UNITED STATES
DEPARTMENT OF THE INTERIOR
GEOLOGICAL
SURVEY
PRELIMINARY STUDIES OF SEISMICITY,
GROUND DEFORMATION, AND GEOLOGY,
COMMONWEALTH OF THE NORTHERN MARIANA ISLANDS,
MAY 20 TO JUNE 8, 2001
by Frank A. Trusdell1, Maurice K. Sako1,
Richard B. Moore2, Robert Y. Koyanagi1 and Steve Schilling3
U.S. Geological Survey
Prepared for the Office of the Governor, the Emergency Management Office,
and the Office of the Mayor of the Northern Islands,
Commonwealth of the Northern Mariana Islands
1Hawaiian
Volcano Observatory; P.O. Box 51;HNP, HI 96718-0051
2Alaska Volcano Observatory; 4200 University Drive; Anchorage, AK 99508
3Cascades
Volcano Observatory, 5400 MacArthur Blvd, Vancouver, WA 98661
June 7, 2001
The Northern Islands municipal government has developed a plan for resettlement of the Northern Mariana Islands (figure 1), with limited economic development of agricultural, fishing, and visitor resources. Critical steps for successful implementation of the plan include the identification of fresh water sources, an assessment of volcanic hazards, and a survey of native and invasive species on the island. We proposed to evaluate the sustainability of island resources and assess the hazard from future volcanic eruptions. This work would demonstrate a commitment to Department of the Interior goals and the Commonwealth of the Northern Mariana Islands (CNMI) government in its resettlement efforts and also show a constructive response to the request for assistance.
Figure1.
The residents of Pagan
Island (figs. 1 and 2), located at 18° 6’ N, 145° 45’ E in the CNMI were evacuated in May 1981 following a
volcanic explosion that covered about 20% of North Pagan with various volcanic
deposits. Most of the former
residents would like to return to their island but the infrastructure of the
island was destroyed during the 1981 and subsequent eruptions. Presently,
Pagan is uninhabited. Resettlement of the island will pose many challenges to the
CNMI government, and a request has been sent for US Geological Survey assistance
with an assessment of volcanic hazards and natural resources.
Figure 2: Map of Pagan Island with monitoring sites.
The
overall goals of this study are to provide information to the CNMI government to
help them assess the volcanic safety of the northern islands.
Furthermore, an integrated science approach including specialists in
volcanology, hydrology and biology (between Geologic Division, Water Resources
Division and Biologic Resources Division) was embarked with a primary focus on
Pagan island to assess water resources, volcanic hazards and biologic resources
of the island.
Volcano monitoring involves a variety of measurements and observations designed to detect changes at the surface of a volcano that reflect increasing pressure and stresses caused by the movement of magma within or beneath it. An eruption occurs when magma rises from its source or from a storage reservoir and finally reaches the Earth's surface. The rising magma causes earthquakes and deformation of the volcano.
The long-term objective of this work is to conduct field surveys and research to determine the seismic and volcanic hazards in the Northern Mariana Islands. Specific goals are to reestablish instrumental monitoring to enhance eruption forecasting, encourage public awareness and promote educating earth science students, and provide information to CNMI as they develop a strategy for safe resettlement of residents in the northern part of the island. The goal of this year’s activity focused on three major areas: geologic mapping with emphasis on volcanic hazards, deformation monitoring, and seismic monitoring.
Seismic Surveys
Seismic instrumentation is needed for continuous real-time detection of eruption-related earthquakes, volcanic tremor and strong tectonic earthquakes. A seismic network was established with the help of the USGS’s Hawaiian Volcano Observatory in the 1980s and 1990s. Recently we assisted CNMI personnel in repairing existing seismic stations. As part of a resettlement effort to ensure public safety, each inhabited island should have a continuous recording seismometer in order to detect small earthquakes that might signify volcanic unrest.
The continuous instrumental monitoring of local seismicity from the CNMI radio-telemetered seismic array can begin to show trends of activity related to volcanism. With many years of systematic seismic data collection, the database becomes increasingly effective for scientific evaluation of volcanic and seismic hazards. It will provide more detail for the short-term (monthly) events, as well as provide researchers with a long-term perspective to observe patterns in seismicity from the long-term (yearly) background levels of activity.
Seismic Summary
Seismogram
sections recorded at Pagan, on a portable seismographic equipment (PS-2), May 24
to 27, 2001 indicate varying tremor patterns at low to moderate amplitude,
relative to past activity. The section between 0900 hrs. to about 1600 hrs on
May 24, shows a moderate swarm of
L-P events. After 1600 hrs, May 24,
low background tremor persists, with occasional higher amplitude bursts. The
record shows weak banding of tremor at intervals of about 2 to 2.5 hours.
