Tides and Currents in Fanning Atoll Lagoonl
B. S. GALLAGHER,2 K. M. SHIMADA,2 F. 1. GONZALEZ, JR.,2 and E. D. STROUP2
As PART OF the Fanning Island Expedition simple fashion. In each case, a straight section 1970, selected physical studies were conducted was established by stringing a cross-channel line, in the atoll lagoon. The major effort was the along which bottom topography was measured. measurement of volume, salt, and heat transports (At North Pass, there is, in addition to a main through the three main atoll openings over a channel, an expanse of shoals which were un 24-hour period. In addition, lagoon and ocean covered at low water. This area was ignored tides were recorded, and a cursory survey was resulting in an estimated maximum local volume made of circulation in a small, reef-enclosed transport error of 13 percent.) A single station pond within the lagoon. was chosen at a point along the section that appeared to be in the main flow (see Figs. 1 and 2). At this location, current velocity, tem TRANSPORT STUDY perature, and salinity were sampled each hour, Fanning Island is a roughly oval atoll whose within 0.5 m of the surface. Water level was lagoon, although almost entirely enclosed, does read hourly from a tide staff mounted in a exchange water with the surrounding ocean. sheltered spot near shore. Figures 1 to 3 show Tidal flow in and out of the lagoon occurs at the measurements. The various transports four locations and amounts to roughly 5 percent through these passes were computed by assuming of the lagoon volume over a semidiurnal cycle. the measured parameters to be constant over the channel cross sections. About 90 percent of this exchange takes place at Flow through English Harbor channel was English Harbor, through a channel with a sampled in far greater detail because it was maximum depth of approximately 8.5 m and a expected to be overwhelmingly important in the least width of 290 m. On the east side of the total-exchange picture. It has been reported atoll, opposite English Harbor, is a shoal open (U.S. Hydrographic Office, 1952) that currents ing (Rapa Pass) which handles about 2 percent of the total exchange. A similar channel to the in the channel exceed 5 knots. Taking measure ments from a skiff would be very difficult at best north (North Pass) accounts for roughly 5 in such a flow. Consequently, transports were percent. Air photos show one additional, mean monitored along a semicircular section enclosing dering route which mayor may not lead from the lagoon end of the channel and through which the lagoon to the sea. This possible fourth a more moderate flow could be expected. Figure opening, located about 5 km south of Rapa Pass, 4 shows the section and a smoothed fathometer was ignored in our study. trace along it. Anchored buoys were emplaced Exchange between ocean and lagoon was at stations 1 to 6. The section was traversed monitored for 24 hours, starting at local noon continuously by a skiff which was made fast to on January 7, 1970. Current velocity, tempera each buoy in turn. Seventeen cycles of measure ture, and salinity were recorded at each of the ments were completed during the 24-hour study. three passes. At each buoy, current velocity, temperature, and Both North and Rapa passes were treated in salinity were measured at the depths shown in Figure 4. These data were integrated over the 1 Hawaii Institute of Geophysics Contribution No. cross section to obtain transport figures. 3 Aver- 359. This research was funded by the National Science Foundation under grant GB-15581. 3 Details of all calculations, lists of instruments 2 Department of Oceanography and Hawaii Institute used, and a description of the buoy-anchoring method of Geophysics, University of Hawaii, Honolulu, are given in Hawaii Institute of Geophysics Report Hawaii 96822. 70-23. 191 192 PACIFIC SCIENCE, Vol. 25, April 1971
oNORTH PASS Channel cross section
160 r-----,-r----,...... ,.-----r----r-----r------, 80
~ 120 / 60 ~ / \ E E c: 80 / \ 40~ ~.- / \ .J:: / \ -CI 40 20 'ijj 'I \ .J:: 'I \ c c 0 0 .." ...Q) \ // , ..... :::>- Current speed '-... _ ./ \ I u 6 40 \ / '-'"" 1200
80 ... 60 o c: ~.- 40 c: ...c ... 20 Q) E :::> o o >:; I o I 20
