbor, the annual cycle is overwhelmed by the fortnightly cycle. North of the river, the annual cycle for enterococci is dominant only at the most northern sampling station, 39N.

11.3.2. Patterns Using Pre- and Post-disinfection FIB Data

Periodograms for total coliform and enterococci were calculated from approximately year-long data sets taken immediately prior to (Figures 11-6a p.1, 11-6a p.2, 11-6b p.1, 11-6b p.2, 11-6c p.1, 11-6c p.2) and after (Figures 11-7a p.1, 11-7a p.2, 11-7b p.1, 11-7b p.2, 11-7c p.1, 11-7c p.2) disinfection of the outfall. Because of the short time series, these periodograms could not resolve the annual cycle. However, we can still determine whether the spatial patterns in the fortnightly cycle change after disinfection.

The spatial patterns in the fortnightly cycle for the pre- and post-disinfection total coliform data sets are similar to those found for the entire five-year data set. The only significant spectral peaks are found north of the Santa Ana River. The most significant fortnightly cycles are located between stations 3N and 9N, though the range extends north of Huntington Pier (station 21N) in the year before disinfection of the outfall began. After disinfection, the fortnightly cycle is confined to a slightly smaller region between stations 3N and 12N. It is not clear if this change will be a consistent feature found in future data sets because the data sets are short. We have only one year of data collected after disinfection.

The fortnightly cycle in enterococci before and after disinfection is significant at most northern and southern stations, except for the stations near and south of the mouth of the Santa Ana River. This is similar to the spatial pattern in the entire five-year data set. Again, the significance levels of the spectral peaks are slightly higher for the pre-disinfection data, but most peaks exceed the 95% confidence level. The fortnightly cycle accounts for relatively more of the variability in each data set in the region between stations 3N and 15N, near the mouth of Newport Harbor and the stations surrounding Newport Pier for both the pre- and post-disinfection periods. This last set of stations also lies inshore of the head of Newport Canyon. The broad spatial range for enterococci is consistent with the spatial patterns found for the entire set of pre-disinfection enterococci data described in Chapter 3 (pattern type 2).

11.4. Discussion

The annual cycle in the FIB data sets is caused by local runoff from storms in the wet winter season. It is well known that runoff contaminates beaches (Reeves, et al., 2004); this entire region tends to exceed AB411 standards after large rains. The annual cycle dominates the variability in all three FIB data sets for most stations at or south of the Santa Ana River. It is also the dominant frequency in coliform bacteria for stations north of station 15N. These data suggest that local runoff is the dominant contamination source for most of the southern and some of the northern beaches in this region. The strong annual cycle at the most northern site (station 39N) for all types of bacteria, coupled with the tendency for fresher water to be found close to shore at the most northern hydrographic sites in summer (Chapter 10), does suggest that runoff from local rivers may also be important at some sites in all seasons. Possibly contaminated water from the San Gabriel or Los Angeles Rivers may enter the northern Huntington Beach region in summer.

The contamination most commonly associated with summer postings and closings on Huntington Beach occurs at the fortnightly, or spring/neap, tidal cycle (MEC, 2000; Grant et al., 2000) (Chapter 3). These data support that result; sites between stations 3N and 15N on Huntington Beach have significant fortnightly cycles in all three FIB data sets. They are most likely to be posted or closed during spring tides. Given that these sites have relatively weak annual cycles, local runoff is not the primary cause of contamination in this limited region of the beach.

A significant fortnightly cycle in enterococci, but not for either species of coliform bacteria, is found at many other surfzone stations. It appears that an enterococci contamination source tied to the fortnightly cycle is present at many more stations than previously thought. It is disturbing to note that a relatively strong enterococci fortnightly cycle is found near and south of the entrance to Newport Harbor (stations 29S, 39S). The fortnightly cycle overwhelms the annual cycle for enterococci. This suggests not only that local runoff is not primarily responsible for the enterococci contamination, but that there may be different Newport Harbor sources for the two species of bacteria. Fortunately, the enterococci contamination levels have not yet been high enough to cause significant posting and closures at these beaches.

It is probable that the source of the fortnightly cycle in contamination events is related to the wetting and draining of the land during the large tidal excursions found during spring tides (Grant et al., 2001; Kim et al., 2004). During spring tides, seawater can: 1) wet previously dry areas of the Talbert Marsh or the banks of the Santa Ana River, 2) enhance the exchange between the flood channels that enter the Talbert Marsh and the coastal ocean, 3) enhance the exchange with Santa Ana River estuary water, and 4) reach higher on the local beaches, causing seawater and any dissolved material to drain back into the coastal ocean. These pathways deserve more investigations in future studies at Huntington Beach and at the other sites with significant fortnightly cycles in enterococci.

The extensive data set collected during the Phase III study showed that the fortnightly contamination events on Huntington Beach were not associated with the transport of bacteria from OCSD’s outfall plume by internal waves during spring tides, as one early hypothesis stated (Boehm et al., 2002). These data support the original conclusion of the Phase III study as well as previous studies (OCSD, 1999; Sea Grant, 2000). Bacteria from the OCSD outfall were not the primary cause of contamination of the local beaches. Post-disinfection measurements of FIB levels at the outfall site are below levels presently found on the beach and well below AB411 standards. There is no significant difference in the spatial patterns for contamination from any species of FIB in the pre- and post-disinfection time periods. Yet, the beaches are still being posted and/or closed, despite disinfection of the outfall.

