REPORT:
The Importance of Field Observations for Monitoring Volcanoes, and the Approach of "Keeping Monitoring as Simple as Practical"

Abstract

Geologic field observations form an important part of any volcano-monitoring program yet are often overlooked in lists of monitoring techniques. Such observations provide the opportunity to integrate many different kinds of data on the spot and to design simple measurements to test key questions resulting from the observations. Field observations go hand in hand with more sophisticated equipment and techniques to form a complete system for monitoring volcanoes. Monitoring programs should explicitly include provisions for geologic field observations and instill in field workers, scientists, and technicians alike, the need to be flexible and clever in designing simple experiments and measurements to test important field observations on the spot.

Introduction

Other chapters in this bulletin describe the many different methods of monitoring volcanoes that have been used by the on-site and affiliated research staff of the Cascades Volcano Observatory (CVO). Most of these methods involve relatively sophisticated equipment, training, or data analysis. Many of the methods have been used successfully at a number of volcanoes around the world in various environmental conditions. They are largely proven commodities that can be put to use for study of active volcanoes anywhere. They reflect the remarkable ingenuity and significant advancement in monitoring volcanoes that characterize the past decade in volcanology.

My aim in this essay is to highlight another means of monitoring that is the oldest that volcanologists have and yet is commonly overlooked in establishing a monitoring program. It cannot take the place of the more sophisticated techniques but can augment them and help guide their use. It is the only means that allows instant on-the-spot integration of a number of parameters that characterize the volcano, and instant reevaluation of that integration. It is old-fashioned yet up-to-date, operates in all but the most extreme environments, and is as durable as the remarkable machine that runs it. The means if direct field observation, and the machine is the observer.

The simplicity of geologic field observations belies their significance, and I provide three examples to demonstrate that fact. I also try to make the point that even simple measurements made on the basis of the observations can be sufficient, and that making the measurements more sophisticated than required does not necessarily improve their value -- and in fact can even undermine their utility.

Examples

Three examples from Mount St. Helens illustrate my thesis that observations and simple measurements can be effective under the right circumstances. They are representative but not exhaustive examples of the importance of this thesis at Mount St. Helens. The examples are some of those with which I am most familiar.

Spirit Lake as a Tiltmeter

Why not use Spirit Lake as a giant carpenter's level? This idea had several advantages. The lake was far enough from the foot of the volcano to be an effective monitor of deep-seated changes under the volcano. The lake was oriented with its long direction radial to the volcano, and quick calculations suggested that tilts on the order of 2 urad would be detectable with reasonably careful measurement procedures. The lake was covered with ice, so that wind-related waves would be damped out. Boat docks and partly submerged stumps in small areas of open water provided sites to install measurement devices.

But what devices could be obtained quickly? We couldn't afford the time to obtain normal water-level sensors and install them. An inexpensive and readily available alternative was immediately apparent. A visit to a local lumber yard resulted in an outright gift of several wooden yardsticks (metersticks), which were easily nailed to piers and stumps. Within a day the level of the lake was monitored by individuals visiting each site, reading the level of the lake on the partly submerged meterstick, recording the time of their visit, and closing back on a station chosen as a reference point. Later we used several observers with synchronized watches, but the results were the same: The shore of Spirit Lake was not tilting. This told us that a large volume of magma was not intruding under the volcano, so that we could concentrate our attention on the edifice of the volcano itself (Lipman and others, 1981).

The North-Flank Bulge

The obvious way to test for deformation of the edifice was to make repeated measurements to targets on snow-free parts of the volcano. Quickly a standard theodolite used by most surveyors was obtained and angles measured to prominent natural features on the volcano. The data suggested outward movement of the north flank, but aiming on natural features was difficult and subject to considerable error. Within a day or two an electronic distance meter (EDM) arrived, and reflectors were placed on the mountain. The EDM is a sophisticated instrument, but the targets were nothing more than clear plastic highway reflectors screwed to boards that were lashed to steel fence posts hammered into the ground. A combination of the theodolite and EDM measurements to these targets soon indicated that the north flank was moving northward at a steady rate, so that clearly the volcanic edifice was deforming in response to intrusion of magma (Lipman and others, 1981). The theodolite measurements to the wooden targets provided sufficient data by themselves to define the rapid displacements of as much as 2.5 meters/day. The EDM data were useful adjuncts to the theodolite measurements but were not necessary to trace the movements at all but the least sensitive targets.