1200
FIG. 1. Channel cross section showing station position (solid dot and vertical line), tidal height and current speed, and volume transport for North Pass during the 24-hour transport study. Dense patches of Turbinal'ia that nearly reach the surface at low tide are located between 20 to 40 m from the east shore. Cross-sectional depths are relative to a tidal height of 13 cm. Absolute sea level is arbitrary. Tides and Currents in Fanning Lagoon-GALLAGHER ET AL. 193
RAPA PASS Or-----,------,-----r->---r---.,...-----.-----r-
..c. Q. Q) 0.8 o
v 60 80 Q) cQ) 0 / , 200 ~~ ""tl ~ ~ / Current speed \ ~ 0 " ~- ..... _-; " u 20 v" , o 1200 1600 1200 7 Jan,1970 v ~ 40 ME -; c 20 0' Q. C 0 o~ ~ ~ ~ 0 Q)E 20 L-. l...... - l...... -::--__l...... - "-- "--__----' ~ 1200 1600 2000 0000 0400 0800 1200 ~ 7 January 1970 8 January FIG. 2. Channel cross section with station position (solid dot and vertical line), tidal height and current speed, and volume transport for Rapa Pass during the 24-hour transport study. Cross-sectional depths are relative to a tidal height of 38 em. Absolute sea level is arbitrary, ages of all the temperature and salinity measure each period of inward flow, a jet developed and ments at buoys 2, 3, and 4 for each measurement extended into the lagoon from the channel itself. cycle are presented in Figure 5, along with the The jet appeared at buoys 2, 3, and 4, where volume transport through the entire channel. speeds exceeding 3 knots could be found. On A recording current meter was later placed either side of the base of the jet, eddying near the surface at buoy 3, the site of strongest motion, which could be observed visually, ap flow. The resulting record is shown in Figure 6. peared in the measurements of volume transport at stations 1, 5, and 6. During the ebb, however, RESULTS OF THE TRANSPORT STUDY there was outward flow at all locations across Volttme our section, funneling into the channel. Figure Currents through the channel at English Har 7 illustrates these patterns in the volume trans bor displayed an interesting pattern. During ports at each buoy. The same jetlike flow also 194 PACIFIC SCIENCE, Vol. 25, April 1971 100 35.5 80 j 0354 60 . ~ ... o o .~35.3 40.s ~ I c c o 1 20 .-... ~ 35.2 Q) o E ..... :::) 35.1 LVolume transport 1200 1600 1200 7 January - RAPA PASS ou 30 .-----=----.------,.------,-----.-----.------...... , u Q) - \ ~ ~ 28 ...... !'_J"emperature ~/ M 35.3 :> .... ,,,,,,,,,, --, ~------" E ~ ~, o - I , I - ~ 26 'v....1 ''../ 40 "t: o-e ~ o >.35.2 / \/ \ Q. c ." c ~ 24 ....Volume :, 20 c ~ ...... , .~ 35.1 transport o .-e c./) Q) :::) E 35.0 20 02 ~ 1200 1600 2000 0000 0400 0800 1200 7 January 1970 8 January FIG; 3. Salinity, temperature, and volume transport for North and Rapa passes during the 24-hour transport study. developed on the seaward side of the channel tion.4 The lagoon's response to the ocean tide during ebb and could often be seen from shore. would be an intriguing subject for further study. Another noteworthy feature of the current was The first semidiurnal cycle of transport studied its rapid reversal, with relatively short periods coincided with a typical tidal excursion. During of weak flow between strong ebb and strong this period (1440 to 0120) 0.025 lan3 of water flood. This indicates that the lagoon has a non was computed to have entered the lagoon linear response to tidal excitation. The presence through the three passes. The computed out- of nontidal overtones is seen in the current and lagoon tide records displayed together in Figure 4 There probably is also some error in the relative 6. As one would expect, the current shows phases of the curves. in Figure 6; the current should not lag the lagoon-tide derivative. We cannot find a strong relation to the tide's rate of change, but timing mistake in either record and can only point both processes clearly reflect nonlinear distor- out this discrepancy. Tides and Currents In Fanning Lagoon-GALLAGHER ET AL. 195 OA 1 2 3 4 5 6 B 2 lagoon 4 2 3__ "~5 N~1''''-'- 6; ..s4- 'A ' B ...c 0.6 Q.) 0 ~ 8 ocean 10 o 2652 0 200 400 600 ~s Distance (m) FIG. 4. Smoothed channel cross section (left) and buoy locations (right) for English