It is unclear whether low concentrations of bacteria from the outfall plume were transported to the beach before disinfection occurred. The data suggest that the relative spectral levels at the fortnightly period are a bit lower after disinfecting the effluent. However, given the changing management policies for FIB sources in the region, we expect that the beaches would be less contaminated even if the effluent plume were not disinfected. Several local storm drains that empty contaminated water onto the beach during the dry summer season have been and are continuing to be found. Water from these storm drains is being diverted from the beach into the sewage system in the critical summer months.

Pursuing a more precise quantification of possible low-level contributions of FIB to the surfzone from the outfall plume before it was disinfected is not as important as trying to find the continuing sources of the summer contamination events. Given the differing spatial and temporal patterns for the coliform bacteria and enterococci discussed above and in Chapter 3 of this report, it is possible that the enterococci contamination has a different source from that of the coliform not only at Newport Harbor, but at many beaches in the study area. The very similar spatial and temporal patterns found for total coliform and fecal coliform bacteria suggest that most coliform bacteria share a similar source.

Unfortunately, further investigations into the issue of different sources for bacteria based on differing spatial patterns for coliform and enterococci bacteria is complicated by our lack of knowledge of how different bacteria species survive in seawater. Recent work suggests that the ecology of bacteria is very complex and changes with temperature, sunlight, salinity, and possibly other environmental factors (Hurst et al., 2002; Noble et al., 2004). An even more complex task is to develop fast and inexpensive methods to determine if the high FIB concentrations on local beaches have human or animal sources. More information on these issues is necessary before one can come to definitive conclusions based on FIB spatial patterns. Hourly samples of FIB concentrations in the surfzone at selected stations during both spring and neap tidal cycles may increase our understanding of this complex process.

11.5. References

Boehm, A.B., B.F. Sanders, and C.D. Winant, 2002. Cross-shelf transport at Huntington Beach: Implications for the fate of sewage discharged through an offshore ocean outfall. Environmental Science and Technology, v. 36, p. 1899-1906.

Chamberlin, C.E. and R. Mitchell, 1993. A decay model for enteric bacteria in natural waters. Water Pollution Microbiology, John Wiley & Sons.

Grant, S.B. et al., 2000. Huntington Beach Water Quality Investigation, Phase II: An analysis of ocean, surfzone, watershed, sediment, and groundwater data collected from June 1998 through September 2000. Final Report.

Grant, S.B. et al., 2001. Generation of Enterococci bacteria in a coastal saltwater marsh and its impact on surf zone water quality. Environmental Science and Technology, v. 35, p. 2407-2416.

Horne, J.H. and S.L. Baliunas, 1986. A prescription for period analysis of unevenly sampled time series. Astrophysical Journal, v. 302, p. 757-763.

Hurst, C.J., R.L. Crawford, M.J. McInerney, G.R. Knudsen, and L.D. Stetzenbach, eds., 2002. Manual of environmental microbiology, 2nd ed. ASM Press, Washington, D.C.

Jones, B., 2004. Huntington Beach Shoreline Contamination Investigation, Phase III, v. II. In press.

Kim, J.H. and S.B. Grant, 2004. Public misnotification of coastal water quality: A probabilistic evaluation of posting errors at Huntington Beach, California. Environmental Science and Technology, in press, 8 p.

Kim, J.H. et al., 2004. Locating sources of surf zone pollution: a mass budget analysis of fecal indicator bacteria at Huntington State Beach, California. Environmental Science and Technology, submitted.

Leecaster, M.K. and S.B. Weisberg, 2001. Effect of sampling frequency on shoreline microbiology assessments. Marine Pollution Bulletin, v. 42, p. 1150-1154.

Lomb, N.R., 1976. Least-squares frequency analysis of unequally spaced data. Astrophys. Space Sci., v. 39, p. 447-462.

MEC, 2000. Huntington Beach closure: Relationships between high counts of bacteria on Huntington Beach and potential sources. Final Report.

Noble, R.T., I.M. Lee, and K.C. Schiff, 2004. Inactivation of indicator bacteria from various sources of fecal contamination in seawater and freshwater. Journal of Applied Microbiology, in press.

OCSD, 1999. Huntington Beach Closure Investigation. Phase I. December 1999, Fountain Valley, CA.

–––, 2004. Receiving water microbiological data: Pre- and post-disinfection. Preliminary Report, May 2004, Fountain Valley, CA.

Reeves, R. l., S.B. Grant, R.D. Morse, C.M. Copil, Oancea, B.F. Sanders, and A.B. Boehm, 2004. Scaling and management of fecal indicator bacteria in runoff from a coastal urban watershed in Southern California. Environmental Science and Technology, in press, 12 p.

Scargle, J.D., 1982. Studies in astronomical time series analysis II: Statistical aspects of spectral analysis of unevenly spaced data. Astrophysical Journal, v. 263, p. 835-853.

Sea Grant, 2000. Huntington Beach closure investigation: technical review. Wrigley Institute for Environmental Studies, University of Southern California, Los Angeles, CA.