Temperature Observations of Crater Lake, Mt Ruapehu, New Zealand, Using Temperathlre Telemetry Bouys A.W. HURqT

Geophysics Division, Department of Scie,di/ic and !,,dustrial "lesearch, P.O Boa. 1320. Wolli,~gton (ig~,w goaland)

ABSTRACT The temperature of the Crater Lake of I;he active volcano Ruapehu has been recorded by Temperature Telemetry Buoys, to determine if lake temperature is correlated with volcanic activity. These buoys had to be specially designed to cope with the mffa~,)ureble eov;ro,~merJt of Crater Lake. A buoy contains a thermistor to measure the lake temperature, and a radio transmitter to transmit a short signal every few minutes, the interval between signals being a functim~ of temperature. The temperatme records obtained from these buoys show that the temperature near the lake surface can vary considerably within a few hours. Some of these variations appear to be caused by disturbances in convective heat transfer occurring in the lake. The occurrence of these short, term temperature variations means that there is no simple relation between Crater Lake temperatures and the volcanic activity of Ruapehu. Some rapid increases in temperature followed volcanic earthquakes, but one of the biggest increases in temperatxtre occurred just b~fore a group of ,,arLhquakes vr,der Lbe lake. INTRODUCTION R u a p e h u is an active avdesitic strato volcano, 2.797 m in height, located near the centre of North Island, New Z~aland (Lat 39°17'S, Long 175°34'E, Cat. o[ Active Volcanoes, 22, 1-10. Co[.~: and NAm.~, 1975). Its active vent is occupied by Crater

Bull. Volcanol. Vc.l. 43-1. 1980.

Lake, which has a current overflow level of 2,531 m. Since 1966 there have b e e n a n u m b e r of ash eruptions, two of which sent lahars down the valleys which radiate from R u a p e b u summit. Phreatic and hydrot h e r m a l eruptions have occurred frequently. Crater Lake is surrounded by perm a n e n t snowfields, which feed several small glaciers through gaps in the, broken ring of peaks around the lake. Wig. 1). All of the heat released by R u a p e h n goes into the lake. T h e t e m p e r a t u r e of the lake is, therefore, a n indicator of the heat output a n d level of activity of Ruap~hu (D1B3LE, 1974). T h e m e a s u r e m e n t of the C,-ater Lake t e m p e r a t u r e has long b e e n a n i m p o r t a n t part of the D e p a r t m e n t of Scientific a n d Industrial Research (D.S.I.R.) program of monitoring Ruapehu. For abeut 12 years temperature m e a s u r e m e n t s hsve b e e n m a d e at approximatnly m o n t h l y intervals. T h e visits to m a k e these m e a s t t r e m e n t s require a considerable a m o u n t of effort, a n d involve a certain a m o u n t of danger, especially when h y d r o t h e r m a l eruptions are frequent. R e m o t e monitoring of the Crnter Lake t e m p e r a t u r e , either continuously or at freq u e n t i n t e r v a l s , can give far m o r e frequent information on the Crater Lake temperature t h a n cotfld ever be obtained by m a n u a l methods. Such m o n i t o r i n g can also reduce the n u m b e r of visits which m u s t be m a d e to the lake, although visits m e still I,eeded to get w~ter s,qmples fo: chemioal analys;s.

122

A.W, H U R S T

•

FIG. 1 - View of Crater Lake, Ruapehu, in summer, looking north. The lake is about 500 m across. TELEMETRy BUOY SYSTEM

