Hydrological changes after the 2016 Kumamoto earthquake, Japan
KeywordsKumamoto earthquake Hydrological changes Streamflow Spring Groundwater Water-holding capacity
Large earthquakes are well known to sometimes cause wide hydrological changes in regions affected by strong ground motion (e.g., Waller 1966). In several cases of such changes, streamflow and spring flow increased in lowlands and the water table dropped in highlands. Rojstaczer and Wolf (1992) and Rojstaczer et al. (1995) comprehensively explained those changes related to the 1989 Loma Prieta earthquake with permeability enhancement. Sato et al. (2000) also used permeability enhancement to explain postseismic groundwater changes after the 1995 Kobe earthquake. However, an analysis of postseismic hydrographs of rivers affected by several large earthquakes, including the 1989 Loma Prieta earthquake, showed that the hydrological changes were caused by liquefaction and not permeability effects (Manga 2001). However, liquefaction seems to be an unlikely cause of water table drops in highlands and large and long postseismic increases in stream water. Actually, in the 1999 Chi-Chi earthquake in Taiwan, no postseismic flow rate increase was found in a certain stream, although heavy liquefaction occurred in the catchment area of the stream (Wang et al. 2004). The observed increases in streamflow and spring flow in lowlands, and the water table drop in highlands, following the 1999 Chi-Chi earthquake may be explained by enhanced vertical permeability in the mountain region, a phenomenon that has less effect on hydrographs (Wang et al. 2004; Wang and Manga 2015). Alternatively, Montgomery et al. (2003) suggested that co-seismic volumetric strain changes may have contributed to postseismic streamflow changes after the 2001 Nisqually earthquake (Mw6.8).
Because Japan is relatively rich in water resources, people may have not paid much attention to earthquake-related hydrological changes. But those hydrological changes are clearly one of the seismic risks and should also be examined in Japan because those changes sometimes continue for a period of several months to years (Rojstaczer and Wolf 1992; Sato et al. 2000).
In this paper, we report the postseismic hydrological changes related to the 2016 Kumamoto earthquake.
The region of strong ground motion during the main shock contains three main river systems: Shira River system, Midori River system, and Kikuchi River system (Fig. 1). We obtained data for these rivers from the water information system of the Ministry of Land, Infrastructure, Transport and Tourism (2019). In January 2019, flow rate data were available up to December 2017. Therefore, we obtained streamflow data collected by eight observation stations along the three river systems during the period 2001–2017. These eight stations have measured streamflow since 2001 or earlier, with no long data gaps through the end of the study period.
Spring water measurement
We surveyed the 11 water springs shown in Fig. 1 six to nine times for flow rate, temperature, and chemistry during the period April 16, 2016, to November 2017. The sampling intervals were a few weeks to several months. The initial surveys were performed several days to 1 month after the April 16, 2016, main shock. Fortunately, 8 of the 11 springs were also surveyed 1 month to a few years before the 2016 Kumamoto earthquake, although the flow rate of one of these eight springs was not measured before the 2016 Kumamoto earthquake. Therefore, we can evaluate the postseismic flow rate changes at 7 of the 11 springs. We can also evaluate postseismic changes in water temperature and chemical composition at 8 of the 11 springs.
We obtained daily precipitation data from the Kumamoto Local Meteorological Observatory (KMMT in Fig. 1) and the Aso-otohime AMeDAS or the Aso-otohime Automated Meteorological Data Acquisition System (ASOT in Fig. 1) from the JMA database (Japan Meteorological Agency 2019b). The daily precipitation is generally larger at ASOT than at KMMT, but the patterns are similar.
Analysis to remove effects of precipitation on streamflow
We also calculated relative precipitation by estimating the average accumulated precipitation. We applied an approximately straight line to the accumulated precipitation data from 2001 to 2015, as shown in Fig. 2b. This straight line is the average accumulated precipitation. We then subtracted the average accumulated precipitation from accumulated precipitation, leaving relative precipitation as the remainder.
Estimation of temporal streamflow response to precipitation
Precipitation was ≥ 20 mm and ≤ 100 mm at both KMMT and ASOT, because a small amount of precipitation produces no clear direct runoff and because a large amount of precipitation causes an irregular streamflow response.
API(10) ≤ 10 mm both at KMMT and ASOT at the start of precipitation.
Direct runoff had finished flowing, followed by several hours of stable streamflow at the same rate observed prior to the start of precipitation ± 5 m3/s.
The ratio of precipitation at ASOT to that at KMMT was ≥ 0.8 and ≤ 2.0, because very local precipitation produces an irregular streamflow response and because the accumulated precipitation at ASOT during the period 2001–2017 was 1.5 times larger than that at KMMT.
Rest accumulated flow rate of the streams
Temporal change in streamflow response to precipitation
Flow rate at springs
Water temperature and chemical composition of the spring water
If the increases in postseismic streamflow had been caused by earthquake-related permeability enhancement, then the streamflow would have increased immediately after the earthquake (e.g., Rojstaczer and Wolf 1992). Therefore, permeability enhancement did not cause the streamflow increase observed 2 months after the 2016 Kumamoto earthquake. Co-seismic volumetric effects (Montgomery et al. 2003) and liquefaction (e.g., Manga 2001) also cannot explain the delayed postseismic streamflow increase in this region. Hosono et al. (2019) reported both of co-seismic drops and rises in stream water level. Similar changes are also recognized in Fig. 5. However those changes were much smaller than the delayed postseismic streamflow increase.