Harbor channel. Depths (in meters) at which measurements were most often taken during the 24-hour transport study are indicated for each buoy. flow was 0.026 km3. The agreement between 107 kg) also matched inflow (86 X 107 kg) these numbers is gratifying, because the tidal within our limits of accuracy (see Fig. 8). Salin records show no net change in lagoon water ity followed no discernible pattern (see Fig. 5). level over this particular period. Moreover, At North and Rapa passes, inflow exceeded the measured exchange, 6 percent of the outflow by 2.0 X 107 kg of salt (Figs. 9 and mean lagoon volume, agrees with the value 10). Salinity in these passes decreased during estimated using tidal range and mean depth in flow into the lagoon and increased during out the lagoon. In spite of these consistency checks, flow. Because these passes are shallow, solar however, the transport figures calculated from heating may cause the levels of evaporation in our measurements should not be regarded as the lagoon to exceed open-ocean levels, at least exact. Inaccuracies in current speeds, directions, locally near the mouths of the passes. The and areas of channel cross sections lead to a higher salinity outflow could be partly caused by volume transport uncertainty of +-15 percent. the mixing of incoming seawater with more There was a net inflow of water through saline lagoon water. However, this cannot be North and Rapa passes during the first, semi proven. If the ocean water merely flowed in and diurnal tidal cycle monitored. The amount was out of the lagoon, with no mixing, estimates 0.56 X 10-3 km3, or 0.1 percent of the lagoon show that local evaporation in the lagoon was volume. Thus, there is an overall flow across the sufficient to account for the observed salinity lagoon which must exit at English Harbor; this increase. flow is probably driven by wind and wave trans port over the windward and northern reefs. Heat In the first transport cycle, during which net Salt volume transport was zero, net heat transport During the first transport cycle, the computed (see Figs. 8 to 10) was also zero to within total salt transport into the lagoon (90 X 107 the accuracy of our measurements. Total inflow kg) matched (within our limits of accuracy) (68 X 1013 calories) was closely matched by the total salt transport out of the lagoon (91 X outflow (66 X 1013 calories). There was a net 107 kg). inflow of heat at the small passes (1.6 X 1013 At English Harbor channel, outflow (90 X calories), but this reflected the particular phas- 196 PACIFIC SCIENCE, Vol. 25, April 1971 Er-N_G_L_IS_Hr--H_A_R_B0r--R---r------,r------,------,3000 u o -, - I\" \ Q.) 29 , \ '- ....:::> _ _ Temperature " " o 28 ~ ------~~~~-~-~-- 2000 '- I \ Q.) ~- ... I \ 0 ,. , I \ 27 I\ E \ I \ , I \ c ~ I \ : \ Volume , \ : \Transport , \ 1000 ~ I \ ~ I , I \ M I I , \ E , , , \ , I , \ .... , I '- o I \ I , , \ o ~ , \ , , 0- o 35.2 I , , o V) >- , c I '\ o \ ' c \ \ .... \ Q.) o \ \ E V> \ \ :::> \ \ 35.1 \ \ 1000 o \ \ \ \ > \ \ .... \ \ :::> \ \ 0 \ \ I .... _ ...L 2000 1200 1600 2000 0000 0400 1200 7 January 8 January FIG. 5. Averages of the salinity and temperature measured at buoys 2, 3, and 4, English Harbor channel, and volume transport across the entire channel, during the 24-hour transport study. 20 ENGLI~H HARBOR u 15 Q) 0 ~c .5.-10 i-= L.------;::-,00;f;0'"""0,------:::12l-;:-0-=--0-~0~00::-:0:-----:::12.l.:-0-:-0--...,..00-,.l0.,....0---12LO-0---=O~OO=O:----==--.J 0 15 Jon 70 16 Jon 17 Jon 18 Jon FIG. 6. Aanderaa current meter record at buoy 3, English Harbor channel, and lagoon tide. Absolute sea level is arbitrary. Tides and Currents in Fanning Lagoon-GALLAGHER ET AL. 197 ENGLISH HARBOR 1200r----.--.----r--~--r__-___. Buoy 1 Buoy 5 ,-, \ II 2 II , 1000 6 I\ II I\ 3 I\ II 4 I .•... \ ,:,''' ".'