Working Environment The environment in and around Crater Lake is very hostile to any measuring equipment. The lake is very acid (pH generally in range 0.7-1.5, GIGGENBACHand GLOVER, 1975), sometimes hot, and prone to eruptions. The slopes around the lake are covered with unstable snow and ice, which gradually move towards the lake. The temperature on these slopes can be as low as:--20°C, with strong winds and icing conditions. These conditions would preclude survival of a fixed installation in the crater area, and ruled out signal transmission by wires to the lake edge. However, a selfcontained instrument such as a radiotransmitting buoy, moored in the lake well away from the shore, would avoid ice

falls into the lake and would not be damaged by eruption surges. To withstand the acid, the entire outside case had to be made of PVC, which is resistant to strong acids up to at least 60°C, a temperature only likely to be exceeded during or immediately after eruptions. The warmth of the lake meant that icing of the aerial should not be a problem. Because the buoy would not survive large eruptions it had to be regarded as expendable, so cheapness and ease of construction were important~

System Design The measurement of temperature can be done with a thermistor to the required accuracy of I°C. A thermocouple is not suitable because no cold junction is possible in a hot |ake. Platinum resistance or

TEMPERATURE OBSERVATIONS OF CRATER LAKE. MT. R U ~ E H U , ETC.

quartz thermometers are more expensive and are not justified for this application_ A self-contained buoy must operate from primary batteries. (Other energy sources, including solar power, thermoelectric power, and batteries with the lake as electrolyte had their attractions, but were not practicable). For a reasonable weight and cost of batteries, it is necessary for the transmitter to be off most of the time, to get aa adequate battery life. The small dutycycle was combined with the method of encoding the information by producing a short ON pulse with a repetition rate which increased with temperature. Since the thermistor is inside the buoy, and therefore has a response time of about 5 minutes to temperature changes of the lake, very little information is lost by only getting readings every 10 minutes or so. The radio link from the buoy is the most critical part of the system. Normally, telemetry links operate at VHF or UHF, but these require line of sight transmission paths. As Crater Lake is in a depression, such a link would need a repeater on one of the peaks or ridges surrounding the lake. The effects of icing and wind ruled this out. Skywave propagation using the ionosphere as a reflector was not possible because a short aerial just above the lake could not be made to radiate upwards. However, if frequencies in the 1.6-5 MHz band are used, their long wavelengths (60-190 In) make the radio waves diffract around peaks and ridges, so a signal from a buoy in Crater Lake can be received at points which are out of sight of the lake. The radio signal was received at the Glacier Shelter, 1 km north of Crater Lake. The audio output from the receiver at the Glacier Shelter is taken via telephone lines to the DSIR Chateau Observatory, where it is recorded. BUOY DETAILS A detailed description of the buoy, including circuit diagrams, can be found in THOMPSON and HURST (1977).

123

Electronics As already mentioned, the temperature sensing element is a thermistor. The thermistor is the timing resister of a multivibrator which uses 2 CMOS inverters as its active elements. This multivibrator drives a 14 stage CMOS divider. When the 14th stage output goes high, the crystal oscillator of the transmitter is turned on, as is a second CMOS multivibrator. After 2 sec, the divider circuit is reset to zero and the oscillator is turned off. The off time typically varies from 16 minutes, for a temperature of 20°C, to 7 minutes for a temperature of 50°C. The second multi~'ibrator oscillates at an audio frequency, and switches the driver stage on and off to modulate the transmitter. The transmitter has a power output of 5 W, at a frequency of 2.452 MHz, into a whip aerial. Because of the corrosive nature of the lake, a capacitive connection to earth is made. A thin sheet of copper lining the inside of the buoy is capacitively coupled to the highly conductive lake which acts as a ground plane. The transmitter is powered by 24 Alkaline D cells, in series-parallel to provide 18V, while the logic circuits are powered by mercury cells providing 5.4V. Temperatures above 60°C, or sustained temperatures above 50°C are liable to damage the batteries.

Mechanical The main case of the buoy is a length of ordinary 100 m m diameter PVC pipe, and standard PVC fittings are used as much as possible (See Fig. 2). The bottom of the pipe is closed off by an end cap, and the thermistor is attached to the inside of the end cap. The copper sheet for the capacitive earth is immediately inside the PVC pipe, and surrounds the electronic components and the battery pack, both of which are enclosed in a potting compound. The helical whip aerial is supported at the top Of the pipe by two spacers. The top spacer uses a silicone compound to give a watertight seal that still allows some flexibility in the aerial mounting.