Low-level
banded tremor became more obvious when we set instrument decibel levels to 24-dB
gain on May 26, 2001. The pattern
looks similar to those observed before at Pagan (and Kilauea, Hawaii).
Mt. Pagan’s internal “plumbing” system seems to be repeatedly
pressurizing to a critical point (low tremor amplitude), and then relieving
rapidly (high tremor amplitude). Pagan still appears to remain lively.
Preliminary
interpretation of the Pagan PS-2 seismogram sections for May 27-30, 2001:
Seismic
records show continuing tremor of variable amplitude and temporal pattern
indicative of subsurface volcanic activity. Swarms of moderate amplitude L-P
(long-period) events (each about a minute in duration and dominant frequency of
1-2 Hertz) on May 29 to 30, appear comparable to earlier episode recorded on May
24, 2001.
There
is no significant number of short period microearthquakes associated with new
intrusion. Approximately 95 small tectonic earthquakes were recorded on
seismogram sections from May 23 to June 3, 2001.
Deformation Surveys
Geodetic networks are set up to measure the changing shape of the volcano surface caused by the pressure of magma moving underground. Techniques we have used in the past at Pagan include electronic distance measurement (EDM) surveys and measurements of tilt. Commonly the ground moves upward and outward above a magma storage volume before an eruption. During a survey in 1994, while Mt. Pagan erupted, volcanologists were able to document seismic and ground deformation associated with its volcanic activity. We have repeated the measurements at the site of previous surveys to determine changes and assess possible unrest at the volcano. The locations of the glass reflectors and the EDM instrument are shown in figure 2.
Lines on three young volcanoes and Saipan were resurveyed for the first time since 1994 (Sako and others, 1994). All results are preliminary until we are able to correct for barometric pressure and air temperatures.
The EDM permanent-glass network was cleared of obstacles and remeasured. The results of the remeasurement of four EDM lines showed no significant changes when compared to the survey of May 1994. Line length changes are all less than one centimeter (within the error of the instrument's precision) and showed minor extension.
The original Billygoat reflector site was recovered, after being buried by about 2 m of ash, by the 1994 eruptive episode of Mt. Pagan. Unfortunately, the reset reflector site is now destroyed. The NOSE reflector site shortened about 80 mm when compared to the last measurement taken in the 1994 survey. No significant changes occurred during our 12 days stay on Pagan. Between May 29 and June 3, there appears to be a small inflationary trend due to the shortening of line lengths on reflector sites located on the upper flanks of the volcano.
Anatahan
Line lengths on the Anatahan EDM network were measured and showed no significant changes. Most line lengths exhibit small contractions when we compared the data to the 1994 survey.
The calibration line on Saipan was measured before and after our work in the Northern islands. All measurements were within analytical uncertainty and therefore the instrument remains stable.
1) Deformation related to Agrihan seems to be relatively quiet.
2) Deformation on Mt. Pagan seems to be more active but still on a small scale. In order for deformation monitoring to be effective as a tool for evaluation of eruptive condition measurements should be made frequently.
3) Deformation on Anatahan appears to be slowing down with no significant changes. It is recommended that cement be poured around the reflector rebars so goats do not damage them.
Geologic mapping
and volcanic hazards assessment:
The best way to assess the hazards of recurrent eruptions is to determine the recent eruptive history of a Mariana volcano. To improve our understanding and extend it back in time to provide a realistic basis for assessing future activity, a detailed temporal and spatial record of eruptive activity should be made by mapping significant surface eruptive units. The age of each unit will be determined by radiocarbon dating in conjunction with stratigraphic position. All new mapping and flow attributes will be combined with those obtained during earlier surveys and compiled into a Geographic Information System to provide a quantitative analyses of the distribution and variation of volcanic eruptive products over time. These analyses will be needed to refine the volcanic hazard assessment for CNMI.
Geologic
Observations:
Islands are listed and described from north to south. Most of the islands in the chain were viewed only from the air, and new geologic fieldwork was carried out only on Pagan Island.
Uracas (also known as Farallon de Pajaros or Uracus)
Uracas was viewed from a fixed-wing airplane on May 31, 2001, and appeared little changed since 1994. No new lava or ash was observed, and a small amount of steam was being emitted from the north rim of the volcano’s crater.
