material. A proven mitigation technique in such cases is for geologists to locate the water surface in fractured rocks and drain off destabilizing water by drilling horizontal wells.
Then there are the regional-scale contributory causes of increased landslide susceptibility such as deforestation. According to the World Resources Institute, approximately 109,000 km2 of tropical forest is being destroyed annually—an area the size of Ohio. Removal of the forest cover increases flooding, erosion, and landslide activity. This deforestation is causing serious landslide problems in many developing countries.
Over a period of about 3 years in the 1980s, during the course of an El Niño episode, regional weather changes in the western United States resulted in much heavier than average precipitation in mountainous areas. That increased precipitation caused a tremendous increase in landslide activity in California, Nevada, Utah, Colorado, Washington, and Oregon. Scientists are coming to understand such cycles through integration of collected data with information found in the historical and geological records. Cycles such as El Niño form the background variation of the climate pattern. But earth scientists do not know what to expect with additional perturbation from a changing greenhouse effect. Will the predicted temperature increase cause a decrease in precipitation, as occurred in mid-America during the summer of 1988? Will it increase storm activity throughout the mid-latitudes? Will it disrupt global climatic patterns, resulting in droughts in some areas and increased precipitation in others? Documented cause-and-effect sequences such as those related to El Niño episodes suggest that if areas prone to landslides are subjected to heavier than normal precipitation, susceptible slopes are likely to fail.
Land subsidence can be currently observed in at least 45 states; it is estimated to cost the nation more than $125 million annually. Subsidence can have human-induced or natural causes; both are costly. In the United States at least 44,000 km2 of land has been affected by subsidence attributed to human activity, and the figure is probably higher. As for natural subsidence, one event—the 1964 Alaskan earthquake—caused an area of more than 150,000 km2 to subside as much as 2.3 m. This event was extreme but not atypical of past or probable future disturbances.
The causes of subsidence are various but well known, and the hazard presented is well recognized; however, the indications of specific imminent danger and the possible cures or preventives are not clear. Subsidence can be induced by withdrawing subsurface support—by removing water, hydrocarbons, or rock without a compensating replacement. In many instances, oil or water is removed from porous host sediments that compact as the interstitial fluid is removed. In such cases, collapse is slow and gentle. More dangerous are situations that leave voids—withdrawal of water from cavernous limestone or mining of coal, salt, or metals. These voids can collapse gradually or suddenly. Not all subsidence is unexpected—ground over longwall coal mines is supposed to subside gradually to a new elevation that is both safe and stable. The ground surface above subsiding land is not a good place for a shopping center or school building, but it may be quite suitable for crops or recreation.
Natural subsidence occurs for several reasons. A basin surface may warp downward in response to recent sediment loading or by dewatering and compaction of sediments; both processes presently affect the Mississippi River delta. Tracts may subside because of folding or faulting, as in the Alaskan example above. Regions such as the Texas Gulf coast are triply vulnerable because they overlie a downward-flexing part of the crust; are above a thick pile—up to 10 km deep—of compacting sediments; and are being mined for groundwater and petroleum, which accelerates deflation of the sedimentary pile.
The primary cause of the most common, and most dangerous, subsidence in the United States is groundwater extraction through water supply wells. In California's San Joaquin Valley, 13,500 km2 of land surface has sunk as much as 9 m in the past 50 years because of removal of groundwater for irrigation. The danger, of course, is in more populated areas, especially those close to sea level. Some cities that are already struggling because of groundwater extraction include Houston-Galveston, Texas; Sacramento and Santa Clara, California; and Baton Rouge and New Orleans, Louisiana. The problem of induced subsidence is international and also threatens London, Bangkok, Mexico City, and Venice. If sea level continues to rise, the cities that are literally on the edge now will be fighting to stay above water.
Sinkholes, another common source of land collapse, can occur unexpectedly on a more local scale than wholesale subsidence. Sinkhole collapse generally results from the slumping of poorly consolidated surficial material into underground caverns.