124

A.W.mmST

Near the top of the pipe is a buoyancy chamber, of polystyrene foam enclosed by P V C sheet, which provides extra buoyancy. The concentration of buoyancy makes the buoy float with a constant waterline. This is necessary because the aerial tuning depends on its height above the lake surface. A polypropylene rope joins the buoy to an anchoring rock, of about 20 kg. T h e buoy, with rock attached, is lowered into Crater Lake by helicopter,using a second rope attached to the buoy. Once the buoy is in the lake, the rock should ground on the lake bottom to prevent tb_ebuoy being washed ashore. OPERATIONAL

atures as high as 70°(] for a few minutes. T h e gaps in the record m a y have been due to the buoy being in a part of the lake from which radio signals could not be received, or the batteries m a y have been temporarily affected by high temperatares.W h e n signals were next received on

Aerial 1430mm

HISTORY

After initialfailures, a buoy (identified as Buoy IV) was put into Crater Lake on 1976 March 9, and operated successfully for over two months. Its eventual failure was probably due to ice damaging the aerial.Part of the record from this buoy, and the record from Buoy VI, are shown in Fig. 3. (Buoy V gave no useful results because of an electricalfault).Buoy VI bad a very short life,as it was put into Crater Lake only a week before an ash eruption on 1977 November 2. (It is fairly likely that the cessation of signals from this buoy was the result of some hydrothermal activity before the eruption). T h e sharp decreases in temperature re- 1120mm corded by Buoy IV on 1976 March 28 and 1976 April 29 (also by Buoy VIII on 1978 November 11) show the effectsof cold water entering the lake during periods of heavy rain and snowfall.Apart from fairly Battery rapid temperature increases after these Pack decreases, the only rapid increase in temperature observed was a few days before the 1977 November 2 eruption. (Fig. 3) Buoy VII, launched on 1978 M a y 30, showed a very differentpattern. O n June 12, the temperature increased by over 6°C, and over the next few days the tem0 500mm perature was consistently over 50°C, with considerable short term variations. T h e highest temperature recorded was 54°(], Scale although on June 17 some weak signals FIG. 2 Cutaway diagram of Temperature from the buoy seemed to indicate temperTelemetry Buoy. -

TE~IPERATURE OBSERVATIONS OF CRATER LAKE, MT. BU.~.P~HU, ETC.

Juixe 18, the temperat~ire was about 48°{2, and it generally decreased after then until the buoy was last heard from on June 25. (By July 18, the temperature had decreased to 30°C at the outlet). During the time ~f these Mgh temperatures, small hydrothermal eruptions were seen to be occurring in th~ lake. P.M. OTWAY (pers. conun.) reported that these consisted of a sudden updorahfig of muddy w~ter across an area of about 100 m 2 to a height of 1 to 5 metres, virtually obsctaed by the release of steam, and accompanied by a ~whooshing~ sound. He measured a lake outlet temperature of 40.8°C (see Fig. 3) at 0030 U.T. on 1978 June 13. This was 3 hrs "after a buoy temperature of 47.7°(] was recorded, showing that the temperature increase took some time to affect the outlet temperature. (The outlet temperature is normally about 2-3°C less than the buoy temperature.) On 1978 October 5, Buoy VIII was launched. It transmitted a strong signal until 1978 November 23, when the signal ceased abruptly. On 1978 November 29 the buoy was seen to be floating on its side in the lake. On 1979 January 11, the buoy was pulled upright by a rope from a helicopter, and signals were heard for about a day. It is thought that the anchor was sinking in the soft sediment at the bottom of the lake, and pulling the buoy under. The record from this buoy (Fig. 3) shows abrupt increases in temperature, including an increase of 8°C in less than an hour on 1978 Oct 31. This increase and some other increases immediately followed B-type earthquakes under the Crater Lake (J.H. LATTER, pers. comm.) and was assumed to be the result of eruptions through the lake, although no visual observations exist to confKrm this. INTERPRETATION OF RESULTS

