NZOI RECORDS - Te Ara

NZOI RECORDSNew Zealand Oceanographic Institute,Wellington, New Zealand

Vol. 4 No. 7 November 1980

A BATHYMETRIC SURVEY OF THE

WAIMANGU THERMAL LAKES

R.F. Keam

Department of Physics, University of Auckland

ABSTRACT

Techniques and equipment used for determining the bathymetry ofhot lakelets in the Waimangu thermal area are described.

INTRODUCTION

The Waimangu hydrothermal system(see Figs 1, 2) is remarkable for a numberof reasons, not the least being that itcurrently exhibits a quasi-cyclic behaviourwith a period of approximately six weeks.During a cycle the hot water lakelet inInferno Crater fluctuates between over-flow and a depth of eight to ten metresbelow this level (Fig. 3). The neighbour-ing Frying Pan Lake in Echo Crater(Fig. 4) exhibits sympathetic variationsin outflow*. During the late 1960s asystem to monitor this behaviour wasestablished, mainly by E.F. Lloyd, and asummary of the data acquired is reportedannually (Lloyd 1973, 1974, 1975; Scott

1975, 1976. 1977, 1978). In an attempt todevelop a quantitative model of the hydro-thermal mechanism responsible for suchsystematic behaviour it has been neces-sary to determine accurately the bathy-metry of the two hot lakes (Keam 1976,1977, 1978, in press a.b). Because of theunusual problems involved in obtainingdata for hot lakes, it is consideredappropriate to detail the methods used.Also briefly described are general meth-ods used for sounding thermal springs.

* In 1963 Tourist and Publicity Departmentstaff re-named Frying Pan Lake as WaimanguCauldron, and re-named Infemo Crater (or itsvent) as Ruaumoko's Throat. In this paper (andmost other scientific publications) the oldernames are retained.

30JUN193!

56 NZOI RECORDS

bO

KEAM - HOT LAKES SURVEY 57

•aI

EH

58NZOI RECORDS


SOUNDINGS FROM SHORE

Soundings in thermal springs of com-paratively small area have been made bymany investigators by lowering a weightedline to the bottom and measuring thelength paid out. When pool diametersexceed 5-6 m it becomes necessary tosuspend a line in some fashion over thevent. In their 1941 temperature survey ofhot springs in Rotorua-Taupo thermalareas, Marshall and Rands (Rands andWilson 1954) tautly held a second linewith a ring on it across large pools, andlowered the thermometer line through thering.

As early as 24 May 1951 E.F. Lloydand I had made a spot sounding of theInferno Crater lake in the following man-ner : a piece of wood, with cork attachedto increase its buoyancy, was floated outon to the lake. Its position was controlledby a tether at either end and by the con-trollers walking along the lake-shore andcrater edge and pulling the tethers taut.A further line with a sinker on it ranthrough a hole in the centre of the woodenfloat. The sinker was raised to touch thefloat and when the latter was in positionthe sinker was lowered until it reachedthe lake bed. Knots were tied by theoperator controlling the sounding line atcontiguous positions on the sounding lineand control tether. When the system wasrecovered, the difference in length, be-tween sinker and knot on the soundingline and float and knot on the control line,gave the lake's depth.

Examination, in 1972, of newly avail-able Waimangu monitoring data led to thesuggestion of a qualitative model forexplaining the observed behaviour of thehydrothermal system (Keam, pers. comm.in Stanton 1978 : 18-23; Keam 1980). Inorder to develop the model further a seriesof spot soundings of Inferno Crater lakewas made on 6 January 1973. The methodwas similar to that used in 1951 exceptthat a plastic bottle replaced the woodenfloat. It was clear, however, that forquantitative work one needed better thansketch locations for the soundings.