The postseismic decreasing trend in water-holding capacity within the catchment following the earthquake is shown in Fig. 6. The 2016 Kumamoto earthquake caused many landslides in and around the Aso caldera (Miyabuchi 2016); these landslides are expected to have decreased the water-holding capacity of the catchment. Such a phenomenon would cause abnormal increases in postseismic streamflow with heavy precipitation, and an overall increase in the streamflow rate at some flow rate stations in the region.
Relative to the flow rate before the 2016 Kumamoto earthquake, it rose at three of the seven springs (e:SOY, g:YSM, and k:DMZ) and dropped at two of them (f:SOI and h:SMR) (Fig. 7). Hosono et al. (2019) also reported both of co-seismic rises and drops in groundwater level. However, large seasonal changes, which were probably induced mainly by precipitation, were recognized at e:SOY, f:SOI, g:YSM, h:SMR, and k:DMZ. Our sampling interval was much longer than that of Hosono et al. (2019). Therefore, the postseismic flow rate changes in Fig. 7 could not be clearly attributed to the earthquake.
Before and after the 2016 Kumamoto earthquake, water temperature and chemical composition were almost stable at the observed springs. Only the concentration of NO3− was somewhat increased just after the earthquake and it soon decreased. Similar results in other groundwaters were also reported by Kawagoshi et al. (2018). NO3− in groundwater is generally considered to be derived from the surface. Therefore, it seems that supply from the shallow groundwater to the spring was slightly increased and then decreased after the earthquake. These results show that hydrothermal fluid did not enter the spring waters after the 2016 Kumamoto earthquake.
We analyzed streamflow data from eight observation stations on three major rivers in Kumamoto Prefecture for the period 2001–2017. We also surveyed 11 water springs in the region several times after the main shock. Some of the eight observation stations recorded large increases in streamflow following a heavy rainfall that occurred 2 months after the earthquake. This may be due to a decrease in the water-holding capacity of the catchment caused by earthquake-induced landslides. In contrast, earthquake-related changes in the spring flow rate were not so clear. Water temperature and chemical composition of the spring waters also changed very little. Only the concentration of NO3−, whose origin is usually considered to be the surface, changed slightly just after the earthquake. These results show that the postseismic hydrological changes were caused mainly by earthquake-induced surface phenomena and that there was little contribution from hydrothermal fluid.
We thank Dr. K. Osaka for his helpful advice. We also thank Dr. K. Kawabata for her cooperation during the field surveys in Kumamoto Prefecture. We thank the residents of Kumamoto Prefecture for their cooperation in this research. We are grateful to the two anonymous reviewers for reviewing our manuscript and valuable comments.
NK carried out the processing and analysis of groundwater and stream water data and drafted the manuscript. SM, TT, and AM carried out the data processing and analysis of the streamflow data. TA, TS, and HT carried out the groundwater observations and chemical analysis of the spring waters. NM helped draft the manuscript. All authors read and approved the final manuscript.
This work was supported by a Grant-in-Aid for Scientific Research on Innovative Areas (KAKENHI No. 26109006) from the Ministry of Education, Culture, Sports, Science and Technology.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
- Geospatial Information Authority of Japan (2019) Maps of geospatial information authority of Japan. http://maps.gsi.go.jp/. Accessed 17 Mar 2019
- Hirata N (2016) The 2016 Kumamoto earthquake, report of the coordinating committee for earthquake prediction research, vol 96, pp 490–491 (In Japanese) Google Scholar
- Ichiyanagi K, Ando S (2017) River water level change affected by the Kumamoto earthquake. J Jpn Assoc Hydrol Sci 47:11–15 (in Japanese) Google Scholar
- Japan Meteorological Agency (2019a) Estimated seismic intensity distribution map of Kumamoto Prefecture Kumamoto region M7.1 at 01:25 on April 16, 2016. https://www.data.jma.go.jp/svd/eew/data/suikei/201604160125_741/201604160125_741_1.html. Accessed 17 Mar 2019
- Japan Meteorological Agency (2019b) Past weather data and download. http://www.data.jma.go.jp/gmd/risk/obsdl/index.php. Accessed 15 Jan 2019
- Ministry of land, infrastructure, transport and tourism (2019) Water information system. http://www1.river.go.jp/. Accessed 17 Jan 2019 (in Japanese)
- Sato T, Takahashi HA, Kawabata K, Takahashi M, Tosaki Y, Miyakoshi A, Inamura A, Handa H, Matsumoto N, Kazahaya K (2017) Anomalous changes in groundwater and hot spring water after the 2016 Kumamoto earthquake. In: Abstracts of JPGU-AGU joint meeting 2017, Makuhari Messe, Chiba, Japan, 20–24 May 2017Google Scholar
- Shirahama H, Yoshimi M, Awata Y, Maruyama T, Azuma T, Miyashita Y, Mori H, Imanishi K, Takeda N, Ochi T, Otsubo M, Asahina D, Miyakawa A (2016) Characteristics of the surface ruptures associated with the 2016 Kumamoto earthquake sequence, central Kyushu, Japan. Earth Planets Space. https://doi.org/10.1186/s40623-016-0559-1 CrossRefGoogle Scholar
- Waller RM (1966) Effects of the March 1964 Alaska earthquake on the hydrology of south-central Alaska. USGS Professional Paper 544A, 28 p, https://pubs.usgs.gov/pp/0544a/. Accessed 7 July 2019
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.