. \\ 800 ! \ \ I ! : , I,,,'" I : ~ , I\ I ! ~ , I\ iI , I .. , ,:"...... \ : ~ c 600 ..., I : '. Ii \ , I: : ,, Ii \ \ I! : , ,.If \.', U Ii ~ , -Q) Ii \ , 400 If ~ , ii, ~ f H E H i /-- i' : /' -,,~' f / \:' JI ~ ~ fI I''" o 200 \~ ~. I, , " i Q. ~ . VI "...... : c ...... : '\ o . : , ~ o -/-.q.L---""' -Q) E ~ :> / ..,.:' ~ 200 :,il :,II -:> ::t, o :, 400 .: I !I I I 600 1200 0000 1200 7 Jan 70 8 Jan FIG. 7. Volume transports assigned to each buoy during the 24-hour transport study at English Harbor chlnne!' ing of the diurnal temperature curves with the temperature and salinity occurred at both passes. semidiurnal current variations, and may have It seems likely that this illustrates the rapid re had no particular significance. It does seem sponse of temperature in this shallow-water en strange, however, that the temperature variations vironment to the large diurnal variation in the were diurnal and did not show the effects of the flux of radiant energy. The diurnal variation tidal flows which appeared in the salinity curves. in evaporation was proportionately very much The disagreement in the periodicities of the smaller, so that no diurnal salinity fluctuation is 198 PACIFIC SCIENCE, Vol. 25, April 1971 100 ENGLISH HARBOR 100 80 80 60 60 Z~ ~Z40 40-~ ...... lI'I o U ~ 0- 20 0 CO 20 P"'"" P"'"" - -~ ..~ o 0 o 0- 0 f/) 0- c lI'I o C ~ 0 20 \ 20 ~ .. .. \ .. \ o .. Q) 0 ~40 \ ~::c V") \ 40 :::> \ / :::> °60 '''' 60° 80 1200 FIG. 8. Salt and heat transports at English Harbor channel during the 24-hour transport study. obvious in the record. At English Harbor, there with large coral formations, but these were iso was no significant temperature variation. This lated and did not seriously baffle the flow. The agrees with other observations which indicate water here was relatively deep and quite clear. that the water which entered and left through The clear water contrasted markedly with the the channel was essentially open-ocean water. extremely turbid water found throughout the Changes incurred by mixing with lagoon water rest of the lagoon, and the boundary between were too small to be detected. clear and turbid water appeared to be vertical and often only a few meters wide. This bound Mixing ary was sometimes observed moving in response Replacement of water in the lagoon appears to tidal inflow and outflow through the channel. to be a slow process. A major reason for this is These facts, together with the temperature and the fact that the lagoon was divided into numer salinity records from the transport study, indi ous, small ponds by interconnecting line reefs, cate that very little mixing occurred between many of which were almost uncovered at low lagoon water and the water that participates in water. Flow was thus restricted to a shoal sur the predominant tidal exchange. Mixing is face layer and to meandering passes through the likely to be relatively more important very lo reefs. An exception was the area receiving flow cally near North and Rapa passes where no topo through the channel at English Harbor. Lagoon graphic or turbidity boundaries were apparent ward from the channel, the bottom was scattered and the water was shoal. Such mixing, because Tides and Currents in Fanning Lagoon-GALLAGHER ET AL. 199 4000NORT H PASS 3000 ..., I\ I, I\ I\ 2000 I\ I\ v I\ Q) I\ ~ I o c V ..... 1000 I .... I FIG. 9. Salt and heat transports at North Pass during the 24-hour transport study. of the line reefs and because the volume ex Tides change through these passes was relatively very The main tidal records were kept near the small, would not greatly reduce lagoon-wide Cable Station on the northwest side of the residence times. island; readings were taken in the lagoon and One possible lower limit on residence time in the ocean. The ocean gauge was fastened can be calculated by using the observed volume to an outhouse piling a few meters seaward from of net inflow through the two small passes. If the water's edge, while the lagoon record was this inflowing water is assumed to mix com obtained at the end of a small pier. The mea pletely throughout the lagoon before exiting at surements were used to make daily predictions English Harbor, then we have for biologists on the expedition who were in volved in nearshore collecting. The tide curves, lagoon volume shown in Figure 11, are useful for examining residence time = rate of net exchange ranges and phases. However, the records are not .41 km3 related to any common reference level. More ------= 11 months. over,. the cur~es contain several interruptions .56 X 10-3 km3/11 hr occasIOned by mstrument malfunctions, and no 200 PACIFIC SCIENCE, Vol. 25, April 1971 2000 RAPA PASS 2000 1500 u - u ~ c 1000 1000 Q) ,.