MARINE
                                            ENVIRONMENTAL
                                              RESEARCH
             Marine Environmental Research 61 (2006) 471–493
                                    www.elsevier.com/locate/marenvrev




       Temporal and spatial variability of fecal
        indicator bacteria in the surf zone off
           Huntington Beach, CA
       L.K. Rosenfeld a,*, C.D. McGee b, G.L. Robertson b,
             M.A. Noble c, B.H. Jones d
   a
     Department of Oceanography, Naval Postgraduate School, Code OC/Ro, 833 Dyer Road, Room 328,
                     Monterey, CA 93943-5122, USA
           b
            Orange County Sanitation District, Fountain Valley, CA 92728, USA
               c
                US Geological Survey, Menlo Park, CA 94025, USA
    d
     Department of Biological Sciences, University of Southern California, Los Angeles, CA 90089, USA

     Received 24 April 2005; received in revised form 16 February 2006; accepted 19 February 2006




Abstract

  Fecal indicator bacteria concentrations measured in the surf zone off Huntington Beach, CA from
July 1998–December 2001 were analyzed with respect to their spatial patterns along 23 km of beach,
and temporal variability on time scales from hourly to fortnightly. The majority of samples had bac-
terial concentrations less than, or equal to, the minimum detection limit, but a small percentage
exceeded the California recreational water standards. Areas where coliform bacteria exceeded stan-
dards were more prevalent north of the Santa Ana River, whereas enterococci exceedances covered a
broad area both north and south of the river. Higher concentrations of bacteria were associated with
spring tides. No temporal correspondence was found between these bacterial events and either the
timing of cold water pulses near shore due to internal tides, or the presence of southerly swell in
the surface wave field. All three fecal indicator bacteria exhibited a diel cycle, but enterococci
rebounded to high nighttime values almost as soon as the sun went down, whereas coliform levels
were highest near the nighttime low tide, which was also the lower low tide.
Ó 2006 Elsevier Ltd. All rights reserved.

Keywords: Sewage effluent; Surf zone; Diurnal; Internal waves; Fecal indicator bacteria; Tides


*
   Corresponding author. Tel.: +1 831 656 3253; fax: +1 831 656 2712.
   E-mail address: lkrosenf@nps.edu (L.K. Rosenfeld).

0141-1136/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.marenvres.2006.02.003
472        L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493

1. Introduction

  Fecal indicator bacteria (FIB) are found in the feces of humans and other animals.
Although some strains are ubiquitous and not related to fecal pollution, their presence
in water is used as an indication of fecal pollution and the possible presence of enteric
pathogens. Epidemiological studies have demonstrated that FIB not only determine the
extent of fecal contamination in recreational surface waters, but their density in recrea-
tional water samples has been shown to have a predictive relationship with swimming-
associated gastroenteritis at marine and fresh water bathing beaches (Cabelli et al.,
1982; Haile et al., 1999; Flesisher et al., 1993).
  Under the standards contained in California Health and Safety Code §115880 (Assem-
bly Bill 411, Statutes of 1997, Chapter 765; [AB411]), from April to October beaches must
be posted when FIB concentration exceeds a single sample, or running monthly geometric
mean, standard, or closed to public contact if the health authority thinks that human sew-
age is the cause of the high bacteria level. There were numerous beach postings and clo-
sures in the summers of 1999 and 2000, due to elevated FIB levels in the surf zone off
Huntington Beach, CA (see Fig. 1). Since this beach is a heavily used recreational area,
postings and closures have a potentially significant economic impact on the local commu-
nity, especially during the summer.




Fig. 1. Map showing Huntington Beach, and the surf zone stations from 39N to 39S, plus the locations of the
OCSD outfall, Santa Ana R., Talbert Marsh, and the AES outfall/intake. The 18 regularly sampled stations are
covered by the shaded rectangle.
        L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493  473

  The Orange County Sanitation District (OCSD) is the third largest southern California
publicly owned treatment works. Between 1998 and 2001, OCSD discharged non-disinfec-
ted effluent through an ocean outfall with a 1.8 km long diffuser extending from 6.4 to
8.2 km offshore from Huntington Beach (Fig. 1). In 2000–2001, the average daily dis-
charge was 244 million gallons per day (MGD) (9.2 · 108 L/d) of treated wastewater
(OCSD, 2001).
  During 15 years of monitoring, required by their National Pollution Discharge Elimina-
tion System (NPDES) ocean-discharge permit, OCSD never found evidence that their
wastewater plume reached the surf zone. In spite of this, it was suggested that bacteria-rich
effluent from the OCSD outfall plume might be responsible for the beach contamination
(Grant et al., 2000), so OCSD undertook an extensive investigation of this hypothesis.
The Huntington Beach Phase III (HB PIII) study took place from May to October 2001
and included moored instrumentation that provided extensive high-frequency measure-
ments of currents and water properties. Since examination of historical monitoring data
(MEC, 2000; Grant et al., 2000) suggested that the elevated bacteria levels coincided with
periods of maximum tidal range (i.e., spring tides), during six 48-h periods near the time of
spring tides, the standard surf zone bacterial sampling program was enhanced with higher
temporal resolution sampling, and hydrographic cruises were performed to help evaluate
cross-shelf transport processes and their potential to move wastewater effluent to the beach.
  The hypothesis tested in HB PIII was whether internal tides transported FIB from the
OCSD outfall plume to the beach. The study concluded that there was no evidence to sup-
port that mechanism (Noble et al., 2003a), but the extensive data set collected during HB
PIII, in addition to the long-term OCSD monitoring data set, afforded an unparalleled
opportunity to examine aspects of the spatial and temporal variability of the FIB concen-
trations and AB411 single sample exceedances in the Huntington Beach surf zone on time
scales from weeks to hours. This examination shed light on potential sources of these FIB.
Particular attention was paid to the relationships between FIB concentrations and the
fortnightly lunar cycle, the diurnal sunlight cycle, and the semidiurnal tidal cycle.
  This paper does not address the impact of rainfall and run-off, which are primarily win-
tertime phenomena in southern California, since that has been the focus of a number of
other studies (Boehm et al., 2002a; Noble et al., 2003b). It should be noted however, that
99% of the annual FIB loading from this watershed occurs during winter storm events
(Reeves et al., 2004).

