Assessing Climate Change Impacts on Extreme Rainfall and Severe Flooding

Hydrological Research Letters 14(4), 155–161 (2020) Published online in J-STAGE (www.jstage.jst.go.jp/browse/hrl). DOI: 10.3178/hrl.14.155 Assessing climate change impacts on extreme rainfall and severe flooding during the summer monsoon season in the Ishikari River basin, Japan Thu Thanh Nguyen1,2, Makoto Nakatsugawa1, Tomohito J. Yamada3 and Tsuyoshi Hoshino3 1Department of Civil Engineering, Muroran Institute of Technology, Japan 2Faculty of Urban Environmental and Infrastructural Engineering, Hanoi Architectural University, Vietnam 3Faculty of Engineering, Hokkaido University, Japan

Abstract: future (Higashino and Stefan, 2019; Hirabayashi et al., 2008). This study investigates the change in extreme rainfall In Japan, extreme flood events from heavy rainfall have and river flooding for a large river basin due to climate been recorded regularly in recent years. For example, large- change during the summer monsoon using a large ensemble scale flooding due to heavy rainfall occurred during July 5– dataset (d4PDF) coupled with the Integrated Flood Analy‐ 8, 2018 in western Japan, causing extensive damage over sis System (IFAS). Frequent severe flooding causes signifi‐ numerous prefectures and resulting in 224 deaths, 21,460 cant damage in Japan. Therefore, we aim to provide useful collapsed houses, and 30,439 inundated houses. In the Oda information to mitigate flood damage. The study area is the River and its three tributaries, levees were breached at eight Ishikari River basin (IRB) in Hokkaido, Japan. We used the points due to the “backwater phenomenon” in which the d4PDF 5-km downscaled rainfall data as input for the IFAS tributary river floods synchronized with the main river model. The results showed that, for a given increase in flood (River Council for Social Infrastructure Develop‐ extreme rainfall, the discharges from the IRB and its main ment, 2018). Considering various climate change scenarios, sub-basins increase to a greater extent. The differences several studies have predicted increased rainfall in the between the time of peak discharge at the reference stations future (Kitoh et al., 2009; Kim et al., 2010). Kim et al. in each tributary and the time of peak water level at the (2010) indicated that rainfall in Hokkaido is expected to confluence points in the main river are evaluated. Climate increase by 6.1% and 10.6% in the near and extended change effects are significant in the southern sub-basins, future, respectively. Additionally, Yamada (2019) reported wherein the amount of extreme rainfall increases by 29%– that extreme rainfall will be more extensive under future 35%, whereas the river discharge increases drastically climate conditions. (37%–56%). Additionally, the time difference decreases by Attempts have been made to develop flood adaptation 1.02–2.14 h. These findings will help policymakers develop strategies to address the critical effects of climate change future flood control measures in flood-prone areas. on the risk of river floods. Several studies have been con‐ ducted in important river basins internationally (e.g. KEYWORDS river flooding; extreme rainfall; time Shrestha and Lohpaisankrit, 2017; Try et al., 2020), as well difference; Ishikari River basin; d4PDF; as regionally in Japan (e.g. Sato et al., 2012; Tachikawa IFAS et al., 2009; Hoshino and Yamada, 2017). A study by Tachikawa et al. (2009) indicated that severe rainfall would INTRODUCTION increase in the Yoshido River basin, and the peak flood dis‐ charge would increase to a greater extent in the future. Additionally, the trend of extreme rainfall events increasing According to the Fourth Assessment Report (AR4) of the in a short period of time should be considered. Sato et al. United Nations Intergovernmental Panel on Climate (2012) indicated that climate change is projected to change Change (IPCC), severe natural disasters due to extreme cli‐ river discharges significantly, especially in northern Japan. mate have become more frequent since 2000 (IPCC, 2012). Thus, understanding the change in the amount of rainfall, Floods are considered as extreme weather events that occur especially considering future extreme rainfall events, and frequently and cause severe damage (Doocy et al., 2013; assessing its effect on the risk of river floods in vulnerable Hirabayashi and Kanae, 2009). Although several factors basins is necessary to create effective flood control plans. contribute to flooding, heavy or prolonged rainfall is con‐ This study investigates the changes in the risk of river sidered the most critical factor that causes floods. The flooding associated with climate change in the Ishikari IPCC 5th Assessment Report (IPCC, 2013) indicated that River basin (IRB), a socioeconomically important basin in rainfall is expected to increase in Asia during future sum‐ Hokkaido, Japan. This study is the first to assess the mer monsoon seasons, and extreme rainfall is likely to changes in extreme short-term rainfall and extreme river become more frequent. Increasing rainfall, especially flooding events during the summer monsoon in the IRB as extreme rainfall, has enhanced the risk of floods in the