Physical Properties o/ Crater Lake Crater Lake is an oval shape, roughly 400 by 550 m: and has a surface area of about 200 000 square metres. A bathymetric survey in 1965 (DIBBLE, 1974) showed that Crater Lake had a deep vent area

125

near the centre, with a maximlwn depth of about 300 m. About 10% of the eu'ea of the lake was more than 100 m deep. At 40°(], a typical lake surface temperature, the heat loss to the atmosphere is about 300 MW (HtIRST and DIBBLE, in prep.). This is at least a thousand times more than could be supported by conduction of heat through the lake, so some form of convection must be occurring. Analysis of temperature and density versus depth measurement made by R.R. Dibble in 1965 and 1966 (HI~F~ST and DIBBLE, in prep.), showed tlmt they were consistent with very turbulent thermal convection throughout Crater Lake. This corrected an earlier study (DIBBLE, 1974), which concluded that thermal convection was not occurring. The temperature gradient within a sti'ongty convecting fluid is normaliy small, with the temperature change concentrated near the boundary layers (TtmNER, 1973). C~ater Lake showed a different pattern, with sharp increases in temperature alternating with regions of nearly constant temperature (fig. 27, DIBBLE, 1974; fig. 6, HURST and DIBBLE, in prep,), and it is assumed that it was convecting in more or less transitory cells. These ceils may have existed because of the shape of Crater Lake, or they may have been an effect of the sediment density gradient in the lake (HURST and DIBBLE, in prep.). In 1970, Crater Lake was resurveyed (IRWlN, 1972), and the greatest depth was found to be only about 80 m. Two mare a o tive vent areas were identifiable from echosounder records by the concentrations of gas bubbles rising from them (DIBBLE and HURST, in prep.). Convection from small sources such as these vents can be treated as a convective plume (TURNER, 1973). In particular, BAINES and TURNER (1969) discussed convection from a source in a confined region. They showed that the effect of a convective plume was to heat the lake nonuniformly, with the lake surface being heated first, and the increase in temperature then moving down into the lake. This analysis was only valid for shallow lakes, with depth less than the surface radius. Deep lakes behaved differently, with a general harbulent circulation in the top part of the lake.

126

A.W. ~IuasT

Crater Lake Records from Temperature Telemetry Buoys 35 34 33 32 31 30 29 28 27 26 25

1976 Buoy IV March-May 1976

i

i

a

24 March

I

i

26

&

i

28

i

.

30

1

,

1 April

,

m

3

|

I

5

I

!

7

I

!

I

9

i

i

i

13

11

35 34 33 p 32

3~:I 33 I.

31

i

i

15

i

i

17

i

i

19

i

21

~ [ ' ~ 1977 Buoy VI #~jv ,1.

3O E 28 ~ 2~

23

54 53 52 51 50 49 48 47 46 45 44 43 42 41 40

25

27

29

1 May

3

5

7

r27,v.29. . . 31

9

.,, 2 November

October

1978 Buoy VII ~ / ~ ' ~

~

o !

i

30 May

I

i

1 June

t

m

3

i

#

5

i

m

7

m

i

9

i

I

11

I

I

13

I

I

15

I

I

17

i

i

19

|

I

21

~

i

i

51

i

23

Date(t~

FIG. 3 - Temperatures from Telemetry Buoy and measured Crater Lake Outlet Temperature, and Earthquake activitynear Crater Lake. (M is average magnitude of earthquakes.) B-type earthquakes are low-frequency volcanic earthquakes, Roof-Rock earthquakes are high-frequency earthquakes less than 0.5 k m deep under Crater Lake (LATTER, 1979).