2. Methods

2.1. Determination of bacteria concentration

  The FIB concentrations and AB411 single sample exceedances in the Huntington Beach
surf zone between July 1, 1998 and December 31, 2001 were considered. The three types of
FIB sampled, and their allowed single sample maxima, expressed as organisms/100 ml,
under AB411 are: (1) fecal coliform (FC) 6 400, or alternately Escherichia coli, EC; (2)
total coliform (TC) 6 10,000, or 1000 if FC/TC P 0.1, and (3) enterococci (ENT) 6 104.
  Samples were collected from 17 mandatory locations, specified by OCSD’s NPDES
permit, plus one additional station, 12N, in ankle-depth water (see Fig. 1). The regularly
sampled stations are named as 1000s of feet north or south of the Santa Ana River (SAR).
The array of stations extends from 39,000 ft (11.9 km) north of the SAR (39N) to 39,000 ft
474      L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493

south of the SAR (39S). Most of the locations are separated by 3000 (914.4 m) or 6000 ft
(1828.8 m), and all but one, 29S, are at integer multiples of 3000 ft from the SAR. Samples
were sometimes taken at additional stations to enhance spatial resolution to better address
perceived localized problem sites (e.g., station 8N) or from specific geographic or man-
made features of interest not within the surf zone, such as the mouth of Talbert Marsh
(TM) and SAR or a nearby power plant (AES). Samples were generally collected between
05:00 and 10:30 (local time) with the typical collection strategy proceeding from north to
south.
  During the late springs and summers (late May–August) of 1998–2001, and continuing
through the end of 2001, samples were collected five days/week, including one weekend
day. During November 1999–March 2000, samples were collected two days per week; dur-
ing parts of September–October 1999 samples were collected seven days per week, and
during the remainder of the 3.5 year time period, samples were generally collected three
days per week. The sampling schedule described above was mandated by the NPDES per-
mit. During the late spring and summer of 2001, six 48-h, ship-based surveys (May 21–22,
June 19–21, July 5–7, July 19–21, August 19–21, and September 15–17) were conducted,
coinciding with spring tides. Surf zone microbiology samples were collected hourly during
these six surveys at a contiguous 11-station subset of the standard OCSD sampling sites
(15S–21N). During the first cruise, hourly sampling was done for only 36 h, with one sev-
eral-hour break.
  Sample collection, preservation and handling procedures are described in detail in
OCSD (2001). OCSD used the multiple tube fermentation method for TC and FC bacteria
at all permit-required surf zone stations (Standard Method [SM] 9221B, E in Eaton et al.,
1995). An alternative method for simultaneous detection of TC and EC, a chromogenic
substrate coliform test commonly known as Colilert-18Ò, was used by OCSD for the addi-
tional samples taken for the HB PIII study (SM 9223B; Eaton et al., 1995). In most envi-
ronments, EC represents the majority of the FC population, but there is no absolute
relationship that can be predicted. FC concentrations were estimated as 1.1 times the
EC concentration for the purposes of this study. This ratio was established based on a sur-
vey of FC/EC concentrations in samples collected from OCSD’s final effluent. EPA 1600,
a membrane filter test method for ENT in water, is the standard methodology used for
OCSD’s daily permit compliance testing. Membrane filter methods (SM 9222A and
9222D in Eaton et al., 1995) were also used to determine TC and FC concentrations in
some HB PIII samples analyzed by the Orange County Health Care Agency’s laboratory
during the hourly sampling events. OCSD also used an alternative chromogenic substrate
for ENT, commonly known as EnterolertÒ (Idexx Corp.) for samples collected during the
hourly sampling events. Minimum and maximum detection limits varied depending on the
method used and sample dilution, as described in Rosenfeld (2004).

2.2. Definition of bacterial contamination events

  Several types of bacterial ‘‘events’’ were defined in order to reduce the three FIB con-
centrations at 18 stations into a few variables. The events take advantage of spatial pat-
terns appearing naturally in the data, that show that coliform contamination events
were more localized and ENT events more widespread (see Fig. 2). Taken together, the
three event types capture all but a handful of days on which there was an AB411 excee-
dance anywhere in the HB PIII data set.
          L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493          475




Fig. 2. The vertical bars indicate days on which bacterial events occurred, blue, type 1; magenta, type 2; gray,
types 1 and 2; gold, type 3. Sea level measured at Los Angeles is shown at the bottom of the figure. A black dot is
plotted at each time and location a sample was taken. Colored symbols are plotted at the time and location of
samples exceeding AB411 single sample standards. The size of the symbol is proportional to the ratio of the
measured bacterial concentration to the AB411 criteria. The symbol size for a sample equal to the AB411 criteria
is shown to the right of the indicator species in the legend.




  When TC or FC concentrations exceeded AB411 standards at one or more of the sta-
tions between, and including, 3N and 12N during any day (00:00–23:59 PST), a type 1
event was recorded for that day. During a day when ENT concentrations exceeded
AB411 standards at three or more of the numbered stations from 39S to 39N, including
at least one station from 3N to 12N, a type 2 event was recorded for the day. A third type
of event was added to ensure that all days on which any of the single sample standards
were exceeded in the Huntington Beach area would be included in at least one type of
event. A type 3 event was defined as occurring on any day during which ENT exceeded
AB411 standards at any station between, and including, 3N and 12N, on which there
was not a type 1 or type 2 event.
  Since hourly round-the-clock sampling was done during the six intensive surveys of the
HB PIII study, it is possible that greater sampling frequency and/or nighttime sampling
led to more frequent observation of bacterial contamination events. Hence, a data set con-
taining no more than one sample per station per day was created. If multiple samples were
available for a station for a given day, the sample closest in time to that of the average
476       L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493

sampling time (calculated over May 1–October 31, 2001) for that station was used. The
type 1–3 event analyses described above were also carried out on this daily sub-sampled
data set.

2.3. Sea level and astronomical data

  Hourly sea level data for Los Angeles (33°43.2 0 N, 118°16.3 0 W, station ID 9410660) for
1998–2001 were obtained through the National Ocean Service web site (http://co-ops.nos.-
noaa.gov/). Sea levels are given in meters above mean lower low water. While technically,
the term spring tide denotes the largest tidal range in a fortnightly period, in order to
assign a time unambiguously to each spring tide, the highest high water in a window of
10–20 days (9.4 days in one case) after the previous spring tide, with a neap tide in
between, was picked as the date/time and height of ‘‘spring tide’’. Note that by using
the actual sea level, rather than the phase of the moon, there is some variance in the time
between ‘‘spring tides’’ due to meteorological effects, although on average it is 14.8 days as
would be expected.
  Times of new and full moons, and sunset and sunrise, were obtained from the US Naval
Observatory web site (http://mach.usno.navy.mil). Averaged over 1998–2001, the spring
tide high water occurs 0.6 days after a new or full moon.

2.4. Event probability versus sea level

  The relationship between the height of the higher high water each day and the proba-
bility of a bacterial event occurring was considered using a simple logistic regression
model. If the probability of an event on day i is denoted by pi, then the logistic model says:
   logðpi =ð1 À pi ÞÞ ¼ b0 þ b1 tidei ;                         ð1Þ
where tidei is the height of the higher high tide on day i, and b0 and b1 are the logistic
regression coefficients.

2.5. Moored data

  An extensive moored instrument array, as well as shipboard surveys, produced a wealth
of data as part of the HB PIII study (see Fig. 1). Only a small subset of those data are used
in this paper and described briefly here. The instrumentation and methodology is
described in much more detail in Hamilton et al. (2004).
  Temperature, salinity, and current records from seven moorings in water depths from
10 to 205 m, extending in a line offshore from a point between 3N and 6N, were analyzed
for the presence of internal tidal pulses. The temperature measured 0.5 m above the bot-
tom in 10 m water depth at mooring M01, located $1 km offshore, was used to indicate
the near-bottom, near-shore temperature. These time-series, with temporal resolution of
2–5 min, are available from mid-June to mid-October of 2001.
  Surface waves were measured using accelerometers and rate gyros in conjunction with a
magneto-inductive compass, on the M07 surface buoy, located 8 km offshore on the 60 m
isobath, just shoreward of the shelf break (see Fig. 1). Daily directional wave spectra, with
directional resolution of 1°, and frequency resolution of 1/256 Hz, were computed from
the earth-referenced north–south, east–west and vertical displacement time series using
        L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493  477

the maximum entropy method (MEM) described by Lygre and Krogstad (1986). Details of
the data processing can be found in Hamilton et al. (2004).