Measurements of Cracks and Thrust Faults Used for Prediction of Dome Growth

During the December-January event, geologists were surprised to find two thrust faults on the crater floor north and northwest of the dome. The thrust faults faced outward, away from the dome. The same questions arose with these faults as with the cracks: were they still moving and, if so, might they provide an indication of future eruptive activity? Again rebar stakes were driven into the upper and lower lates of each thrust, and the distance between them was measured with the steel tape. The distance between each pair of stakes shortened with time and indicated that the upper plate was moving across the lower plate. Repeated measurements of the thrusts and associated radial cracks soon showed a distinctive pattern. Rates of displacement were slow after a dome-building episode but increased nearly exponentially as the onset of the next dome-growth event neared (Chadwick and others, 1983). This pattern and various other details of the entire deformation process of the crater floor were nicely traced with the simple rebar-and-tape method and formed the basis for a series of predictions of dome growth in 1981-1982 (Swanson and others, 1983). Moreover, the simple measurements with a steel tape provided the primary data for two interpretative papers about how and why the crater floor deformed (Chadwick and others, 1988; Chadwick and Swanson, 1989).

Discussion

These three examples each show the value of on-the-spot field observations and the application of simple measurement techniques to resolve important questions and stimulate significant interpretations. Within hours to several days the basic questions raised by each set of field observations were answered, although of course refining the details took much longer. The answers came quickly, in part because the questions were easy (yet very important), but in part because the measurements could be started quickly owing to the simplicity of the methods used. Moreover, the simple methods were reliable and didn't depend on a complex chain of equipment, any link of which could fail unexpectedly. Total reliance on sophisticated equipment can backfire if the means is unavailable for its rapid repair. In general, a good guide to follow for devising a monitoring method is to "keep it as simple as practical."

Since 1980 notable improvements have been made in portable electronic-monitoring equipment, as several chapters in this bulletin show. Nonetheless, given the same circumstances as those in the three examples, I believe that the answers would still come faster with the simpler methods, largely because the sophisticated equipment requires significant installation and "settling-in" time. There is no question that the more sophisticated techniques fill important gaps, because most have the capability to acquire data almost in real time, 24 hours a day under any weather condition. But one of the principal limitations of these techniques is one of the principal strengths of field observation and simple measurements: flexibility and the ability to integrate several observations on the spot and to design a measurement to test that integration. Continuous measurements often can proxy for an on-site observation, but even the most broadly based -- television monitors -- are less adaptable and ingenious than a human observer in the field.

The flexibility provided by field geologic observations cannot be overemphasized. Most electronic sensors are designed to detect and report one parameter, such as tilt, displacement, seismicity, or changes in a particular species of gas. If some unmonitored parameter changes, the sensor either doesn't detect it or may provide spurious information. On-the-spot geologic observation clearly is not restricted to a single parameter, although of course it is restricted to what can be observed with the eyes, ears, and nose. Small changes may go unnoticed by field observers, but large changes may go unnoticed by electronic sensors not monitoring the proper parameter. Only on-the-spot observers can quickly assess the situation and determine which parameters are likely to provide vital information. Measuring an unimportant quantity accurately and continuously using sophisticated equipment can be a waste of time and resources; it is much better to get to the heart of the matter by the simplest means possible.

Another important point is that even the best remote-monitoring techniques require verification by on-site field observations and integration with other data. For example, an electronic tiltmeter may indicate a change, but his change could reflect an electronic problem, an unstable installation, or real deformation of the volcano. Only independent information can determine which interpretation is most likely. As a rule of thumb, never blindly accept data obtained remotely until on-site verification can be made. Accept no substitute!

Many colleagues emphasize the importance of sophisticated equipment and telemetered data during times of great hazard to field workers, and I agree with their reasoning. It is far better to lose a piece of equipment than a life. No reasonable person would advocate fieldwork under conditions that he or she feels are life-threatening. But there are many times in the activity of a volcano when conditions are not so hazardous, and it is these times that are suitable for field observation and indeed would benefit from such observations. Moreover, those of us who study active volcanoes must admit that certain risks exist, just as they do for firefighters and police officers. An integral part of our job is to assess those risks and establish personal guidelines for the relatively safe conduct of our research.

