HUGO: Hawaii Undersea Geo-Observatory

HUGO Newsletter #3

February 9, 1998


The Hawaii Undersea Geo-Observatory (HUGO) is a system designed to supply the infrastructure necessary for scientists from around the world to conduct experiments at an active submarine volcano with real-time control and access to their data. Experiments will be installed by ROV and submersible and powered by cable from the Island of Hawaii.

The main HUGO Cable and Junction Box were successfully installed on October 11 and 12, 1997, from the R/V Independence with the help of the people and cable handling equipment from SAIC Maripro and Tyco International. The weather and ship were perfect, and the Junction Box was placed in the water at 1645 HST on October 11, was on the bottom at 2045, and the cable was passed to the shore party at Honuapo at 1115 on October 12. After laying 47 km of cable, only about 100 m of extra cable were left when the cable was passed to the shore - an excellent job by the SAIC cable laying team. All systems were functioning when the cable was terminated to pass to shore from the ship.

Junction Box Tilt during deployment. Before the anchor hit bottom the Junction Box is tilted roughly 70 during lowering. When the anchor hits bottom (20:36), the tilt of the Junction Box dropped to about 12, and when the Junction Box came to rest on the bottom (20:46) it had a 6 nose-down tilt. About 36 minutes later (21:23), the main cable became taught and tugged on the Junction Box, changing its tilt to less than 1.

In addition to orthogonal tilt sensors, the external pressure and engineering sensors were also recorded. Voltage and current supplied to the Main System Pressure Vessel (MSPV) remained constant, as did the pressure inside the MSPV. The DIGIQUARTZ pressure sensor recorded the expected increase in pressure as the Junction Box was lowered, and the internal temperature sensor recorded the appropriate temperature drop. Once the Junction Box was on the bottom, the geophone package and high-rate hydrophone (HRH, sampled at 64 KHz), were activated. Recording of data began as soon as the Junction Box was in the water and continued until the cable had to be terminated to pass to shore.

Loihi Erupting Again. Records from the HRH during deployment are particularly interesting, and sound samples will be made available through the HUGO Web page shortly. The time series is rich in transient signals with frequencies from below 10 Hz to above 1 KHz, most of which display a 2nd arrival 0.8 s after the first arrival. Transients described as "shotgun blasts, fireworks, crackles, pops, pistol shots, and roars" are common and often tens of dB above background noise. These signals are tentatively interpreted as water vapor explosions and lava movement at the surface of Loihi, and, if this is true, Loihi was erupting strongly during deployment. The gain of the HRH, sensitive to signals at frequencies from 0.01 Hz to over 20 KHz, was limited by microseism noise at 3-4 sec period, which caused data to clip at gain 64. 4.5 hours of DAT recording of the HRH were made during deployment. Other than some contamination by 60 Hz noise that disappeared after recording was moved to shore, the data quality is excellent.

The pinger/transponder was also exercised and performed nominally. It has the ability to ping on command at any frequency from 3 KHz to 20 KHz, or to transpond at 16 KHz when interrogated at 15.5 KHz. The pinger is heard easily on the HRH.

Recording of data at the Shore Station. It was not possible to arrange for the splice of the main cable to the Shore Station until about one week after the cable was passed to the shore, and data monitoring operations were resumed on October 20. Three fibers were spliced to the Shore Station, the others had been damaged at the Junction Vault near the shore during splicing. At that point, the HUGO pow er regulator was functioning, but refused to send power monitoring data back to shore. The experiment package - including the external temperature sensor, low-rate hydrophone, and three-component geophone was not operating. When power was supplied to the connector that it is plugged into, it would trip its circuit breaker. Based on this observation, we assumed that the connector to the Experiment package is flooded.

The HRH, pinger, and pressure sensor were still operational, and an additional 6 hours of DAT recording from the HRH were made on October 21 using both the high and low frequency data transmission options on one of the fibers. A particularly interesting measurement was detection of the reflection of the pinger signal from the ocean surface to aid in calibrating the depth of the Junction Box. The reflection was recorded 1.616 0.001 s after the ping, in good agreement with a depth of 1200 m for the Junction Box.