2.6. Identification of internal tidal pulses

  For the purposes of this study, internal tidal pulses were identified by the appearance
of cold water near shore, as determined by two methods. In the first, the temperature of
the sewage outfall plume was determined by looking at the temperature/salinity distribu-
tion from the moorings in two-week increments from June 17 to October 15, 2001. The
maximum temperature of the plume T/S anomaly was picked as an upper limit of the
temperature of the plume and this value was compared with the temperature time series
at the near-shore moorings (M01 and M03). If the 10 m temperature at mooring M01
was less than, or within 0.25 °C of, the warmest temperature in the sewage outfall
plume, it was designated as a cold water near-shore event. The results were found to
be insensitive to whether mooring M03 or M01 was used. In the second method,
designed to identify the times when water from the shelf break was brought closest to
the beach, the temperatures measured by the moorings along the main line were used
to determine when the 12 °C isotherm was inshore of the 30 m isobath and the water
at M03 was cooler than 13 °C.

3. Results

3.1. Three and a half year record: July 1, 1998–December 31, 2001

3.1.1. Statistics
  There were 14,866 samples from the 18 regularly sampled surf zone stations (from
39S to 39N) collected between July 1, 1998 and December 31, 2001 and analyzed for
the three bacterial indicators. The majority of samples had bacterial concentrations less
than, or equal to, the minimum detection limit (58% for TC, 73% for FC, 66% for
ENT). Only 219 (1.5%) had TC concentrations (or a combination of concentration
and FC/TC ratio) exceeding the AB411 single sample standard, 284 (1.9%) exceeded
the FC standard, and 834 (5.6%) had ENT concentrations greater than the AB411 stan-
dard. There were 217 samples that exceeded AB411 standards for both ENT and coli-
form bacteria.
  While many more samples exceeded the ENT standards than exceeded the coliform
standards, when the exceedances were grouped into daily events, as described in Section
2.2, the number of days affected by coliform events slightly exceeded the number affected
by ENT events. Referring to events calculated using the daily sub-sampled data set, it was
found that out of the 1280 days, of which 692 were sampled, type 1 events occurred on 148
days, type 2 events on 67 days, and type 3 events on 75 days. Without sub-sampling to no
more than one sample per day per station, an additional six days qualified as type 1 events
and an additional 10 days as type 2 events.

3.1.2. Fortnightly variability
  Analysis of the relationship between tidal height (or range) and the occurrence of bac-
terial contamination in the surf zone found that there was a higher likelihood of a type 1 or
2 bacterial event occurring the day after the spring tide high water than on any other day
478       L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493

in the fortnightly cycle. Of all the days in a fortnightly cycle, the day that spring tide
occurred had the greatest chance (60%) of experiencing an AB411 exceedance (i.e., type
1, 2, or 3) in the 3N–12N region. Classification of all bacterial events according to what
day they fell on relative to the spring tide showed that about 50% of them occurred within
±2 d of the spring tide.
  The logistic model for each of event types 1 and 2, plus one in which the response was
type 2 or type 3 (see Fig. 3), all indicated that the tide height term was ‘‘statistically sig-
nificant,’’ meaning that the size of the effect we see in the model is such that it is very unli-
kely an effect of that size would have arisen in the sample if, in fact, there were no such
relationship in the population. The results show that the higher the height of the high tide,
or correspondingly the greater the tidal range, the more likely there is to be a bacterial con-
tamination event.

3.2. HB PIII: June–October 2001

3.2.1. Surf zone bacteria spatial variability
  All three indicator species showed higher concentrations preferentially in a band
between 3N and 12N (see Fig. 4). This was particularly evident for FC. Occasionally, rel-
atively high levels of ENT (>300 MPN/100 ml) were found almost simultaneously at sta-
tions all the way from 9S to 15N. Total coliform rarely exceeded 300 MPN/100 ml south
of the SAR, but on one occasion (in mid-August) TC exceeded 200 MPN/100 ml from 15S
to 12N.

3.2.2. Surface waves
  The daily wave spectra were examined by eye and characterized as southerly swell, wes-
terly wind waves, southerly wind waves, a combination of the above, or no waves (see
Fig. 5). Fifty-five percent of days with a bacterial event coincided with southerly swell,
while 45% of days with a bacterial event occurred on days with either no waves to speak
of, or westerly wind waves alone. These are nearly identical to the percentages of all days
in the study period with (54%) and without (46%) southerly swell, thus indicating that bac-
teria events did not occur preferentially during southerly swell. Considering each of the
three event types separately, type 1 (58%) and type 3 (50%) events occur in concert with
southerly swell in close to the same proportions as southerly swell occurs in general, while
type 2 bacterial events have a slight bias towards southerly swell, with 75% of them occur-
ring on southerly swell days.

3.2.3. Internal tidal pulses
  The dates on which internal tidal pulses occurred, according to the criteria described
in Section 2.6, are shown in Fig. 5. Subsequent complex demodulation analysis (Noble
et al., 2004a) confirm that all but the June 28–29 cooling event determined by these
methods, were in fact associated with the presence of an enhanced internal tidal signal
at the shelf break. These large cold water sloshing events were also identified by the
14 °C isotherm reaching 5 m, or shallower, at another mooring (SIO2) on the 11 m iso-
bath. Judged by all of these criteria, the largest internal tide pulses of the HB PIII study
occurred during July 23–26. Note the general lack of correspondence between the days
on which cold water was found near shore and the days on which bacterial exceedance
events occurred.
          L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493            479




                 1.0
                                          Type 1



                 0.8
           Prob. of Event
                0.6
             0.4   0.2
                 1.0 0.0
                 0.8




                                          Type 2
             Prob. of Event
                  0.6
              0.4
           0.2    1.0 0.0




                                          Type 2 or 3
                   0.8
           Prob. of Event
                0.6
            0.4   0.2
                 0.0




                        1.2   1.4   1.6    1.8   2.0   2.2
                      height of daily higher high tide (m, relative to MLLW)

Fig. 3. The solid lines show the predicted probability of bacterial events, as a function of tide height. The open
circles plotted at 1 or 0 on the y axis indicate event or no event, respectively, versus height of higher high water on
the x axis. The two dotted lines give estimated 95% confidence levels for the probabilities at each point.

