All above mentioned cases of water loss due to earthquake activity are known from published sources such as newspapers, internet web sites, and scholarly papers/reports. However, detailed geographic data are required to study the effect of earthquakes on hydrologic environments. Such data should include the geographic location of wells, their characteristics regarding water level before and after an earthquake, impact on households, condition of casing (e. g. collapsed vs. intact), and detailed description of water quality. While there are many documented studies of earthquake effects on water level in USGS monitoring wells, there are only a few cases known to the scholarly community that are well documented with maps and data on the conditions of the domestic wells serving water supply. Both geological and cartographic data are critical to understanding the environmental conditions that create a hazard and consequently lead to a disaster.
Data on water loss in domestic wells and associated data on geology and topography would allow scientists to create models based on geological and environmental conditions to study the repercussions of seismic events in the past and present. For example, we already know that the potential impact of an earthquake on domestic water supply can cause reduction in water discharge in wells located on top of ridges and mountains and increase in discharge in wells located in valleys and low-lying areas. However, we still do not know specifics about hydrogeologic conditions that cause these effects; furthermore, we do not know geological and hydrologic conditions that set the stage for these events.
Some of the influencing geological factors and environmental conditions related to the effect of seismicity on loss of water supply in wells are discussed in Gorokhovich and Fleeger (2007) using the Pymatuning earthquake as an example that was well documented and described previously by Armbruster et al. (1998) and Fleeger et al. (1999). The model which we will be referring to later as the ‘Pymatuning model’ contains two potential hazardous conditions: (1) Heterogeneity of geological material composing aquifers or existence of a perched aquifer; (2) Location of wells on any high elevation (plateau, slopes above valleys, etc.) that can be treated as a “hydrologic island”, i. e. a structure that has a limited and geographically narrow source of recharge (Toth, 1963; Norvatov and Popov, 1961).
These two conditions favor a situation when aquifers located above the valley bottom can lose water after an earthquake due to increase in their permeability. This increase can be drastically exacerbated by the heterogeneity of the aquifer material that can cause an unexpectedly rapid increase in permeability. This, in turn, causes fast discharge of groundwater down the valley bottom with the subsequent increase in river discharges or overflow of water in wells located at the valley bottom or in its vicinity.
The Pymatuning model can also explain the effects of the Matsushiro earthquake that took place in Japan in 1965 and produced series of seismic shocks through March 1968, causing among other disasters (such as faults and landslides) a loss of water in two villages (Terakawa and Matsuo, 1996). A published map and GIS topographic data helped restore the geological settings in the area affected by the earthquake. Figure 10.1 Shows a recreated map of the affected area between the cities of Nagano and Ueda using data from Terakawa and Matsuo (1996) and global digital elevation SRTM (Shuttle Radar Topographic Mission) model. Oval symbols show areas where water discharge increased and decreased.
According to the geologic map of Japan posted by the Geologic Survey of Japan (GSJ, 2008), area between Ueda and Suzaka (Fig. 10.1) consists mainly of mafic and plutonic volcanic rocks overlaying early miocene marine and nonmarine sedimentary rocks. These rocks are represented by porphyrites, diorites, pebbly mudstone, and sandstone covered by alluvial deposits; high quantities of clays and clay minerals were found on slopes within the Matsushiro area. These sediments exacerbated slide-prone conditions (Morimoto et al., 1967). Intrusions of diorite-porphyrite penetrated sedimentary rocks during the Miocene and later during Pliocene and Pleistocene were covered by Pleistocene volcanic complexes (Tsuneishi and Nakamura, 1970). Accumulation of sediments still continues in the area and results in a composition of 328-656 ft (100-200 m) thick alluvial fans and unconsolidated sediments (Ono, 1967; Nakamura and Tsuneishi, 1967).
Fig. 10.1 Geographic settings of the Matsushiro earthquake (areas of water discharge/overflow were adapted from Terakawa and Matsuo, 1996)
Geological conditions in the Matsushiro area provide favorable environment for heterogeneous aquifers with variable permeability. In such conditions, aquifers could easily loose water during earthquakes. The terrain is mountainous and some small mountains with elevations of less than 3,280 ft (1,000 m ) are separated from more sizeable mountains with much higher elevations (more than 9,842 ft or 3,000 m a. s.l.) located in the east.
Figure 10.1 Shows that areas with dried up wells are on high elevations (2,6253,280 ft or 800-1,000 m a. s.l.) while areas with excessive flows are in the valley near the Tikuma river. Both conditions (high elevations and heterogeneous aquifers) are in accordance with the Pymatuning model, which was also applied to the interpretation of possible causes of crisis on Crete (Gorokhovich and Fleeger, 2007).
World History