E.F. Lloyd established several surveystations and began a systematic series ofplane-table surveys of the edge of InfernoCrater lake at intervals of approximately1 m from overflow to the lowest level itreached during a recession. These werethen supplemented by further soundingsusing the float method when the lake wasnear a minimum level. Sounding locationson this occasion were determined by aplane-table survey of each float position.The method of measuring the fall of thesinker was simplified for each soundingby attaching a flexible metric tape to thesounding line when the sinker was at thefloat and noting how far this was pulledout when the sinker was lowered to thelake bed. The tape was detached betweensoundings while the float was reposition-ed.

The water in the main basin of InfernoCrater lake covers a circular area 60 m to65 m diameter when at its minimum level.A larger lake would present considerabledifficulties in sounding by the floatmethod, mainly because the sinker wouldhave to be inconveniently heavy in ordeito pull the sounding line through the floatrather than allowing the portion of sound-ing line between float and operator to sagunder its own weight to the lake-bed. ForFrying Pan Lake, whose width reaches200 m, the alternative is to take sound-ings from a boat.

BOAT SOUNDINGS

Maji Moto, the boat used, was speciallydesigned incorporating specific safetyfeatures for use on hot lakes. (See Appen-dix for details on construction and use.)Certain types of wooden dinghy could bemodified to include these features.

During December 1974 and February1975 staff and students of the PhysicsDepartment, University of Auckland,established a network of more than 40survey stations at Waimangu and thesehave been accurately fixed with respectto the Bay of Plenty Circuit. Station ele-vations were determined by levelling frombench marks and by using vertical angle

60 NZOI RECORDS


^V_^*, ™

O to

g o | S g .

§ o § -c

.9

a

62 NZOI RECORDS

REAM - HOT LAKES SURVEY 63

theodolite observations and a leastsquares fitting procedure. A level datumhas been established for each lake.

Sounding locations were determinedfrom theodolite observations with theinstruments being set up on stations ofthe established network (Fig. 5). Thetheodolite target was a small sphere ontop of an aluminium tube directly abovethe sounding reel. Soundings were madeby hand-lowering a lead weight at the endof a stainless steel or fibre-glass tapewound on the reel.

Communication between the boat crew,tether controllers and theodolite operatorswas predominantly by whistle signals butwas also by direct voice and radio-telephone. Various configurations andnumbers of tethers were used. The mostefficient method utilised two tethers, oneattached to each side, with the drawbackthat the boat was still able to moverelatively freely normal to the line of thetethers. Wind gusts tended to producemovement, so soundings were made whenmovement was judged to be least, awhistle sighting-signal being given forthe theodolite operators (who were follow-ing movement of the target sphere withtheir telescopes) to freeze movement oftheir instruments simultaneously at theinstant when the weight made contactwith the lake-bed. A third tether, tied tosome shore object and operated from theboat, was used if persistent wind made itotherwise difficult to stay on station. Arecent modification was the provision ofa wide board able to be lowered into thewater through slots parallel to and hardagainst the stem. Its drag in the waterwhen lowered reduced the effect of windgusts. It was disadvantageous to use it,however, when there were significantcurrents.

Maji Moto was also used for soundingsbelow low water level in Inferno Craterlake during the periods 16-17 December1976, 15-17 March, 30-31 October 1977and 14-16 March 1978. A total of 786soundings were recorded from the boattogether with 78 earlier ones using thefloat method (Fig. 7).

THEODOLITE TECHNIQUES

The earlier plane-table determinationsof the location of the lake-edge over itsrange of levels in Inferno Crater weresuperseded by a method which uses asingle theodolite ("Angle of Depression*technique). The geometry is such thatfrom a suitable elevated position, andprovided the instrument's height abovelake-level is known, observations ofvertical and horizontal angles to pointson the lake edge by a single theodoliteare sufficient to define the co-ordinatesof each point so observed. Both with thismethod and when sounding it was neces-sary to monitor carefully the movementsof lake-level which typically are around20-25 mm/hour.

Steam severely restricted visibilityduring periods of high lake-temperatureand it was found impractical to use eithersounding or angle-of-depression methodsfor some littoral portions of the InfemoCrater lake basin. Such regions of thebasin which were accessible in safetywere surveyed by establishing a sequenceof temporary stations (by laying a colour-ed golf-ball on the surface) and observ-ing these with two theodolites on net-work stations. Regions not so accessiblewere comparatively economically survey-ed by a single operator observing identi-fiable physical features alternatelythrough two theodolites set up on adja-cent network stations.

