Early warning capabilities for geological hazards

Robert M. Hamilton
National Academy of Science/National Research Council, Washington D.C., USA.
Tel./Fax: +1-202-334-3600,
E-mail: bhamilto@nas.edu

Geological processes like earthquakes, volcanic eruptions and landslides have the potential to cause great harm and can constitute hazards. There are two major strategies for reducing these impacts:

  1. Avoiding the hazards, by building structures and lifelines out of harmís way or by evacuating a threatened area, for example.
  2. Building structures to withstand the effect of the hazard.
Implementing these strategies in a cost-effective and acceptable manner requires reliable information about where and when hazards are likely to occur, and what their consequences might be. The capabilities for providing such information vary among the hazards, with some common aspects. In general, capabilities presently exist for identifying the areas where the occurrence of geological hazards is possible, but without indicating when they might occur. In some situations, a statement about the likelihood of occurrence may be made with specified reliability. The earthquake-prone regions of the Earth are well delineated in a general sense, and in some areas with high resolution, which provides an adequate basis for developing land-use plans and implementing seismic-resistant building practices at a national or regional level. Similarly, the location of volcanoes are generally recognised, and much is known about the level of activity of, and danger posed by many of them. Hazard assessments have been prepared for some volcanoes, but they are lacking for many other volcanoes that may threaten populated areas. Zones prone to landslides have been delineated in some regions based on topographic and geological information, but landslide hazards in most areas are generally only dealt with in conjunction with construction or development projects on a local basis. Tsunamis, which can be triggered by earthquakes, volcanic eruptions, or landslides that displace water, threaten low-lying coastal areas, usually around the Pacific Ocean, Mediterranean Sea, and Caribbean Sea, and some of these areas have been mapped. On some occasions, anomalous phenomena occur indicating that conditions for an earthquake, a volcanic eruption, or a landslide are building up, but without indicating exactly, or even approximately, when an event may occur. Interpretation of these phenomena could lead to a conclusion that the event may occur sooner rather than later. The capabilities for making such judgements depend on the specific situation, including knowledge of past occurrences, monitoring data, and any other relevant information. Generally, recognising symptoms for an earthquake are highly uncertain, whereas signs of impending volcanic eruptions or landslides could be fairly reliable in certain situations. With respect to short term predictions or forecasts, within four hours or days of an event, capabilities vary substantially among the geological hazards. Although successful short term predictions have been made for a few earthquakes, most predictions have been unsuccessful and there presently is no reliable capability that provides a useful basis for issuing short term warnings of the occurrence of an earthquake. A capability is emerging, however, to predict and warn of effects within tens of seconds after an earthquake occurs, but before the seismic waves reach a specific place, using high-speed data acquisition, processing and communication systems. In contrast, successful predictions of volcanic eruptions have been made in several situations that have resulted in saving many lives, although each impending eruption must be dealt with as an almost unique case. Landslides can also be predicted with some success, either in the general sense that they are likely in an area, due to, for example, heavy rainfall, or in a specific sense based on a recognition of incipient failure criteria. For earthquake-generated tsunamis there is an operational warning system for the Pacific Ocean. The occurrence of a large, shallow earthquake underwater triggers a tsunami watch, pending confirmation by tide gauges. Arrival times of a tsunami can be estimated accurately and useful warnings can be given. Further research on geological hazards is essential to improve understanding of the processes that cause them and of their effects. Monitoring systems need substantial improvement as very few areas are monitored with sufficient numbers of modern, high sensitivity instruments. Every country that is prone to geological hazards should complete a national hazards assessment, and, for the most active areas, regional and/or local assessments. These assessments should be incorporated into the on-going activities of the government agencies and communities responsible for economic and social development, particularly those involved with land-use planning and construction. Proper utilisation of mitigation measures and early warning systems can avoid enormous economic losses and save lives. When phenomena are detected on monitoring systems or reported by observers indicating that a geological hazard may be developing, experts should be assigned to conduct appropriate investigations. Every hazard-prone nation should either develop the necessary expertise itself or establish access to it through some form of bilateral or multilateral assistance agreement.