As during the deployment 9 days earlier, the observed HRH data were very rich in transient events, indicating that the eruption was continuing. A measurement of 0.8 s between the initial and secondary arrivals on nearly all events is interpreted to indicate that the sounds are all arriving from approximately the same location. If the first arrival is a direct arrival from the source (it is most often impulsive and richer in higher frequencies than the 2nd arrival), and the second arrival is the reflection off the ocean surface (it is more emergent and dispersed in time), then the time difference between these arrivals (0.8 s) allows an estimate of how far the source of these sounds is from the Junction Box depending on the depth of the source. In this way, the source (and eruption site) is determined to be between 1.3 and 2 km away, and a locus of allowable locations was determined on a map. Based on this analysis, possible sources include the upper south rift zone, portions of west pit, and areas to the north of the Junction Box. As the deeper locations are likely to have their first arrivals blocked by intervening topography or by refraction, the most likely sources are shallow, which limits probable sources to those north of the Junction Box.

Mark Rosen went to the site to terminate the remaining fibers on 11/3/97, and managed to splice two additional fibers, while breaking two of the completed splices, and, in the end, five fibers are now complete from the Shore Station to the Junction Box. OTDR recordings show that all five connected fibers are in good shape at both 1330 and 1500 microns with strong reflections from the HUGO Junction Box at 47,460 m from the Shore Station. The 6th fiber (green) cannot be spliced until the cable is re-terminated and stripped back another 3' or so. This task will be accomplished in late February.

OTDR of main fibers shows an optical loss in four of the five fibers of up to 2 dB at 1550 microns at a distance of 5050 m from the Shore Station This loss is probably caused by a sharp cable bend at that point, which lies on the steep slope of Mauna Loa as it rises towards the shore.

Of the 9 HUGO connectors, all electrical circuits appear to be operating (assuming that the circuit breakers that trip connectors 2 (experiment) and 7 (hydrophone) trip because of leaks beyond the connectors). All other connectors do not trip when 300 V is applied to the connector.

System Down. On October 23, Bob Jordan, resident HUGO Manager at the Shore Station, called to say that he could no longer bring up the HRH, and that it was also tripping its 300V circuit breaker. On November 7, the main HUGO power supply breaker at the Shore Station began tripping whenever power was supplied to the Junction Box, effectively terminating all data monitoring. Current measurements at low voltages suggests that the main cable is intact and that there is a short (through some resistance) to sea water at or near the Junction Box.

A diode in the MSPV in the Junction Box prevents the Junction Box from drawing current when the supply from shore has the wrong polarity. If the MSPV were flooded, that diode would no longer be active and current could flow equally well to the sea water short in both polarities. A test was run on November 8 to determine if the diode was active, and it was indeed found that current was allowed to flow only in one direction. Therefore, the MSPV is NOT flooded, and the leak is on the other side of the connector to the Power Regulator.

It has been determined that the system can function without power regulation at relatively low currents (few experiments), so that when the Power Regulator is unplugged the system should again function. At this point we knew of no optical data that is successfully crossing a HUGO connector. The last data seen through an optical connector link were the high-rate hydrophone data. We assumed that slow leaks are occurring around the optical lenses in the lens stand, and repeated test failures support the likelihood of this conclusion.

Junction Box Repair. Based on our findings and guesses, it was highly desirable to visit the HUGO Junction Box as soon as possible to perform the following tasks:

  1. cut off the anchor cable,
  2. adjust the attitude of the Junction Box if necessary and survey the area,
  3. remove the Power Regulator connector to see if the system begins operating again.
  4. if so, test all optical and electrical circuits in the HUGO connectors with a new test connector designed to evaluate all connector functions
  5. replace the High-Rate Hydrophone with a new unit, and return the other to the surface,
  6. deploy the sea-water return electrode
  7. return the Experiment module to the surface and replace the connector and cable to the experiment package (assuming that the connector as leaked), then return the unit on the following dive and re-install it.
  8. install a new connector designed to allow voice conversation between the submersible and HUGO Shore Station via acoustic link.
  9. attach the Power Regulator Module and connector to the Junction Box float and release the float to be picked up by the KoK.
January, 1998, PISCES V Dives. A 6-day Kaimikai-o-Kanaloa/ Pisces V dive cruise from January 15-20, 1997, funded by existing HUGO resources, was planned. Towards this end we tested methods to prevent the leaks in the connectors and built up two new hydrophones, a test connector, and experiment connector -cable. Four dives were planned to accomplish the above tasks and to attach cable protection to the main cable at "hard spots" where it shows attenuation near shore, and where it passes through very rough territory towards the north end of Loihi.