During two separate week-long exped-itions in May 1975 and November-Decem-ber 1976, and on 17 March 1977, 28-29October 1977 and 16-17 March 19781173 soundings were made in Frying PanLake at well determined positions and thebathymetry was defined (Fig. 6).

ACCURACY

Precision in determining the positionsof sounding locations has varied, depend-ing on environmental ^ i J

64 NZOI RECORDS

25 METRES 50

CONTOURS IN METRESAT 2m INTERVALS

Fig. 7. Outline bathymetry of InfernoCrater Lake. [In February 1978 a hydro-thermal disturbance formed a new ventENE of the deepest existing vent. Thebathymetry shown here is that whichexisted before this event.]

causing boat movement, steam obscuringsighting sphere) and increasing withexperience with any given set of environ-mental conditions. At times three theodo-lites were used simultaneously. The in-circle radius of the horizontal triangleformed by intersection of vertical planesthrough their axes seldom exceeded 10 cm.

Further checks were available throughthe theodolite vertical angle readings,which were also systematically recorded.The position of the sighting sphere cantherefore, in most cases, be regarded ashaving been determined to within ca.10 cm.

Whether the sphere, in these circum-stances, was directly over the sinkerdepended on inclination of the tape in thewater. This in turn depended on whether


significant drift had occurred during thetime the tape was being lowered and be-tween lake-bed contact and the sightingsignal and on the effects of any currents.The top metre or so of the tape was ob-served through the centre-board casing,and when there was any noticeable depart-ure from the vertical, or too much boatmovement, the sinker was raised andlowered again. A departure from the ver-tical of > 5° would probably have beennoticed. Maintained over the length ofthe tape, this would give a maximumpositional error of 8% of the water depth -for Frying Pan Lake with an averagedepth of approximately 6 m this would beabout 0.5 m. The error region for locationcan therefore be regarded as an ellipsewith semi-minor axis around 10 cm,semi-major axis < 8% of water depth andoriented with the minor axis along theline of tethers. (Two sets of soundings inInferno Crater were achieved in complete-ly calm conditions with expected locationerrors of no more than a few centimetres.)

The sounding tape was a 30 m carbonsteel measuring tape graduated in milli-metres but replaced by a fibre-glass tapefor the March 1978 soundings. Depthswere read to the nearest centimetre. Witha hard bottom and no-drift conditions thiswas found to be completely repeatable.With a soft bottom one had to define thedepth and I took it to be the level atwhich the sinker stopped sinking at anoticeable rate. The main depth errorwould probably have arisen from non-verticality of the tape, as above, with afive degree slant contributing a 0.3%excess depth error. Allowing for soft bot-tom uncertainty and some curvature(catenary effect) a 10.00 m measureddepth would almost certainly imply a realdepth within the range 9.95 m to 10.02 m.Corrections where appropriate have beenmade for thermal expansion of the steelsounding tape. At most these amount to1 cm.

Soundings by the float method wereless accurate than boat soundings, withan expected error of ± 0 . 1 m. Those usingplane-table locations of the float havebeen disregarded and only soundings usingtheodolite determinations of position have

been used in defining the bathymetricmaps. Location accuracy can therefore beregarded as comparable with that for theboat soundings.

ACKNOWLEDGMENTS

I am grateful to D.A. Jones, SeniorTechnical Officer, Physics Department,University of Auckland for designing theMaji Moto and for incorporating the specialsafety features. The boat was suppliedby the Plywood Association of New Zea-land and constructed by R.L. Fink ofTamahere.