We saw no sign of the submarine volcano that was erupting in April 1994 (Sako and others, 1994) south of Uracas. Closer to Maug, we viewed Supply Reef, a submarine volcano that reaches to within a few meters of the ocean surface. It is marked by lighter green water and crashing waves.
Maug appeared to have changed little in the past seven years. There was no evidence of a recent grass fire that had been reported to the Emergency Management Office.
The amount of water at the surface of the boiling hot spring at the A.D. 1917, vent on the caldera floor continued to decrease since 1994, to the point where it can only be heard, not seen. This development is not unusual for a geothermal area, however, for deposition of mud, clay, and sulfur often effectively clog such systems and drive the water elsewhere. The highest fumarole temperature measured was 98.1oC, which is boiling for the elevation and the same as we have reported previously.
The attached table 1 summarizes the current but incomplete understanding of the timing of eruptions in the past 350 yr, (from Banks et al., J. Volcanol. Geother. Research, 22:225-269).
The bulk of this year’s efforts were to improve our understanding of Pagan Island. We collected charcoal for radiometric dating of eruptions. We also collected rocks for chemical analyses, for the composition of lava can be a major factor in its explosivity. Analyses of these and rock samples commonly requires several months and the results are a key to deciphering the frequency of eruptions and understanding their attendant volcanic hazards. We collected fifteen new charcoal samples that will augment our previous dates.
We spent many hours studying the deposit at the top of the caldera wall whose explosive emplacement apparently led to collapse of the previous Mt. Pagan a few thousand years ago. The deposit generally is 10-20 meters thick with rocks ejected outward being commonly 0.3m in diameter where we have examined it. Much of the erupted material undoubtedly fell into the sea. This eruption represents the maximum possible hazard from the volcano, but preliminary assessment indicates the likelihood of a recurrence is quite low. We will be compiling our geologic map of Mt. Pagan back in our offices in Hawaii, Alaska and Washington.
South Pagan volcano also has a caldera, although its diameter of 2-3 km is much smaller than the 5-6 km of Mt. Pagan’s caldera. Deposits of explosive origin north and south of the caldera probably led to its formation. Only two or three post caldera vents are visible on its floor. The largest and youngest of these extruded a voluminous blocky `a`a flow that completely filled the caldera and spilled down the volcano’s west and east flanks. We collected charcoal that should provide a date for this eruption.
Alamagan appears unchanged from 1994. Steaming areas, some quite impressive, surround the summit crater.
Guguan appears unchanged.
Sarigan shows no sign of activity. Its youngest lava flows appear to be several thousand years old.
Anatahan was observed on June 4 by helicopter. The permanent-glass EDM network and water temperature measurements were made during this trip.
Anatahan has had no historic eruptions. The geology team measured the temperature of several boiling pots and springs on the floor of the eastern crater. The temperature of the ponds as well as fumaroles ranged from a minimum of 96.7° C to a maximum of 100.3° C.
1. People who live in the northern islands should be made aware of volcano and earthquake hazards. To this end, we have encouraged each of the villages to operate a portable seismograph and learn to recognize the common signals caused by earthquakes, volcanic tremor, weather conditions, cultural noise, or instrumental problems. Detection of warning signals should be communicated by radio with EMO in Saipan. Residents in conjunction with civil authorities should prepare safety procedures and evacuation plans. It is essential that a log of seismic activity is kept, especially by those who reside on Agrihan, Alamagan, Anatahan, and by people who visit Pagan. This information should be archived and a copy be faxed or relayed to us for our database and further analysis. Data, such as estimates of the height of the ash cloud above the summit of the volcano and the vigor of the eruption should also be recorded in a log book of eruptive activity to correlate with the seismicity.
2. Continued support for maintaining the seismic instrumentation and data analysis is critically important in monitoring the internal state of the volcanoes in the northern islands currently occupied by residents. It is essential that field repairs of seismic equipment be made as soon as possible and that evaluation and communication of the seismic data occur in a timely fashion. Funding allowances to obtain specialized training and assistance in instrumental maintenance and record analysis are essential, as the seismic network improves and expands. Spare seismic components, particularly seismometers, telemetry radios, and discriminators are a critical need.