3.3. Hourly round-the-clock sampling during 48-h surveys

3.3.1. Spatial variability
  Contour plots of the FIB concentrations during July 19–21 (see Fig. 6) are shown as a
representative example of the hourly surf zone bacteria data collected during six 48-h
480        L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493




Fig. 4. Log10 of total coliform (A), fecal coliform (B), and enterococci (C) concentration are plotted versus time
and distance alongshore. The data set subsampled to no more than daily values was used. X’s indicate the day and
location of each sample. Data is contoured with PlotPlus on a 180 · 13 grid (time and distance, respectively) using
cay = 5 and nrng = 2. The cay value determines the interpolation scheme. Cay = 0 means Laplacian interpolation
is used. As cay is increased, spline interpolation predominates over Laplacian. For pure spline interpolation
cay = infinity. Grid points are set to ‘‘undefined’’, and not used, if farther than nrng away from the nearest data
point.


periods coinciding with spring tides. Similar figures for other periods may be found in
Rosenfeld (2004). While there are similarities in the spatial patterns for all three FIB,
there are also differences. The higher concentrations tended to be found north of the
          L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493            481




Fig. 5. Vertical bars indicate days on which bacterial events occurred, blue, type 1; magenta, type 2; gray, types 1
and 2; gold, type 3. A black dot is plotted at each time and location a sample was taken. Cruise periods are shown
at the top of the figure. A characterization of the directional wave spectra is shown below that ( westerly wind
waves only,     all include southerly swell), with the occurrence of cold water near shore indicated below that
( 12 °C inshore of 30 m,    M01 T < max plume T,     both; the largest internal tide pulses, July 23–26, are
circled). Black dots in row 4 indicate days on which bacterial samples were taken. The day’s higher high water as
measured at Los Angeles is shown at the bottom.


SAR, but when there were elevated levels south of the SAR, they occurred to some
degree in all three indicators (see for example 2000–2300 PDT July 19 and 2100–2200
PDT July 20, Fig. 6). The highest values of TC tended to occur at station 0 (next to
the mouth of the SAR), and there was some suggestion of upcoast (northwestward)
propagation from there, with a speed of about 30 cm sÀ1 (see Fig. 6a), with elevated val-
ues subsequently found in the 0–12N region. FC values generally peaked close to station
6N (see Fig. 6b), while ENT values tended to be high not only near 6N, but also at the
southern (see Fig. 6c), and occasionally at the northern (not seen in this example), ends
of the range. ENT had a minimum from 6S to 0. These patterns also held for the daily
sub-sampled data shown in Fig. 4.

3.3.2. Diel and tidal variability
  The hourly data reveal a day–night cycle in all three FIB, with the highest values occur-
ring at night. The onset of elevated TC values at station 0 occurred between about mid-
night and 03:00 (see Fig. 6a), just prior to the nighttime low tide which occurred between
03:00 and 06:00 over the course of the study. Bacterial concentrations were not generally
elevated near the daytime low tide (the higher low tide of the two). On the few occasions
when there were high values of coliform in the middle of the day (not seen in this example),
ENT values were only slightly elevated, if at all, while at night all three indicators showed
notably higher concentrations. Note that the mid-day increases in bacteria level had their
maxima at station 0 for all three indicators, and occurred close to the time of the higher
low tide.
482        L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493




Fig. 6. Log10 of total coliform (A), fecal coliform (B), and enterococci (C) concentration (MPN/100 ml) are
plotted versus time and distance alongshore. Black xs indicate the time and location of each sample. Data is
contoured on a 48 by 13 grid (time and distance, respectively) using cay = 5 and nrng = 2. Time is Pacific daylight
time (PDT).




  The ENT exceedances, wherever they occurred, did so overwhelmingly between sun-
set and sunrise (see Fig. 7a). Note that the onset of ENT exceedances preceded both
the high tide (the start of the ebb tidal flow) and the sudden drop in near-bottom
                                  L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493              483

near-shore temperature, as seen on the nights of 19 and 20 July (see Fig. 7a). The coli-
form exceedances, however, did not show this strong relationship to the day–night
cycle, nor did such a pattern appear when a cut-off even lower than the AB411 single
sample standard was used for visualization (see Fig. 7b). The predominance of high




                              60                                    2.3 m
                                                                   21.5 o C

                                                               A
                              50
               Enterococci / 104 MPN/100 ml




                                                                   M1 T 0.5 mab Sea level
                                                                               21N
                              40
                                                                               15N,S
                                                                               12N
                                                                               9N,S
                              30
                                                                               6N,S
                                                                               3N,S
                                                                               0
                              20


                              10

                                                                   12 o C
                               0                                    -0.45 m
                              Noon              Noon               Noon
                              7/19              7/20               7/21

                                                                   2.3 m
  Total coliform / 1000 MPN/100 ml - solid symbols




                               8
  Fecal coliform / 400 MPN/100 ml - open symbols




                                                                   21.5 o C

                                                               B
                               7
                                                                   Sea level




                               6

                               5
                                                                   M1 T 0.5 mab




                               4

                               3

                               2

                               1
                                                                  12 o C
                                                                  -0.45 m
                               0
                              Noon              Noon               Noon
                              7/19              7/20               7/21

Fig. 7. The ratio of ENT concentrations (A), and TC and FC concentrations (solid and open symbols,
respectively, in B) to the AB411 single sample standards* at stations 15S–21N for the 4th intensive sampling
period are plotted together with sampling times at station 0 (pink dots), Los Angeles sea level (red line), and
temperature 0.5 m above bottom at M01 (blue line). Only ENT samples with ratios exceeding 1, and coliform
samples with ratios exceeding 0.5 are shown here. A local time reference is used; the period between sunset and
sunrise is shaded gray. *Note that not all samples with TC >1000 MPN/100 ml exceed AB411 standards, as the
TC/FC ratio may not be <10.
484                 L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493

coliform values near the time of low tide is evident however. The temporal patterns are
similar for the other five intensive sampling periods not shown in Fig. 7.
  Figs. 8–10 show the concentration of all three FIB at all stations 15S–21N for all
hourly data from the six intensive sampling periods plotted as a function of time of
day, sea level, and time since the previous higher high water, respectively. ENT values
rebounded as soon as the sun set, while coliform values did not rise until later in the
evening, and did not reach values as high as seen in the early morning (see Fig. 8).
Coliform values were highest (with a few exceptions) when sea level was low, whereas
ENT had no obvious relationship to sea level (see Fig. 9). The TC peak that was seen
5–10 h after higher high water (HHW) at station 0 also occurred north of the SAR
(represented by station 6N here), while the weaker peak 15–20 h after HHW only
appeared at station 0 (see Fig. 10). Fecal coliform showed a pronounced peak at sta-
tion 6N, but not at station 0, 6–10 h after HHW. The ENT pattern was quite different
from total and fecal coliform.