- ~ 0> ..;:,L. 0 U -0 0 .... r-- ~ 500 -.... 0 ~ a. 0 ." a. c ." 0 C ~ 0 ...... ~ .... 0 0 .... 0 0 lI) Q) I .... :::> 0 500 1200 1600 2000 0000 0400 0800 7 January 1970 8 January FIG. 10. Salt and heat transports at Rapa Pass during the 24-hour transport study. attempt was made to maintain a constant zero The lagoon tide lags that of the surrounding level between fragments of a given record. ocean by a typical period of 1 hour, 40 minutes The observed tidal range in the lagoon was as shown in Figure 12. The figure also shows typically 40 em. This provided an immediate tides measured in North and Rapa passes dur estimate of the volume of tidal flow. Half the ing the transport survey and the tide outside the tidal range was 5 percent of the mean lagoon Cable Station. Although the measurements in depth estimated by the geologists, which implies the passes were taken roughly halfway between that 5 percent of the lagoon volume took part lagoon and sea, the curves appear to be about in tidal exchange. Direct measurement of the the same in phase and in range as the ocean tide volume transport through the passes confirmed measured at the Cable Station. this value. The range in the lagoon was roughly There is a temptation for practical reasons to half that in the ocean outside. see whether the ocean tide at Fanning Island is Tides and Currents in Fanning Lagoon-GALLAGHER ET AL. 201 related in some simple way to Honolulu tides. sively along the windward boundary of the The Honolulu tide is shown plotted against the pond, indicating the directional consistency of Fanning Island tides in Figure 11. No relation the wind-generated waves and currents over the ship that could be used to give rough Fanning reefs, and supporting the finding that currents predictions based on Honolulu records is ap at depth are weak. parent. Semidiurnal phase lags, based on very Tidal effects were superposed and sometimes scanty records, varied from 16 minutes to 1 masked by the wind-driven flow. Currents were hour, 42 minutes for high water and from measured at several locations on the shoal reefs -48 minutes to 2 hours for low water. The bounding the pond, both during rising and dur amplitudes were also quite different. ing falling tides. At Suez, in a channel roughly 1 m deep and 5 m wide dredged through the Suez Pond leeward reef, an ebb flow moved against the A large part of Fanning Island Lagoon is wind at 0.1 to 0.2 knots. Tidal reversals were subdivided into small ponds by interconnecting not detected at any other location. Tidal cur line reefs that reach nearly to the surface. A rents almost surely crossed the reefs at other typical example, called Suez Pond in this paper, places, and future measurements should include was chosen for intensive geological and chem a study under conditions of no wind. To the ical study; we attempted to provide some sup north and west (and not communicating with porting information about water circulation. the pond) is a meandering pass which can be (See Roy and Smith in this issue of Pacific Sci followed through the reefs from the Cable Sta ence for a further description of the pond it tion to English Harbor. It is possible that this self. ) serves as a channel for overall southward flow Water movements in the pond proved to be in the area; the ebbing tide and the necessary extremely complex and usually too weak to mea return of wind-driven surface transport may be sure quantitatively with standard current meters. minimal in Suez Pond itself. Winds, tides, and thermohaline forces all ap We attempted to trace subsurface flow with peared to be important in the circulation. Fur dye, but found this to be of limited value be thermore, processes on the bordering reefs and cause of the high turbidity in the pond. On one in adjacent ponds seem to have had an influence occasion, however, we were able to detect sub as well. Our time and instrumentation were al surface movement against the wind. Dye intro most totally inadequate for the task of de duced in a vertical streak near the windward scribing the circulation, and