Received 21 August, 2020 Correspondence to: Nguyen Thanh Thu, Muroran Institute of Technology, 27-1 Mizumoto, Muroran, Hokkaido 050-8585, Japan. E-mail: Accepted 30 October, 2020 [email protected] Published online 12 December, 2020

© The Author(s) 2020. This 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.

—155— T.T. NGUYEN ET AL. well as in its main sub-basins (IM-SBs) using an Integrated 60 years (2051–2110) and comprises 90 members (total: Flood Analysis System (IFAS) coupled with a large- 5,400 events) for the +4K future climate simulation, and 60 ensemble rainfall dataset (d4PDF) with a high resolution of years (1951–2010) and 50 members (total: 3,000 events) 5 km (Yamada et al., 2018). Additionally, the differences for the historical climate simulation. between the time of peak discharge at the reference stations This study used the d4PDF downscaled rainfall data with (TOPD-RS) in each tributary and the time of peak water 5-km resolution from a previous study (Yamada et al., level at the confluence points (TOPWL-CP) in the main 2018). The 5-km resolution rainfall data were downscaled river are evaluated. The shorter the time difference, the from 20-km resolution data via the non-hydrostatic regional greater the flood risk. These results will provide additional climate model (Sasaki et al., 2011). The target period for information about the effect of climate change on the risk downscaling was set to 15 days of maximum rainfall for of river floods, and thus guide climate change adaptation each event in Hokkaido for 3,000 historical simulation and flood damage mitigation strategies in vulnerable areas. events and 5,400 future simulation events (Yamada et al., 2018; Hoshino et al., 2020). After downscaling, the rainfall METHODOLOGY amount, hourly rainfall intensity, and spatiotemporal distri‐ butions of rainfall were similar to those of the recorded Study area rainfall events. Additionally, the downscaled results can represent the topography of the study area more precisely The study area is the IRB. The river flows through 48 (Yamada et al., 2018). These results suggest that the dataset municipalities (including Sapporo, the prefectural capital), after downscaling can be used to evaluate the effect of cli‐ accounting for roughly 52% of Hokkaido’s population. The mate change on a regional scale. mean annual precipitation in the IRB is 1,300 mm In this study, we selected the rainfall data for locations (Ministry of Land, Infrastructure, Transport and Tourism of within the IRB. Then, we chose the annual maximum rain‐ Japan, 2004). The hydrologic peaks occur from March to fall (mm/72 h) (AMR-72h) for assessing the change in May during the snow-melt period, and in August and short-term extreme rainfall and its effect on river flooding September during the rainy season. At 268 km in length between the historical and future simulations in the IRB. and with a drainage area of 14,330 km2, the river is the According to the report (Japan River Association, 2003), longest in Hokkaido and the second largest in terms of the degree of safety for the Ishikari River in Hokkaido was basin area in Japan (Japan River Association, 2003). The set as 1/150, giving a return period of 150 years. Therefore, IRB has experienced severe damage from large-scale his‐ this study focused on assessing extreme river flooding torical floods. For example, the flood in August 2016 events for the top 20 and top 36 rainfall events (T20-T36- caused damage of approximately 260 million USD and REs) out of 3,000 and 5,400 rainfall events corresponding agricultural losses on 40,258 ha of land (Japan Society of to return periods equal to or more than 150 years for the Civil Engineers, 2017). Therefore, the projection of flood historical and future simulations, respectively. risk is significant to reduce future flood damage in this Hydrological model basin. Figure 1 shows the locations of the IRB and IM-SBs. d4PDF Dataset This study used the IFAS model developed by the Inter‐ national Centre for Water Hazard and Risk Management The “database for Policy Decision making for Future cli‐ (ICHARM). The IFAS uses a Public Works Research Insti‐ mate change” (d4PDF) contains data from numerous tute (PWRI)–distributed hydrological model developed in ensemble climate experiments with 60-km resolution on the the 1990s (Yoshino et al., 1990) as the runoff simulation global scale (Mizuta et al., 2016) and from a regional scale engine (ICHARM, 2014). A schematic of the IFAS model at 20-km resolution (Sasaki et al., 2011). The dataset spans is shown in Figure S1, and its parameter values are shown in Table SI. It has been used to estimate the flood risk for many river basins globally and has demonstrated good sim‐