                   3000
       Conc. (MPN/ 100ml)




                                               Total Coliform
                   2500
                   2000
                   1500
                   1000
                   500
                    0
                      12 13 14 15 16 17 18 19 20 21 22 23 0 1 2 3 4 5 6 7 8 9 10 11 12
                                   Hour of the day
                   3000
       Conc. (MPN/ 100ml)




                                               Fecal Coliform
                   2500
                   2000
                   1500
                   1000
                   500
                    0
                      12 13 14 15 16 17 18 19 20 21 22 23 0 1 2 3 4 5 6 7 8 9 10 11 12
                                   Hour of the day
                   1000
        Conc. (MPN/ 100ml)




                                                Enterococci
                   800

                   600

                   400

                   200

                    0
                      12 13 14 15 16 17 18 19 20 21 22 23 0 1 2 3 4 5 6 7 8 9 10 11 12
                                   Hour of the day

Fig. 8. Concentrations of fecal indicator bacteria at all stations 15S–21N for all hourly data from the six cruise
periods are plotted versus time of day using a 24-h clock. Three total coliform, one fecal coliform, and nine
enterococci samples had concentrations greater than the maximum values plotted in this figure.
                  L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493     485

                  3000


       Conc. (MPN/ 100ml)
                                           Total Coliform
                  2500
                  2000
                  1500
                  1000
                  500
                   0
                   -0.5     0     0.5     1     1.5     2      2.5
                             Sea Level (m, relative to MLLW)
                  3000
       Conc. (MPN/ 100ml)




                                           Fecal Coliform
                  2500
                  2000
                  1500
                  1000
                  500
                   0
                   -0.5     0     0.5     1     1.5     2      2.5
                             Sea Level (m, relative to MLLW)
                  1000
       Conc. (MPN/ 100ml)




                                             Enterococci
                  800

                  600

                  400

                  200

                   0
                   -0.5    0     0.5     1     1.5     2      2.5
                             Sea Level (m, relative to MLLW)

Fig. 9. The same surf zone bacterial data as shown in figure 8 is plotted versus sea level measured in Los Angeles.
Values measured at station 6N are shown as red xs, those measured at station 0 as blue +s, all other stations are
black os.



4. Discussion

4.1. Statistics and sampling bias

  As has been previously noted by Noble et al. (2003b), Grant et al. (2000) and others, we
found that ENT exceeded AB411 single sample standards in many more samples than
total and fecal coliform did. However, the analyses presented demonstrate that, due to
the fact that samples from many stations have ENT exceedances on the same day, the
number of beach postings or closures in this area are as likely to be due to high coliform
levels as to high ENT levels.
  Since the measurements of FIB concentrations are not continuous in time, it is impor-
tant to consider how representative these data are, given the sample schedule used to
obtain them. Using five years of daily sampling at 24 Los Angeles area sites, Leecaster
and Weisberg (2001) estimated that 70% of AB411 single sample exceedances last only
one day, and less than 10% last more than three days. Assuming similar temporal variabil-
486                  L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493

                   3000


        Conc. (MPN/ 100ml)
                                              Total Coliform
                   2500
                   2000
                   1500
                   1000
                   500
                    0
                      0     5      10      15      20       25
                                  Hours Since Prev. HHW
                   3000
       Conc. (MPN/ 100ml)




                                              Fecal Coliform
                   2500
                   2000
                   1500
                   1000
                   500
                    0
                      0     5      10      15      20       25
                                  Hours Since Prev. HHW
        Conc. (MPN/ 100ml)




                   1000
                                              Enterococci
                   800

                   600

                   400

                   200

                    00       5      10      15      20       25
                                  Hours Since Prev. HHW

Fig. 10. Concentrations of fecal indicator bacteria at stations 0 (blue +s) and 6N (red xs) for all hourly data from
the six cruise periods are plotted versus time since the previous higher high water as measured at Los Angeles.
Two total coliform, no fecal coliform and three enterococci samples had concentrations greater than the
maximum values plotted in the figure.



ity, the five times per week sampling OCSD did during the summers and the last half of
2001 may have missed about 20% of the total and fecal coliform exceedances. Based on
two weeks of hourly sampling at four stations during May 2000 and 12 h of sampling
every 10-min at six stations (3N–12N) during September 2001, Boehm et al. (2002a) esti-
mated that more than 70% of single sample exceedances last less than 1 h, so the likelihood
is that many more days on which exceedances exist, at least fleetingly, are also missed. This
means the five times/week summertime sampling and the $12 days worth of hourly sam-
pling analyzed here should be representative of the spatial and temporal patterns of sum-
mertime variability with time scales greater than 2 h.

4.2. Spring–neap cycle and internal tidal pulses

  The incidence of high concentrations of bacteria confirm a previously identified rela-
tionship with the lunar cycle (MEC, 2000; Grant et al., 2000). The larger the tidal range
(or equivalently, the lower the daily lower low water, or the higher the daily higher high
         L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493      487

water), the higher the probability of a contamination event (see Fig. 3). Our results (Sec-
tion 3.1) indicate that bacterial contamination events were most likely to occur on the day
of the largest tidal range and that 50% of the events occurred within ±2 days of the spring
tide. Based on summer 1999 Huntington Beach surf zone samples, Kim and Grant (2004)
found that the error rate for beach postings would have been substantially reduced if they
had been based on ENT concentrations predicted from daily tidal range, rather than on
measured concentrations. This is because the correlation between ENT concentration
and daily tidal range often exceeded that between the concentration in the current sample
and the one preceding it; and the time between the collection of a sample and posting a
beach based on that sample’s concentration usually equals or exceeds the time interval
between sample collections.
  While it is clear that the surf zone was most contaminated near the time of spring tides,
the reasons for this are not as clear. The tidal variations in sea level are primarily due to
the barotropic, or depth-independent (neglecting bottom friction) component of the tidal
signal, also known as the surface tide. In stratified water however, the tidal period iso-
therm displacements and currents may be much more heavily influenced by the depth-
dependent, or baroclinic, component of the tidal signal, known as the internal tide. At
the latitude of Huntington Beach, freely propagating internal waves can exist at the semi-
diurnal, but not the diurnal, period. A detailed analysis, description, and discussion of the
internal tide off HB can be found in Noble and Hamilton (2004) and is the focus of a paper
in preparation by Noble et al.
  As has been found in many other coastal areas, Boehm et al. (2002b) identified inter-
mittent pulses of cold water near shore, and attributed them in part to cross-shore advec-
tion by internal tides. They could not rule out the possibility that this phenomenon, which
we will refer to as internal tidal pulses, could transport FIB in high concentrations from
the OCSD effluent plume into the surf zone, the hypothesis that had been proposed by
Grant et al. (2000). If the internal tide were phase-locked to the surface tide (i.e., the larg-
est internal tides occurred during spring tides), and if the internal tidal pulses transported
high concentrations of FIB to the surf zone, then this mechanism would help explain the
fortnightly variability observed in the FIB concentrations. However, the largest internal
tidal pulses during HB PIII occurred during July 23–26, not coincident with a spring tide
(the previous spring tide occurred on July 20), and the only exceedance of the AB411 sin-
gle sample standards in the Huntington Beach area on those days was a single sample
taken at 3N on July 23 with a measured ENT concentration of 110 MPN/100 ml. The tim-
ing of the cold pulses and the surf zone sampling at 3N–12N (see Table 1) was such that
the samples could have captured high bacterial concentrations if they were brought in to
shore together with the cold water.