we are able to make edge of the pond broke into two patches. A only qualitative comments. surface patch moved with the 9-knot wind at a During much of the year Fanning Island lies maximum estimated speed of 0.4 knot, and in the southeasterly tradewinds, which blow became too diffuse to see after about 40 minutes. across the lagoon at mean speeds of 10 knots A second patch, extending downward from from 135°. There are diurnal speed variations about 0.5 m, moved up against the windward with afternoon winds being 10 to 20 percent reef at an estimated speed of 0.06 knot and stronger than morning winds (New Zealand could not be seen coming to the surface. The Meteorological Service, 1956). Short, choppy dye may have moved downward, or it may have waves up to 0.3 m high were generated within ascended very slowly and become too diffuse to the lagoon itself. With exceptions to be noted detect in the surface flow. Dye injected as a later, surface flow across the pond and adjacent point source in the same location at 3 m depth reefs was in the direction of the wind. Within was never detected throughout 2 hours of ob the pond, the directly wind-driven layer was servation. less than 1 m deep with speeds on the order of Distributions of properties showed a compli 0.1 knot. We found no simple pattern of return cated pattern of inhomogeneities. Salinity was flow at depth; currents at 3 m were variable sampled at depth at five locations during a 3 in direction with speeds too low to measure hour period. At four locations salinity increased (< .05 knot). Sand from the tops of the adja by .01 to .03%0 from the surface to the bottom, cent reefs was found deposited almost exclu- and one station showed a decrease of .01%0' 100 , /1", \ -, ,.. /1II' \\ :\ ,'" Fanning: \ !\ E80 Honouu/II \ ocean! \\ : " ,\ r'\ \ I \ .~ ,\/\ " f' :\ C '\ '\ /\:\ .~ 60 !\ I \ f ~ i \ ~ !\ I \ I /\ f A\ ~ 40 f I I \ I I \ f 'I ..- , I \ I'I \ : \ / ," ~\J! ",;" , \,' , \ , , \' '" \ 20f-, /\ V \\' 'J D --A.Fanning: ,- o .- lagoon ILUU 70 100 ," -E 80 " \Honolulu ~ I\ .- I\ ~ I\ ,-, o I \ , 0Qi , \ ., 60 , \ \ ~ \ \ o , I \ ,: \I \ ~ 40 I I ..- , / I \ , \ \ , I I \ \ \ \ \ \, , ,, , ..., ... , 1200 0000 1200 0000 1200 0000 1200 10 Jan 11 Jan 12 Jan 80 E u ~60 ,..., , ..., I ' .Ql Honolulu,'" " I , I " ,., I\ 1/\ ' I , Q.) ", I, ...c '" I , I' \' ...... ')(\ ' ' \ 1'\ 40 I" " \ /1 \ ',_..... / \\ "\ ...... , -0 , -"' /1 ,/ \ I \\ I ..... , -0 I ' "'- 'I \ ', ~ , /"' 20 \\ ,'1"IFanning\\ //"'\ I \."1/ ocean ~ - I J,... ' ' 1200 0000 1200 0000 1200 0000 14 Jan 15 Jan 80 ", I\" I\ I \ , ... I -E I , \ , , , \ \ I \ U " I , \ , \ \ I , I \ '="60 , , \ , \ ...c " Honolulu \ C> , \ , , , \ , \ I I Q.) , , \ I \ , \ ...c I \ , I \ , ~\ -40 \ , \ 0 \ , , -0 I .- \-..... A., I I ~ ,'" I , I ... A, I 20 , ,...... , '.' lagoon I 1 1 I I I II I 1200 0000 1200 0000 1200 17 Jan 18 Jan FIG. 11. Tidal height measured in the lagoon and the ocean near the Cable Station, Fanning Island, compared with tide at Honolulu (CGS record). Ab· solute sea level is arbitrary in the Fanning records. 204 PACIFIC SCIENCE, Vol. 25, April 1971 c ~20 I- 1200 FIG. 12. Tidal heights at North and Rapa passes, and in the lagoon and the ocean near the Cable Station, during the 24-hour transport study. Absolute sea level is arb:trary. Variations of .04%0 over distances of a few hun peared to bring turbid water from the leeward dred meters occurred at all depths. One of the reef into the pond to the east of Suez. At depths stations was repeated after 24 hours, and salin between 2 m and 4 m additional horizontal ity had increased almost 0.3%0 at all depths. We patches of suspended load could be seen. They found no water of this higher salinity in the were less intense and had no apparent connec pond on the previous day, so the rapid replace tion with those at the surface; they may have ment of the 8-m water column at this station been remnants of surface patches in the process does indicate vertical movement. The salinity of settling. (Particulate matter of this size has data indicate thermohaline circulation which was a settling rate on the order of 1 m/day.) Below very irregular in space and in time. Probably 4 m the