I should also like to record my grati-tude to the following peop1^ and organ-isations : E.F. Lloyd, now of GeologicalSurvey, for help in our early investigationsand later survey work and the initialcontouring within Inferno Crater lakebasin; B.J. Scott, W. Crafar, P . van derWerff, also of N.Z. Geological Survey;the Tourist and Publicity Department forfreedom of access into the WaimanguScenic Reserve and the willing personalhelp of its employees, particularly A.S.Marx, P . Scott, S. Hitchens, H. McNickle;the Lands and Survey Department for con-venient accommodation; and other staffand students of the University of Auck-land for many hours of both rewarded andunrewarded labour often in trying weatherconditions. This work was also supportedin part by a research grant from the Uni-versity of Auckland.

REFERENCES

KEAM, R.F. 1976: Bathymetric survey of Fry-ing Pan Lake. N.Z. volcanol. Rec. 5 : 68-69.

KEAM, R.F. 1977: Research at Waimangu byUniversity of Auckland Physics and Geol-ogy Departments. N.Z. volcanol. Rec. 6 :49.

66 NZOI RECORDS

KEAM, R.F. 1978: Research at Waimangu byUniversity of Auckland Physics Depart-ment personnel. N.Z. volcanol. Rec. 7 :63-65.

KEAM, R.F. 1980: Waimangu. Pp. 52-54 inHochstein, M.P.; Hunt, T.M. (eds) "Guideto geophysics of the volcanic and geo-thermal areas of the North Island, NewZealand*. Royal Society of New Zealand,Wellington. 93 p.

KEAM, R.F. (in press, a): Inferno Crater LakeProvisional Bathymetry, 1:250. N.Z.oceanogr. Inst. Chart. Misc. Ser. 52.

KEAM, R.F. (in press, b): Frying Pan LakeProvisional Bathymetry, 1:500. N.Z.oceanogr. Inst. Chart. Misc. Ser. 53.

LLOYD, E.F. 1973: Geothermal observations,1971, Waimangu. N.Z. volcanol. Rec. 1 :34-37.

LLOYD, E.F. 1974: Hot spring monitoring,Waimangu hydrology and temperaturemeasurements (for 1972). N.Z. volcanol.Rec. 2 : 55-57.

LLOYD, E.F. 1975: Hot spring monitoring,Waimangu hydrology and temperaturemeasurements (for 1973). N.Z. volcanol.Rec. 3 : 6 1-64.

RANDS, M.B.; WILSON, S.H. 1954: A survey ofthe temperatures of the hot springs of theRotorua-Taupo Thermal Areas. Geo-thermal Chem. Rep. : Ninth Rep. (Dom.Lab.) Unpublished.

SCOTT, B.J. 1975: Hot spring monitoring,Waimangu hydrology and temperaturemeasurements (for 1974). N.Z. volcanol.Rec. 4 : 73-76.

SCOTT, B.J. 1976: Hot spring monitoring,Waimangu Thermal Area, instrumentalmonitoring of Inferno Crater and Frying PanLake (for 1975). N.Z. volcanol. Rec. 5 :55-59.



STANTON, T.P. 1978: Seismic noise char-acteristics of Inferno Crater, Waimangu.Unpublished M.Sc. Thesis, University ofAuckland. 204 p.

WARBRICK, A.P. 1934:"Adventures in Geyser-land". A.H. and A.W. Reed. Wellington.157 p.

APPENDIX

Details of Boat Construction, SafetyFeatures and Usage

(Construction details have been providedby DA. Jones, Physics Department,

University of Auckland)see Fig. 8

1. The bottom skin of the boat was cutfrom a single sheet of 2.44 m x 1.22 m x10 mm marine ply to B.S.S. 1088.

2. All joints were glued/screwed/nailedusing boil-proof resorcinol glue.

3. Two foam buoyancy "tanks" were eachcut from a single slab of high densitypolyurethane foam. To fit the side of theboat small saw cuts were made at inter-vals vertically across the mating face toallow the foam to take up the curve of thehull. It was first glued to the hull usingP.V.A. glue and then clamped to the sideswith battens and 'G' clamps. After the