141o0’0”E 142o0’0”E 143o0’0”E ulation performance (Aziz and Tanaka, 2010; Kimura et al.,

River basins Total Length 2014). Area(km2) (km) This study used the August 1981 flood to calibrate the Ishikari 14,330 268 hydrological model. The floods in September 2001, Uryu 1,722 177 September 2011, and August 2016 were chosen to validate Chubetsu 1,063 59 N ” Japan map Hokkaido island 0 0”N the model. These four extreme flood events were large-

’ Sorachi 2,618 194 0 ’ 0 o o 44 44 Ikushunbetsu 343 59 scale historical flood events that caused severe damage in Yubari 1,417 136 Uryubashi Akatsukibashi the IRB. The flood event in August 1981 was particularly Chitose 1,244 108 Akabira large, being the largest flood event observed in the IRB. Toyohira 902 72 The simulations were conducted for an additional period of

Lake Shikotsu two weeks prior to the main period of flooding events for River channel Nishikawamukai model warmup and to allow time for the water to reach the N

” Reference station Kiyohorobashi 0 0”N ’ 0 ’ 0 Kariki downstream area. o Confluence point Uranosawa o 3 3 4 4 Ishikari Ohashi station We calculated the river discharge at the Ishikari Ohashi

Ishikari river basin station (IOS), which is approximately 26.6 km upstream from the river mouth (Figure 1). The rainfall data for the 141o0’0”E 142o0’0”E 143o0’0”E August 1981 flood were obtained from rain gauge stations Figure 1. Ishikari river basin, Hokkaido, Japan provided by the Hokkaido Regional Development Bureau