turbidity became almost uniform hori patches of high-salinity water originated over zontally, with an intensity that appeared to be the wide, shoal, boundary reefs; were advected close to an average value for the overlying into the pond; and, perhaps with nighttime cool water. ing, gave rise to vertical circulations which can The salinity and suspended load data suggest have small horizontal dimensions compared to some possible general conclusions about mixing the pond. Below about 3 m, the pattern of such in the pond. The uniformity of turbidity at circulation would be further complicated by the depth, along with the settling rate of the parti very irregular topography. cles, suggests that the pond had a mixing time The distribution of extinction coefficients in of less than about 5 days. Salinity indicates that the water column is shown in Figure 12, Roy the mixing time certainly exceeded a few hours, and Smith, this issue. Several factors appeared. and that at least part of the mixing was asso A patch of high turbidity, covering about one ciated with diurnal, thermohaline processes. One third of the pond's area, was apparent at the might suspect that deeper portions of the pond surface and extended down to about 2 m. are filled by relatively dense water that is re Within the patch, suspended load decreased newed only infrequently by unusual events. with depth, indicating that the patch was quite However, there were no indications of stagna recently formed. The horizontal distribution tion; oxygen concentrations were close to satura strongly implies that this highly turbid water tion at all depths. Thus, there must be a fairly entered the pond through Suez Canal during regular downward transport of buoyancy, indi the falling tide and flowed out over the surface. cating an external source of mixing energy. The Shallow tidal flow against the wind also ap- most probable agent is the wind-driven flow, Tides and Currents in Fanning Lagoon-GALLAGHER ET AL. 205 but we have insufficient measurements for pro occur rapidly and over short distances are to be posing a circulation model which could explain expected. On-the-spot instrument readout would the mixing and the observed distributions of be extremely valuable. properties. Should members of a future expedition wish ACKNOWLEDGMENTS to obtain a comprehensive picture of physical We thank C. J. Berg, Jr., R. E. De Wreede, processes in the pond, they would face a diffi E. B. Guinther, Dr. E. A. Kay, G. S. Key, and cult task. We suggest from our experiences that J. P. Villagomez for their help during the 24 they employ very sensitive current meters (re hour transport study; Dr. D. C. Gordon for sponsive to 1 em/sec or less) to see whether making all salinity determinations; Dr. K. J. any average, overall patterns of movement can Roy for his help in obtaining fathometer records be discerned. Great effort would be required to of the bottom topography between buoys at ensure no motion of the meters themselves; English Harbor channel; and E. G. Gilley for they cannot simply be lowered from a single keeping the boats operational. anchored skiff. Hourly vertical profiles should be obtained at as many stations as possible (at least four on a line roughly parallel with the LITERATURE CITED wind and through the center of the pond) . We NEW ZEALAND METEOROLOGICAL SERVICE. advocate this direct approach because the patchy 1956. Line Islands. Wellington, Meteorologi and transient nature of property distributions cal Notes II-B. would limit the value of these profiles for in RoY, K. J., and S. V. SMITH. 1971. Sedimenta ferring motion unless they were based on very tion and coral reef development in turbid dense sampling. If a general circulation pattern water: Fanning Lagoon. Pacific Science, vol. were found, then temperature, salinity, and tur 25, no. 2, pp. 234-248, 15 figs. bidity measurements could be planned to com U.S. HYDROGRAPHIC OFFICE. 1952. Sailing di plement and take advantage of it. Such a plan rections for the Pacific islands. III. The south should be designed to maximize the density of central groups. Washington, Hydrographic readings in both time and space; variations that Office Publication 80.