Ancillary KoK cruise tasks include deploying and monitoring sonobuoys to listen for and locate any continuing eruptive activity, and work with the FOCUS vehicle by Dr. Roy Wilkens. If time permitted, and a likely target was available, the sub would also visit the Loihi eruption site. The PISCES V was also be equipped with two external broadband hydrophones to listen for and locate eruptive sounds.

There was a strong incentive to complete this work and re-commence operations prior to a large seismic cruise on the southeast side of Hawaii by the R/V Maurice Ewing scheduled to begin in late January.

HUGO On-Line Again. The R/V Kaimikai - o- Kanaloa departed Honolulu on January 18 for four dives at Loihi. Sonobuoys dropped at the summit indicated that Loihi was very quiet, and no natural activity was heard.

The 1st dive located the HUGO Junction Box without incident at 18 54.716' N, 155 14.651'W on a flat mud plane east of East Pit, as planned. What was not anticipated was that the Junction Box was half buried in mud, within a few inches of the connector bays, which must be clear to install experiments. While the submersible found the Junction Box to be very stable in the mud, such a large cloud of mud rose every time the submersible changed position that we would have to delay work up to an hour before visibility cleared.

The Power Regulator was disconnected, and the system began operating again as hoped. The flooded hydrophone was removed and replaced with one of the new ones, which also began operating immediately, blasting the Shore Station personnel with noises of the submersible working on the Junction Box. The new hydrophone has extra gain, and it can be filtered to decrease the amplitude of the microseism peak. The two hydrophones mounted externally to the submersible were found to have too little gain to hear ambient noise.

A demodulator allowed the Shore Station to listen in the voice traffic between the submersible and the K-o-K centered at 20 KHz. The test connector designed to test the functionality of each of the Junction Box connector bays, was plugged into bay #9, and testing began.

The cable from the anchor behind the Junction Box was cut with a guillotine to allow the submersible to allow the submersible to move the Junction Box if desirable. At this point, Terry Kerby, the Pisces V pilot, found it very difficult to maneuver, and had trouble moving the sub in any direction. Rather than wait until the batteries were drained, we left bottom under full thrusters and headed slowly to the surface. The old hydrophone was indeed found to be flooded, as expected.

The 2nd dive was able to quickly home in on the Junction Box pinger. Currents on the bottom were very slow, and the sub had to wait more than an hour before work could begin. Testing of connectors on the port side of the Junction Box were done, and the Seawater Return was deployed. A second hydrophone was also installed, but a cockpit error required it to be removed again to be re-set on the ship. The Seawater Return electrode, three plates of titanium, immediately sunk in the mud about three feet. The submersible then surveyed the HUGO cable towards the north.

The 3rd dive recovered the experiment package and seismometer to replace the connector. Terry Kerby had to dig into the mud to recover this equipment, as there was no trace of it above the mud line. Without force feedback, this was a heroic operation, but the equipment was recovered with negligible damage. The cable and connector to the experiment package were replaced and the 2nd hydrophone was reset, but we could not get the experiment to work on the ship, and gave up in time to get a bit of sleep before the last dive. A5 foot long rod was inserted into the mud to test the depth of mud under the Junction Box. The rod met no resistance for its full length, thus the mud thickness is more than 5'.

The 4th dive reinstalled the 2nd hydrophone, which still refused to work, and tested the remaining connector bays. The remainder of the dive was spent surveying the cable route out to two km north of the Junction Box. The cable is resting comfortably on the bottom, or slightly buried for most of the route surveyed, but in at least three areas, a span of more than 50 m is found where the cable crosses a crevice or valley. The ends of the span terminate in mud, and appear to be quite safe, although the submersible will need to take great care when approaching these areas.