REAM - HOT LAKES SURVEY 67

68 NZOI RECORDS

P.V.A. glue had cured the foam wascovered with epoxy resin and fibre-glasscloth. First a small 10 cm wide fillet ofdynel cloth was "epoxied" to the hulland to the surface of the foam at both thetop and bottom of the foam slab (dynelcloth had been chosen for its ability tomould to sharp angles). When the resinhad cured, 285 g fibre-glass cloth wasspread over the foam slab and overlappedon to the dynel cloth. A small stiff brushwas helpful in forcing the glass and epoxyinto place. A second fillet of dynel cloth10 cm wide was then epoxied over the topedge of the 285 g cloth and laid up thesides of the hull thus locking firmly the285 g cloth to the hull structure. Finallya 15 cm strip of fibre-glass cloth wasepoxied to the sides of the foam forminga protective belting to the buoyancy tank.

4. Additional buoyancy was provided byfoam slabs locked in position under theseats so that in the event of the boat fill-ing the occupants could sit on the bowand stern decking and place their feet onthe seats clear of the water.

5. On completion of the fibre-glassing theboat was painted with Everdure (a clearepoxy preservative), primer reaction lac-quer undercoat and sprayed with severalcoats of reaction lacquer. All nails werepunched down and the nail holes filledwith epoxy filler.

6. Dimensions :Sides 0.41 m highWell 0.30 m x 0.30 mFoam Slab Section 0.25 m x 0.10 mOverall Length 2.49 m

7. Carrying handles were fixed to thetransoms. They also served, together withmetal loops on the gunwales, as securingpoints for the rope tethers.

Safety features comprised the following :1. The wide shallow floor combined withlow sides ensured maximum stabilityagainst tipping of the boat.

2. All soundings and samplings were donethrough the centre-board casing, preclud-ing the necessity for any crew member to

lean over an end or side. A gantry, withthe sounding and theodolite-target gear,was fitted over the casing, i.e., in themost stable location.

3. "Tanks" of high-density polyurethanefoam along the sides and beneath theseats provided sufficient buoyancy tosupport two crew members above water-level in the event of a catastrophic flooror casing failure.

4. Nylon tethers, attached to handles onthe ends and loops on the sides, providedmeans of propulsion, positioning andholding the boat on-station, and also ofenabling rapid rescue from the shore. (Atest on a cold lake showed that the boatcould be hauled 100 metres in about45 sees.)

5. Gumboots were provided for the crew.These found normal use in launching theboat or wading in the hot lake for anyother purpose, but also would haveenabled the crew to move around in theboat had a serious leak occurred.

6. Gas-masks were provided. (In theevent these were later usually dispensedwith as no significant problem withfumes was encountered.)

7. Oars were provided as an alternativemethod of propulsion.

So far Maji Moto has been afloat onFrying Pan Lake (whose measured surfacetemperature ranged between 49.3'JC and61.8°C) for approximately 93 hours. It hasalso been on Inferno Crater lake at com-parable or lower temperatures for 142hours - including five times being mooredthere overnight - and on other occasionsat temperatures of around 74°C for a totalof 4 hours. No significant structural re-sponse has occurred and there appears tobe no reason to prevent the boat's use oneven hotter bodies of water.

It is perhaps interesting to recall (War-brick 1934) that another hot lakelet onceexisted at Waimangu which was soundedfrom a boat. This was the Waimangu Gey-ser whose site in Echo Crater is now a


sediment-filled basin lying between Infer-no Crater and Frying Pan Lake. On 10 Au-gust 1903 A.P. Warbrick and H.E. Bucke-ridge rowed a dinghy across and aroundthe pool for 18 minutes during a quiescentstage between eruptions. A maximumdepth of 48 feet was found. Water temper-ature was not recorded but Warbrick notes

that when hauling up the lines betweensoundings Buckeridge "hurriedly droppedthe rope in the boat to avoid scalding hishands*. Safety precautions were non-existent and perhaps superfluous sincethe main danger undoubtedly lay in thepossible occurrence of an untimely erupt-ion.

(Received for publication - 24 July 1978,as revised, 4 August 1980)

Documents

NZOI RECORDS - Te Ara