Table 1
The timing of the cold surges, defined as the 13° isotherm being shoreward of the 20 m isobath, are shown
together with Huntington Beach surf zone sample times
Date             Cold surge time (PST)            Stn 3N–12N sample times (PST)
7/23/01           02:00–07:00                 07:25–07:55
7/24/01           01:00–09:00                 06:35–07:10
7/25/01           03:00–06:00                 06:30–07:05
7/26/01           02:00–07:00                 06:30–07:05
488       L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493

  We also considered whether the fact that the sampling times for each station vary very
little over the 3.5 year record, in combination with the change in time of the low tides each
day, could result in the effects of the timing of the sample relative to low tide being mis-
interpreted as spring tide effects. If the tidal cycle were due only to the largest constituent,
M2 with a period of 12.42 h; and the spring tides were exactly 14.77 d apart (which they are
not based on our determination of the largest sea level ranges from measured tidal
heights), and the sample was taken at exactly the same time each day, then the sample time
relative to low tide would be the same on all spring tide days. For example, if the samples
at station 9N were always collected an hour after low tide on days when the spring tide
occurred, then it would be impossible to tell if the preferential occurrence of high bacterial
concentrations on those days was due to the large tidal range, or the timing of the sample
relative to the stage of the tide. This would be particularly problematic for the coliform
bacteria data which, based on the hourly sampling, show a relationship to the stage of
the tide.
  We examined the timing of when the bacteria samples were taken at station 9N relative
to the time of the previous low water at Los Angeles, and found that on summertime
spring tide days, samples were always taken between 0 and 6 h after the previous low water
(i.e., generally on the rising tide). During the winter, spring tide sampling occasionally
occurred more than 6 h after the previous low water; but sampling frequency was less
in the winter than the summer. The fact that the number of hours after low tide that
the samples were taken on spring tide days are relatively evenly distributed between 0
and 6 h, and that on many non-spring tide days samples were also taken 0–6 h after
low tide, lends credence to the hypothesis that the increased bacterial concentrations at
spring tide are due to larger tidal ranges, rather than the timing of the sample collection
relative to the phase of the tide. However, with the given sampling scheme, it is not pos-
sible to unequivocally separate out those two effects. To get an even distribution of sam-
ples on falling versus rising tides on spring tide days, one would have to collect half the
samples in the evening, which would then introduce another set of issues concerning the
diel cycle.
  As other potential tidally related mechanisms are ruled out, increasingly it appears that
the larger land area flushed by seawater during the extreme high tides may be responsible
for the fortnightly variability. This mechanism would increase FIB contributions from
beach areas where dog feces, for instance, might be a significant source as was found in
the San Diego area (MEC, 2003), as well as from the TM where bird droppings have been
implicated as a source of ENT (Grant et al., 2001), and from the SAR (Kim et al., 2004).

4.3. Spatial patterns and sources; tidal current and surface wave direction

  Increased levels of all three FIB were found more frequently in the surf zone north of
the SAR than south of it (Figs. 2 and 4; and Grant et al., 2000), though high concentra-
tions of ENT were found throughout the range on a number of occasions. Excessive levels
of total and fecal coliform were generally confined to the 0–15N and 3N–9N regions,
respectively. There was some indication that a local source of bacterial contamination,
particularly high in FC, existed near 6N.
  The highest values of TC were found preferentially near low tide at station 0 (next to
the mouth of the SAR) and nearby stations to the north of it. Hourly TC data also
indicate that there was upcoast (northwestward) propagation from station 0, with a
        L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493  489

speed of about 30 cm sÀ1 (Fig. 6). Based on measurements made just outside the surf
zone during May 2000, Kim et al. (2004) state that the alongshore currents during flood
tides are directed upcoast, which is consistent with results from the HB PIII studies
(Noble, personal communication), and hypothesize that they distribute the FIB that
entered the coastal ocean from the SAR and TM on the previous ebb tide preferentially
to the north of the SAR. A mass budget analysis for July 5–7, 2001 which included
measurements at the outlets of TM and SAR, in addition to the surf zone data used
here, concluded that SAR contributed more TC than did TM (Kim et al., 2004). It
should be noted, however, that the subtidal currents often exceed the tidal currents,
so that there may be no change in the sign of the alongshore flow over the course of
several days (Hamilton, 2004).
  Earlier work by Grant et al. (2001), using May 2000 data, identified TM as a net source
of ENT. Confirming results presented here, recent modeling of mass and momentum flux
in the surf zone are consistent with TM and SAR being the primary source for the TC, but
not the FC and ENT, found in the Huntington Beach region (Grant et al., 2005).
  Kim et al. (2004) have suggested that the predominance of high bacterial concentrations
north of the SAR may be due in part to northwestward alongshore currents in the surf
zone resulting from southerly swell. Grant et al. (2005, 2001) found that dye introduced
into TM on two ebb tides in May 2000 subsequently moved upcoast or downcoast in
the surf zone at $30 cm sÀ1, depending on whether surface waves were out of the south
or west, respectively. Boehm et al. (2002a) noted propagation of an ENT pulse from
6N to 9N, also at 30 cm sÀ1, in 10-min data collected on September 15, 2001, also during
southerly wave conditions. The results presented here indicate that bacterial contamina-
tion events in the Huntington Beach region show no preference for days with southerly
swell versus days without them. Southerly swell was present on some occasions when
upcoast (northwestward) propagation of high TC concentrations from the SAR was
observed, and not on others. For instance, there were only westerly wind waves during
July 19–20 when upcoast propagation of high TC concentrations was observed (see
Fig. 6).

4.4. Diurnal and semidiurnal variability

  Consistent with conclusions drawn by Grant et al. (2000) and the results of Boehm et al.
(2002a) for data from May 2000, all three FIB showed a day–night cycle, with the highest
values occurring at night (see Figs. 6–8). The hourly data revealed no relationship between
ENT concentration and sea level (see Fig. 9). That, and the fact that the rapid increase in
ENT concentration preceded both high water and the sudden drop in near-bottom near-
shore temperature (see Fig. 7a), suggests that the day–night cycle in ENT is more strongly
influenced by sunlight-induced die-off than by tidal influence on either a landward or sea-
ward source, since the former would be expected to commence with the start of the ebb
tide and the latter to be coincident with the cooler temperatures which have an offshore
source. In contrast, high coliform values occurred predominantly near low tide (see Figs.
7b and 9). The high TC levels found near station 0 near the end of the larger nighttime ebb
tide were also found at stations to the north (see Fig. 10), with evidence of 30 cm/sÀ1
propagation speeds (see Fig. 6); suggesting that they may enter the coastal ocean from
a landward source and then be advected alongshore. The smaller TC peak at station 0
associated with the daytime ebb tide did not affect station 6N, perhaps indicating die-off
490      L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493