I see no point in providing further discussion of putative advantages and disadvantages of geologic field observations versus electronic measurements. The real point to be made is that we should think no in those terms (although many of us do), but instead in terms of an integrated monitoring effort that incorporates the best of all observations into unified interpretations. No one approach in inherently superior. There is a tool for every job, and the trick is to find the best tool or set of tools, whether they are close-in observations, complex equipment, sophisticated telemetry systems, or a combination of all three.

Nothing is new in this discussion, but the emphasis on sophistication and telemetered data in this bulletin seems to me to require a counterpoint, even though an obvious one. Field observations and related measurements are a vital component of volcano monitoring, just as are electronic sensors, radios, and high-priced surveying equipment. In fact, the two approaches commonly merge. For example, sophisticated (and expensive) EDM's measure to painted wooden boards with plastic reflectors screwed on them -- a real marriage of the aristocracy and the proletariat!

Indonesia provides a remarkable example of how a combination of simplicity and complexity has saved lives. In 1988, Indonesia experienced seven explosive eruptions, all of which had seismic precursors. As a result of the warnings and preliminary, closely observed, minor eruptive activity, 33,000 people were evacuated from their homes, and only four lives were lost (all on Banda Api, where those who were killed were knowingly evading the evacuation order) (T.J.Casadevall, written, commun., 1989). Indonesia has about 150 active volcanoes and 50 volcano observatories. Each observatory has from two to four observers, generally only one seismometer with a smoked-drum seismograph, and typically only radio contact (a few have telephones) to central headquarters in Bandung. Nonetheless, the observers are well trained to note changes in seismicity and many other parameters (such as the presence of new ejecta, changes in plume behavior, fumarole temperatures, characteristics of crater lakes, and other factors), and they are form the region around the observatory, know the local people well, and have their trust. Perhaps there are lessons here for all of us. One lesson is that the simplest kind of monitoring and warning system that works is the best one to use; in this example, the monitoring system consisted of the simplest kind of complex instrument -- a single seismometer and smoked-drum seismograph, and the warning system involved the observers who were intimately familiar with the monitored volcanoes and with the people of the region. Another lesson, not germane to the theme of this chapter yet of great importance, is that the familiarity (and even closeness) of the observers to the local populace is a key that we should consider more in our efforts to save lives.

During some volcanic crises, many scientists and technicians converge on the scene (or to a field observatory) but may stay for only short periods of time, returning to other duties as required. Continuity of observations and data gathering is difficult to maintain with such a rotation of staff, especially for field observations that commonly involve relatively subjective descriptions. One way to minimize such problems is to develop and use a check list of the kinds of field observations that should be made routinely. Such a list would in detail be specific to each volcano and would be susceptible to modification as the activity of volcano developed. Another key element in maintaining continuity of field observations is the thorough transfer of information between departing and arriving observers, preferably by visits to the volcano together.

Conclusions

My purpose in writing this essay is not to minimize in any way the modern monitoring techniques that are helping make volcanoes safer and better understood. I have used many of these techniques myself and have championed their implementation by others. Instead I simply want to point out that any monitoring effort is incomplete if trained observers are not made a part of it, and if the opportunity is not given for these observers to influence the gathering and interpretation of the data acquired by sophisticated techniques. In my view a volcano observatory or monitoring program of any sort starts with the people on its staff, not with the equipment in its locker or planned in its budget. Each geologist or geophysicist who spends much time on a volcano should be trained to observe field conditions, to think about those observations while in the field, and to be flexible and clever in devising simple measurements that can be made quickly and definitively once changes are noted. Those scientists and technicians should not have tunnel vision for only their specialty but instead should integrate all available information and be ready to respond to observable changes of any significant parameter. These individuals should be given the field time to spend on active volcanoes, and in fact such time should be an integral part of the monitoring program. Nature is too complex for us to learn enough about a volcano by monitoring it only remotely; we must also observe and monitor it personally and closely.

A volcano is too complex to be understood by individuals working alone. Free communication and exchange of data, observations, and ideas among all workers involved in monitoring are essential. Integration of field observations with telemetered data is vital to preparing a unified assessment of the volcano and its hazards. The stakes are so high that the data a observations, and the resulting development of interpretations, must be shared in an atmosphere of intellectual curiosity and social responsibility, rather than compromised in an atmosphere of competition or one of conflict among senior and junior scientists and technicians.

References Cited

[not online]