127

TEMPERATIPRE OBSERVATIONS OF C R A T E R LAKE, Mt, RUAPFStTJ, ETC.

and snowstorms (e.g. 1976 March 28-29, April 29-30 and 1978 November 11-12. See Fig. 3). Although cold fresh water has a density less than the lake water, convection was effective in lowering the temperatare at 1 m depth by several degrees Celsius. This is probably because the fresh water would tend to be mixed with the denser lake water by the turbulence associated with the convection in Crater Lake. Comparison of the estimated annual precipitation with calculated cold water inflows to Crater Lake suggested that about half of the precipitation actually mixed with the lake water CLIURST and DIBBLE, in prep.). If 50% of the precipitation is as-

No bathymetric surveys of Crater Lake have been done since 1970, so its present maximum depth is unknown. Neither have any temperatures at depth been measured since 1966. However, some information on the heat transfer mechanisms within Crater Lake can be deduced from the telemetry buoy records.

Lake Temperature Variations The telemetry buoy record of temperatures at 1 m depth showed rapid decreases in temperature when large quantities of cold water entered the lake during rain 42 41

40 39 38

37 36 35

1978 Buoy VIII

34

33 32 31 30 29 28

27 26

o

I

I

I

xo

o

~

I

7

o•~ I

9

I

I

11

I

I

13

I

I

15

I

I

17

I

I

19

I

1231

21

^~_~

o

I

25

I

I

27

I

1291

I

I

31

October

I I 2 November

Q.

E =

KEY EARTHQUAKES

40

39

1978 Buoy VIII continued

38

~. All M< 2.5 Roof-RockType

37 36 35

o M< 2.5

J

34

TEMPERATURES

33

x Outlet Temp. Measurement Buoy 2-Hourly Aw Temp.

32 31 30 29 28

27 26'

"1

O 2,5~ M<2-9 ~ B-Type • M~2.2

© ~

4

'

November

, 6

q

i 8

i

i 10

i 12

~

i 14

!

i 16

i

i 18

Date (UT)

FIG. 3 (cont).

i

20

i

1221

128

A.W.HURST

sumed to mi~ then the temperature changes recorded in the three storms mentioned above are about the values expected, if the cooling effect was evenly distributed through the top 1 m of the lake, at the time the minimum temperature was recorded. The temperature record of Buoy VII showed a very sharp increase in temperature during 1978 June 12 and 13. If the whole lake had increased in temperature, this could have been caused by the heat entering the lake rapidly increasing from about 300 MW to 3000 MW, but this seems unlikely for several reasons. There was no volcanic tremor or shallow earthquake which might be expected to accompany such a change in heat output. (An earthquake did occur within 1 km of the lake at 0135 U.T. on 1978 June 13, but the lake temperature had already risen to nearly 50°C by this time). T h e chloride ion concentration did not increase as much as one would expect if such a massive injection of steam occurred, nor did the magnesium concentration rise, as might have been expected if lava had been injected into the lake (See GIGGENBACH and GLOVER, 1975, for discussion of heat and chemical relationships). Also, the drop in temperature about June 17 was too rapid to be reasonably explained if the whole volume of the lake had increased in temperature. An alternative explanation is that a comparatively small change in the heat source at the bottom of Crater Lake changed the mode of heat transfer. With convection through the whole lake, as observed in 1966 (HURST and DmBLE, in prep.), the temperature in Crater Lake increased with depth. On the other hand, BAINES and TURNER (1969) showed that a convective plume from a small source would heat the top of the lake first. The temperature rise on 1978 June 12 and 13 could be explained by the development of a plume, which rapidly heated the top layers of the lake. The small hydrothermal eruptions which occurred frequently during the period 1978 June 12 to 16 (P.M. OTWAY, pets. comm.), were probably caused by the expansion of bubbles of steam rising in the plume. T h e momentum of this steam as it entered the lake may