or dilution to background levels before the source waters were advected past station 6N.
There was a peak in FC concentration at 6N, but not at station 0, associated with the
nighttime ebb tide only (see Fig. 10).
  Noble et al. (2003b) proposed that ENT have a higher rate of failing the AB411 stan-
dards because, as suggested by earlier studies, ENT survives longer in seawater than TC
or FC. However, recent work by the same author (Noble et al., 2004b), using sewage as
an inoculant, and environmental conditions similar to those found off HB, demonstrates
that while ENT die off more slowly than FC under dark conditions, ENT are more sus-
ceptible than FC to sunlight inactivation, which is consistent with our results (Fig. 8)
and with Boehm et al. (2002a) who noted that ENT falls below detection limits earlier
in the day than total and fecal coliform. Sinton (2004) found that the relative inactiva-
tion rates of ENT and FC are different for raw sewage versus the effluent from waste
stabilization ponds, with ENT having higher survival rates than FC over the course
of one day in the former, while the reverse was true for the latter. The T90 for ENT
in temperature and sunlight conditions common in southern California coastal waters
was only 8–10 h (Noble et al., 2004b), indicating that these bacteria are very unlikely
to last from one day to the next. It has also been suggested (Chamberlain and Mitchell,
1978) that photochemically produced oxidants, such as peroxide, may play a role in the
daytime die-off of FIB. Boehm et al. (2002a) measured increasing concentrations of per-
oxide over the course of a day in a sunlight exposed tank filled with water from the
Huntington Beach surf zone.
  Though the results presented in Section 3.3 indicate that TC levels were controlled
more by the phase of the semidiurnal tidal cycle, and ENT more by the diurnal light
cycle, the fact that all the hourly sampling in this study was separated by about two
weeks means that we cannot definitively separate the two effects since the time of low
tide varied so little among the six intensive sampling periods, and the larger ebb tide
always fell at night. Grant et al. (2001), using May 2000 data, characterized the temporal
variability of ENT concentration in terms of flood versus ebb tides, noting that TM was
a source of ENT to the surf zone during ebb tides, but they did not look at the influence
of time of day.
  The temporal and spatial variability in FIB concentrations is due not just to temporal
and spatial variability in their sources, and advection and dispersion once they enter the
ocean, but also to the fact that bacteria do not behave in a conservative fashion. Also, dif-
ferent types of bacteria die off at different rates, and those rates vary as a function of envi-
ronmental factors such as temperature, salinity, nutrient concentration, predation, and the
presence or absence of bacterial toxins, in addition to solar radiation. Furthermore, coag-
ulation, flocculation, adsorption on particles, and sinking, all impact the ultimate fate of
the bacteria. The survival rate of FIB under the prevailing environmental conditions is not
well enough known for us to be able to say with any certainty how the temporal patterns
were influenced by survival rate.
  It is possible that the effects of sunlight in combination with the timing of the flood/ebb
cycle could contribute to the predominance of high bacteria levels near summertime spring
tides. In a mixed, predominantly semidiurnal tidal regime, such as is found off southern
California, the nearly twice daily tidal cycles are of unequal value. This is known as the
diurnal inequality of the semidiurnal tide. The larger of the two ebb tides (higher high
water to lower low water) falls at night during the summertime spring tides, with the
higher high tide at 8–10 p.m. local time. During the summertime neap tides, the larger
         L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493  491

of the two daily ebb tides is smaller than during spring tides, obviously, but much of it also
falls during daylight hours.

5. Conclusions

  The TC, FC and ENT concentrations measured in the surf zone off Huntington Beach
between July 1, 1998 and December 31, 2001, including 5-day per week summertime sam-
pling, and 12 days of hourly sampling during 2001 summertime spring tides, were analyzed
with respect to their spatial and temporal variability on time scales from fortnightly to
hourly. Most samples had FIB concentrations below detection limits; only a small percent-
age exceeded AB411 single sample standards.
  Areas where coliform bacteria exceeded AB411 standards were generally localized in
the region within 15,000 ft (4.6 km) to the north of the SAR. The data do not support
the hypothesis (Kim et al., 2004) that surface wave-driven alongshore flow is
responsible for this distribution, as these events did not occur preferentially during times
of southerly swell. The ENT exceedances covered a broader area both to the north and
south of SAR and tended to show up at many surf zone stations on the same day. Hence,
even though more ENT samples than coliform samples exceeded AB411 standards, Hun-
tington Beach closures were as likely to be due to coliform as to ENT.
  Higher bacteria concentrations were associated with maximum tidal ranges, i.e., spring
tides. Fifty percent of the FIB exceedance events (defined as either coliform or ENT excee-
dances local to the Huntington Beach area, or widespread ENT exceedances) occurred
within ±2 days of the spring tide. The data predict that on any given spring tide day there
is a 60% chance of an AB411 single sample standard being exceeded in the Huntington
Beach surf zone. A companion paper (Noble et al., accepted) found that the fortnightly
variability continued through 2003, even after OCSD fully implemented disinfection of
its effluent in October 2002. In southern California, the largest spring tides fall in the sum-
mer (May–August) and winter (December–February). The timing of the largest spring
tides relative to the annual cycle does not change year to year. Hence, the probability
of beach contamination will remain high in the summer for contaminants related to the
spring–neap cycle. Large internal tidal pulses were not coincident with spring tides, and
were not associated with high bacterial concentrations in the surf zone.
  All three FIB exhibited a day–night cycle, with the highest values occurring at night.
This pattern was very pronounced for ENT, which rebounded to high nighttime values
almost as soon as the sun went down. High coliform counts appeared preferentially at
low sea level heights, particularly after the larger ebb tide. Since the larger ebb tide always
occurred at night during the six hourly sampling periods, separated by two weeks, we can-
not definitively separate the effect of the diurnal inequality in the semidiurnal tide from the
day–night cycle. However, we suggest that total and fecal coliform levels are controlled
principally by the phase of the tide, whereas ENT levels are controlled more by the
day–night cycle.
  The widespread nature of the high levels of ENT, combined with the fact that it dies off
very quickly during the day (Noble et al., 2004b), would seem to indicate there may be
multiple sources. Boehm et al. (2002a) suggested that the most likely answer to the ques-
tion ‘‘Why is the surf zone so rapidly re-supplied with indicator bacteria after the sun goes
down?’’ is that there is a continuous supply of indicator bacteria to the surf zone. These
data would support that suggestion for ENT.
492        L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493

Acknowledgements

 The Orange County Sanitation District provided the funding for this study. We
acknowledge the following people from NPS for their assistance: Fred Bahr and Todd
Anderson for data analysis and visualization, Sam Buttrey for statistical analyses, and
Ken Davidson, Dick Lind, and Paul Fredrickson for the wave data.