Current Status.

  • Data recording: We are currently continuously recording the hydrophone at a rate of 500 samples/s while the M. Ewing shoots reflection profiles in the area. They have made three passes directly over the HUGO Junction Box, and we have recorded their outgoing pulse at normal incidence at 4,000 samples/s. In addition, pressure and tilt are recorded once per second. We expect to telemeter the hydrophone data to the Hawaiian Volcano Observatory later this month for inclusion with their seismic data.

  • Eruption: Loihi continues to be noisy with explosive and popping noises nearly continuously, we thus believe that an eruption is occurring in the summit area. One type of event begins with a low frequency roar followed by a high frequency hiss lasting for several minutes. This type of event is repeated several times per day. Nearly all the transient events from Loihi have a sharp arrival followed 0.8 s later by a more dispersed arrival. We suspect that these are the direct and surface-reflected arrivals, and the 0.8 s delay suggests that the source is between 1.5 and 2 km from the Junction Box. The summit was active during each pass of the Ewing, and we expect that the records from their 4 km long streamer should contain several Loihi events which can be used to locate the source of these events.

  • System: HUGO continues to operate well, although we have lost the data link for some of the engineering data from the system for an unknown reason. Of the nine Junction Box connectors available for experiments, four are known to be operating, three have questionable test results, and two could not be tested because they are blocked by the chain holding the float to the Junction Box. If only the connectors we know now are operating were available to support experiments, HUGO could still support 21 experiments, in addition to the high rate hydrophone. If the bandwidth taken by the hydrophone were split to support other experiments, and additional 49 experiments could be supported.

    Future of HUGO. Operation of HUGO will continue indefinitely. The highest priority tasks for the HUGO team are:

    1. Direct data transmission to the Hawaiian Volcano Observatory. We will establish a RF link to HVO to make it possible to record and store the Loihi hydrophone data as part of the HVO seismic network, as well as to monitor the higher frequency acoustic signals.
    2. Direct data transmission to the Manoa campus. We will establish a frame-relay link to the Manoa campus of the University of Hawaii so that the system can be controlled remotely without need of an on-site operator.
    3. Stabilize the Junction Box. The burial of the Junction Box about one foot in the mud is troublesome. Should it sink another four inches, it will make it impossible to make new connections. We are considering different options for raising and stabilizing the Junction Box to prevent further sinking. Until that is done, we do not believe it wise to remove the Junction Box float, which supplies about 1,000 lbs of buoyancy to the package. We may decide to move the Junction Box onto a flat plate, jack up the corners of the Junction Box, or strap the float down using for-aft straps, allowing us to remove the chain blocking two of the connector bays. At least one PISCES V dive will be available during the September, 1998 dive series.
    4. Install a 2nd hydrophone and seismic experiment. These systems, which could not be installed during the January dives, are now functional, and will be installed in September, 1998.
    5. Construct and emplace a Level-2 Multiplexer Node. This system will be installed at some distance from the Junction Box, and will allow an additional eight experiments to be installed. While there is pressure to install this node in or near Pele's Pit, the logistics and distance to the pit are formidable enough that we will try an easier installation first, probably at the south side of the Thousand Fingers Field.

    Call for Experiments. We are now confident enough that HUGO will be around for several more years, that we encourage investigators to consider constructing experiments for installation in HUGO. The main constraints are less than 10 Watts of electrical power and no more than 2048 samples per second. Please see the HUGO Experimenter's Guide for details. Please write for the latest copy if you're interested.

    Acknowledgments: I wish to thank the more than 100 individuals who were responsible for the successful emplacement of the HUGO Junction Box, Cable, and Shore Station, and with the first service dives, which got the system operating again. This was a team effort in every sense, and it has been a true joy working with so many competent and patient people.

    MORE about Loihi and HUGO can be found at the following WEB sites:

    Fred Duennebier
    Dept. of Geology and Geophysics
    University of Hawaii at Manoa
    Honolulu, HI 96822

    Fred Duennebier HUGO Newsletter #3 2/9/98 8

    This page created and maintained by Jackie Caplan-Auerbach

    Last updated Monday, Feb 18, 1998.