3. Potential activity on Agrihan, Alamagan, Anatahan and Pagan Island is of concern because of the ongoing resettlement of residents. Vigilance is of greater importance now with the planned reoccupation of the islands. In particular the EDM lines on Mt Pagan should also be remeasured more often. Since Mt Pagan is the most active volcano (with recolonization planned) in the active Mariana Island Arc, as much data as possible should be collected in order to ascertain how these types of volcanoes work. Data from Mt. Pagan could be used to develop models for the other volcanoes in the chain. Ideal times for remeasuring the EDM monitor would be whenever visits to Pagan Island to repair seismic equipment are scheduled. Additional deformation monitoring should be placed around the volcano to provide additional constraints on the volcanic processes coincident with volcanic activity. The CNMI government may wish to consider purchasing the equipment necessary to conduct the ground deformation studies.
In concert with the increase in the seismic coverage, EDM networks should continue to be monitored and measured on Agrihan. And monitoring established on islands that are presently being resettled (i.e. Alamagan). A lower priority would be to establish deformation monitoring networks on all the potentially active volcanoes in the Northern Mariana Island chain.
Measurement of these EDM networks
should be determined by the amount and type of seismicity.
The results of EDM measurements from Mt. Pagan in the past showed that
there was significant, measurable ground deformation. Yet with such infrequent
reoccupation of the EDM monitoring networks, we are unable to determine whether
the deformation was episodic (coincident with changes in seismicity) or were
accumulating gradually.
4. Reconnaissance geologic studies have been carried out on Pagan, but more work is required. The eruption frequency on all islands in the chain is unknown, and any plans for development of the northern islands should consider this information. Geologic mapping provides us with the long-term insight necessary to better understand possible future eruptive activity, especially to determine the magnitude and frequency of eruptions that could be expected at each volcano.
5. We understand the tremendous amount of resource expenditure associated with carrying out a volcano monitoring survey such as this one. We continue to support efforts to identify local resource people who can become knowledgeable in volcano monitoring. Ultimately, we would recommend that we take on the roles of resource and support personnel, and that our roles be secondary to those local experts, instead of having them supporting us as primary observers and data gatherers. The CNMI needs to identify people who are willing to learn and can fulfill this role in order that the CNMI attain self-sufficiency in volcano monitoring. Sending individuals to training courses and encouraging others to pursue degrees in Earth Sciences is a good beginning.
6. Education and outreach regarding volcanic hazards is the first step to mitigation. We recommend that the government begin educating the public about hazards. The USGS volcano hazards websites have lots of material towards this purpose and provides a good starting point.
ACKNOWLEDGMENTS
We thank Governor Pedro P. Tenorio and EMO Director the staff of the Emergency Management Office for excellent support during our survey. We especially appreciate the support of the CNMI field crew stationed on Pagan Island, including Ramon Chong, Juan Takai Camacho, Ben Lieto, Tony Tenorio, Joe Kaipat, Tom Crisostomo, Charlie Klakidm, Paul Cruz, Joe Basa, and Mark Pangelinan. In addition, the support of DPS Mike Camacho and Fire Division Medic Donald Taitano. The seismic survey was conducted cooperatively with the resident seismic specialist Ramon Chong and technicians "Ike" Ignacio Borja of the CNMI Office of Disaster Control. Special thanks are due to Greg DeLeon Guerrero, Director the Emergency Management Office, for his overall supervision of this campaign, and concern for hazard mitigation planning. We also would like to acknowledge Dr. Buffington who provided seed money for this integrated science project between GD,WRD and BRD. Finally, we would like to recognize northern islands Mayor’s Office (Joseph Ogumoro) for their assistance.
REFERENCES
Banks, N.G., Koyanagi, R.Y., Sinton, J.M., and Honma, K.T., 1984, The eruption of Mount Pagan volcano, Mariana Islands, 15 May 1981: Journal of Volcanology and Geothermal Research, v. 22, p. 225-269.
Moore, R.B., Koyanagi, R.Y., Sako, M.K., Trusdell, F.A., Kojima, George, Ellorda, Renee, and Zane, Sandra, 1991, Volcanological investigations in the Commonwealth of the Northern Mariana Islands, September-October 1990: U.S. Geological Survey Open-File Report 91-320, 31 p.
Moore, R.B., Koyanagi, R.Y., Sako, M.K., Trusdell, F.A., Ellorda, R.L., and Kojima, George, 1991, Volcanological investigations in the Commonwealth of the Northern Mariana Islands, May 1992: U.S. Geological Survey Open-File Report 93-541, 37 p.
Sako, M.K., Trusdell, F.A., Koyanagi, R.Y., Kojima, G., and Moore, R.B., 1994, Volcanic investigations in the Commonwealth of the Northern Mariana Islands, April to May 1994: U.S. Geological Survey Open-File Report 94-705.