References

Boehm, A.B., Grant, S.B., Kim, J.H., Mowbray, S.L., McGee, C.D., Clark, C.D., Foley, D.M., Wellman, D.E.,
  2002a. Decadal and shorter period variability of surf zone water quality at Huntington Beach, California.
  Environmental Science and Technology 36, 3885–3892.
Boehm, A.B., Sanders, B.F., Winant, C.D., 2002b. Cross-shelf transport at Huntington Beach. Implications for
  the fate of sewage discharged through an offshore ocean outfall. Environmental Science and Technology 36,
  1899–1906.
Cabelli, V., Dufour, A.P., McCabe, L.J., Levin, M.A., 1982. Swimming-associated gastroenteritis and water
  quality. American Journal of Epidemiology 115, 606–616.
Chamberlain, C.E., Mitchell, R., 1978. A decay model for enteric bacteria in natural waters. In: Mitchell, R.
  (Ed.), Water Pollution Microbiology. Wiley, NY.
Eaton, A.D., Clesceri, L.S., Greenberg, A.E. (Eds.), 1995. Standard Methods for the Examination of Water and
  Wastewater, 19th ed. American Water Works Association, Washington, DC.
Flesisher, J.M., Jones, F., Kay, D., Stanwell-Smith, R., Wyer, M., Morano, R., 1993. Water and non-water-
  related risk factors for gastroenteritis among bathers exposed to sewage-contaminated marine waters.
  International Journal of Epidemiology 22, 698–708.
Grant, S.B., Web, C., Sanders, B.F., Jones, B., Boehm, A., Kim, J.H., Redman, J., Chu, A., Mrse, R., Gardiner,
  N., Brown, A., 2000. Huntington Beach Water Quality Investigation Phase II: An analysis of ocean, surf zone,
  watershed, sediment, and groundwater data collected from June 1998 to September 2000. Technical Report
  for the National Water Research Institute and the City of Huntington Beach, 2000.
Grant, S.B., Sanders, B.F., Boehm, A.B., Redman, J.A., Kim, J.H., Mrse, R.D., Chu, A.K., Gouldin, M.,
  McGee, C.D., Gardiner, N.A., Jones, B.H., Svejkovsky, J., Leipzig, G.V., Brown, A., 2001. Generation of
  Enterococci bacteria in a coastal saltwater marsh and its impact on surf zone water quality. Environmental
  Science and Technology 35, 2407–2416.
Grant, S.B., Kim, J.H., Jones, S.A., Jenkins, S.A., Wasyl, J., Cudaback, C., 2005. Surf zone entrainment, along-
  shore transport, and human health implications of pollution from tidal outlets. Journal of Geophysical
  Research 110, C10025. doi:10.1029/2004JC002401.
Haile, R.W., Witte, J.S., Gold, M., Cressey, R., McGee, C., Millikan, R.C., Glasser, A., Harawa, N., Ervin, C.,
  Harmon, P., Harper, J., Dermand, J., Alamillo, J., Barrett, K., Nides, M., Wang, G., 1999. The health effects
  of swimming in ocean water contaminated by storm drain runoff. Epidemiology 10, 355–363.
Hamilton, P., 2004. Chapter 4: Subtidal circulation pathways. In: Noble, M., Xu, J., Huntington Beach Shoreline
  Contamination Investigation, Phase III, Final Report, Coastal Circulation and Transport Patterns: The
  Likelihood of OCSD’s Plume Impacting Huntington Beach Shoreline, USGS Open-File Report 2004-1019,
  Vers. 1.0. Available from: <http://pubs.usgs.gov/of/2004/1019>.
Hamilton, P., Jones, B., Largier, J., Noble, M., Rosenfeld, L., Xu, J., 2004. Chapter 2: Methodology. In: Noble,
  M., Xu, J., Huntington Beach Shoreline Contamination Investigation, Phase III, Final Report, Coastal
  Circulation and Transport Patterns: The Likelihood of OCSD’s Plume Impacting Huntington Beach
  Shoreline, USGS Open-File Report 2004-1019, Vers. 1.0. Available from: <http://pubs.usgs.gov/of/2004/
  1019>.
Kim, J.H., Grant, S.B., 2004. Public mis-notification of coast water quality: a probabilistic evaluation of posting
  errors at Huntington Beach, California. Environmental Science and Technology 38, 2497–2504.
Kim, J.H., Grant, S.B., McGee, C.D., Sanders, B.F., Largier, J.L., 2004. Locating sources of surf zone pollution:
  a mass budget analysis of fecal indicator bacteria at Huntington State Beach, California. Environmental
  Science and Technology 38, 2626–2636.
Leecaster, M.K., Weisberg, S.B., 2001. Effect of sampling frequency on shoreline microbiology assessments.
  Marine Pollution Bulletin 42, 1150–1154.
          L.K. Rosenfeld et al. / Marine Environmental Research 61 (2006) 471–493          493

Lygre, A., Krogstad, H.E., 1986. Maximum entropy estimation of the directional distribution in ocean wave
  spectra. Journal of Physical Oceanography 16, 2052–2060.
MEC Analytical Systems, Incorporated (2000). Huntington Beach closure: relationships between high counts of
  bacteria on Huntington Beaches and potential sources. Final Report.
MEC Analytical Systems, Incorporated (2003). San Diego River–Ocean Beach Water Quality Improvement
  Project. San Diego river sampling and source identification investigations. Phase I – Final Report.
Noble, M., Hamilton, P., 2004. Chapter 7: Tidal transport pathways. In: Noble, M., Xu, J., Huntington Beach
  Shoreline Contamination Investigation, Phase III, Final Report, Coastal Circulation and Transport Patterns:
  The Likelihood of OCSD’s Plume Impacting Huntington Beach Shoreline, USGS Open-File Report 2004-
  1019, Vers. 1.0. Available from: <http://pubs.usgs.gov/of/2004/1019>.
Noble, M., Xu, J., Rosenfeld, L., Largier, J., Hamilton, P., Jones, B., Robertson, G., 2003a. Huntington Beach
  Shoreline Contamination Investigation, Phase III. Executive Summary, USGS Open-File Report 03-62.
Noble, R.T., Moore, D.F., Leecaster, M.K., McGee, C.D., Weisberg, S.B., 2003b. Comparison of total coliform,
  fecal coliform, and enterococcus bacterial indicator response for ocean recreational water quality testing.
  Water Research 37, 1637–1643.
Noble, M., Jones, B., Xu, J., Rosenfeld, L., Hamilton, P., Robertson, G., Largier, J., 2004a. Non-linear internal
  tides move cool sub-surface waters from the mid-shelf to the beach. EOS Transactions AGU, 85, Fall Meeting
  Supplement, Abstract OS32B-05.
Noble, R.T., Lee, I.M., Schiff, K.C., 2004b. Inactivation of indicator bacteria from various sources of fecal
  contamination in seawater and freshwater. Journal of Applied Microbiology 96, 464–472.
Noble, M.A., Xu, J.P., Robertson, G.L., Rosenfeld, L.K. Distribution and sources of surf zone bacteria at
  Huntington Beach before and after disinfection on an ocean outfall – a frequency domain analysis. Marine
  Environmental Research (accepted), doi:10.1016/j.marenvres.2006.02.006.
[OCSD] Orange County Sanitation District, 2001. Marine Monitoring Annual Report FY 2000-01.
Reeves, R.L., Grant, S.B., Mrse, R.D., Copil Oancea, C.M., Sanders, B.F., Boehm, A.B., 2004. Scaling and
  management of fecal indicator bacteria in runoff from a coastal urban watershed in southern California.
  Environmental Science and Technology 38, 2637–2648.
Rosenfeld, L., 2004. Chapter 3: Surfzone bacteria patterns. In: Noble, M., Xu, J. Huntington Beach Shoreline
  Contamination Investigation, Phase III, Final Report, Coastal Circulation and Transport Patterns: The
  Likelihood of OCSD’s Plume Impacting Huntington Beach Shoreline, USGS Open-File Report 2004-1019,
  Vers. 1.0. Available from: <http://pubs.usgs.gov/of/2004/1019>.
Sinton, L.W., 2004. Dynamics of pathogens in marine systems. ASLO/TOS Ocean Research Conference,
  Honolulu, HI, Abs. SS9.06.