have been a significant factor in producing the convective plume (The possibility of steam driving Crater Lake convection was suggested by DmBLE, 1974). In contrast to the results from Buoy VII, two sudden increases in temperature recorded by Buoy VIII in 1978 October-November immediately followed B-type earthquakes, which often accompany eruptions. It is therefore probable that eruptions occurred at those times due to interaction of the lake with a heat source below and that this was the cause of the temperature increases. T h e large daily variations on the record from this buoy probably reflect the cooling effect of meltwater entering the lake during the daily thaw. CONCLUSIONS A temperature telemetry buoy can give useful records of lake temperatures, of adequate accuracy for most purposes. The length of record obtained from such buoys depends mainly on how often eruptions occur in the lake, although other environmental problems, such as ice forming on the aerial, can also give trouble. The usefulness of temperature telemetry buoys for predicting eruptions of Ruapehu, is not yet established, because not enough is yet known of the temperature changes before eruptions. Some previous eruptions of Ruapehu have followed months of lake heating (e.g. 1971 April and May eruptions), others such as 1975 April 24, have come from a lake which was comparatively cool a few days before the eruption. (HURST and DmBLE, in prep. has a graph of Crater Lake Temperature and eruptions since 1966). The changes in convective regime in the lake also produce tempemixtre changes which m a y be hard to distinguish from temperature changes due to changing heat flow into the bottom of Crater Lake. ACKNOWLEDGEMENTS T h e original suggestion which Ied to this project was made by W.I. Reflly, and

T E M P E R A T U R E OBSERVATIONS OF CRATER LAKE, MT. RUAPEHU, ETC.

helpful discussions with W. Ireland indicated that an H.F. radio buoy was practicable. I would like to t h a n k all the persons who assisted the construction, testing and operation of the buoys, especially G.K. Sorrell, P.C. Whiteford and the helicopter pilots who put the buoys into Crater Lake, a n d P.M. Otway for his outlet temperature measurements. J.H. Latter provided the details of earthquakes shown in Fig. 3. REFERENCES BAINES, W.D. and TURNER, J.S., 1969, Turbulent Buoyant Convection from a Source in a Confined Region. J. Fluid Mech, 37.

p. 51-80COLE, J. W. and NAmN, I. A,, 1975, Catalogue of the Active Volcanoes of the World Including Solfatara Fields. Part XXII New Zealand. IAVCEI, Rome, 156 pp. DIBBLE, R. R., 1974, Volcanic Seismology and Accompanying Activity of Ruapehu Volcano, New Zealand. In L. Civetta, P. Gasparini, G. Luongo and A. Rapolla (Editors). Physical Volcanology. Elsevier, Amsterdam. p. 49-85.

129

HURST, A.W. and DIBBLE, R.R., in prep., Bathymetry, Heat Output and Convection in Ruapehu Crater Lake, New Zealand. Sub-

mitred to Jour. Volc. and Geoth. Res. GIGOENBACH, W.F. and GLOVER, R,B., 1975, The Use of Chemical Indicators in the Surveillance of Volcanic Activity Affecting the Crater Lake on Mt Ruapehu, New Zealand.

Bull Volcanol., 39, p. 70-81. IRWIN,J., 1972, New Zealand Lakes Bathymetric Surveys 1965-1970. NZOI Records 1, No. 6. p. 107-126, New Zealand Oceanographic Institute. Wellington, New Zealand. LATT~,R, J. H., 1979. Volcanological Observations of Tongariro National Park. 2. Types and classification of volcanic earthquakes, 1976-1978. Report 150, Geophysics Div.

D.S.I.R., Wellington, New Zealand. THOMPSON, G.E.K. and HURST, A.W., 1977, Temperature Monitoring at Crater Lake, Mt Ruapehu. Technical Note 68, Geophysics

Div. D.S.I.R., Wellington, New Zealand, 28 PP. TURNER, J. S., 1973, Buoyancy Effects in Fluids. Cambridge University Press. Cambridge, 367 pp. Ms. received March 1979; reviewed May 1979. Revised ms. received Sept. 1979