—156— FLOOD RISK ASSESSMENT

(HRDB). The rainfall data for the September 2001, Then, future changes in extreme rainfall between the September 2011, and August 2016 floods, and the observed T20-T36-REs were evaluated for the IRB and IM-SBs. river discharges for the calibration and validation processes After validation, the IFAS model was used to estimate the at the IOS were obtained from the Ministry of Land, Infra‐ river discharge at 8 reference stations located in the IRB, structure and Transport, Japan (MLIT) (Ministry of Land, each IM-SBs, and 7 confluence points (Figure 1) for the Infrastructure, Transport and Tourism, 2020). Additionally, T20-T36-REs. Changes in extreme river flooding between the observed AMR-72h data (1926–2018) for the IRB were the T20-T36-REs in the IRB and IM-SBs were investi‐ provided by the Sapporo Development and Construction gated. Finally, the differences between the TOPD-RS in Department, HRDB (MLIT). each tributary and TOPWL-CP in the main river were eval‐ To quantitatively evaluate the performance of the IFAS uated. The average values of the time differences over all for the historical flood events, we used the Nash–Sutcliffe the flood events in each of the sub-basins were estimated coefficient (NS) (Nash and Sutcliffe, 1970) and the three and compared between the T20-T36-REs. indices of wave shape error (Ew), volume error (Ev), and peak discharge error (E ) (Japan Institute of Construction p RESULTS AND DISCUSSION Engineering, 2011; Aziz and Tanaka, 2010), defined as fol‐ lows. Hydrological model calibration and validation n 2 ∑i = 1 QM i − QC i NS = 1 − n 2 (1) Figure 2(a) compares the simulated and observed dis‐ ∑i = 1 QM i − QAVG i charges for the flood event of August 1981 after calibra‐ 2 tion. Figures 2(b), (c), and (d) compare the simulated and 1 n Q − Q E = M i C i (2) observed discharges in the validation process for flooding w n ∑ Q i = 1 M i events in September 2001, September 2011, and August n n 2016, respectively. The model calibration was performed ∑i = 1 QM i − ∑i = 1 QC i Ev = n (3) using a “trial and error” process. As shown in Figure 2, the ∑i = 1 QM i IFAS closely reproduced the flood duration and peak dis‐ charge in most cases. For the August 1981 flood, the simu‐ QMP − QCP Ep = (4) lated discharges showed close agreement with the observed QMP discharges, as indicated by a high NS value of 0.95, an Ew 3 Here, QM: observed discharge (m /s); Qc: simulated dis‐ of 0.08, Ev of 0.04, and Ep of –0.01. 3 charge (m /s); n: number of data points; QAVG: average In the validation process, the simulated discharges 3 observed discharge (m /s); QMP: peak value of the observed matched well with the observed discharges for the Septem‐ 3 discharge (m /s); and QCP: peak value of the simulated dis‐ ber 2001, September 2011, and August 2016 floods, with charge (m3/s). The simulation model is acceptable if NS > high NS values of 0.96, 0.92, and 0.90, respectively. The

0.7; and the smaller the Ew, Ev, and Ep errors are, the better simulated peak discharges were evaluated to be slightly the model is. lower than the observed peak discharges for the September Analytical procedure 2011 and August 2016 flood events. However, the statisti‐ cal performance indices suggested good performance for all This study was conducted to assess the effect of climate cases. The detailed indicators are shown in Table SII. These change on extreme rainfall and severe river flooding in the results suggest that the IFAS model can perform reasonably IRB using the following steps. First, the IFAS model was well for the IRB. calibrated and validated against historical flood events. Future changes in extreme rainfall Next, future changes in extreme rainfall were estimated based on the large ensemble rainfall dataset d4PDF. The Figure 3 shows the relative frequency of AMR-72h in Student’s t-test was used to find a significant difference the IRB for the observed rainfall data, and the historical between the two sets of samples (observed rainfall data and and future simulations. The results indicate that the rainfall data obtained from the historical simulation). The AMR-72h in the historical simulation and the rainfall hypothesis of the test is stated as follows. amount (mm/72 h) from the observation data have similar frequencies. The result from Student’s t-test demonstrates a Ho: μ1 = μ2 (5) H1: μ1 ≠ μ2 p-value of 0.214, which exceeds α. This result implies that Here, μ1 and μ2 are the means of the observed rainfall data the mean of the observed rainfall data is similar to the mean and rainfall data obtained from the historical simulation, of rainfall data obtained from the historical simulation. respectively. The null hypothesis is rejected if the p-value is Additionally, the rainfall amount is projected to increase less than the significance level α of 0.05. Additionally, the significantly in the future. The mean value of the rainfall Student’s t-test was used to find the significant difference amount in the future simulation is 86.3 (mm/72 h), which is between the two sets of samples (historical and future sim‐ 1.13 times that in the historical simulation (76.6 mm/72 h). ulations); the hypothesis of this test is stated as follows. The Student’s t-test result shows that the p-value is less than 2.2×10−16, i.e. much smaller than 0.05. This adequately Ho: μ2 ≤ μ1 (6) H1: μ2 > μ1 proves the validity of H1. Therefore, the frequency and Here, μ1 and μ2 are the means of AMR-72h in the histori‐ magnitude of extreme rainfall are expected to increase in cal and future simulations, respectively. The null hypothe‐ the future. sis is rejected if the p-value is less than the significance level α of 0.05.

—157— T.T. NGUYEN ET AL.

0.8

0.7

Min Mean Max 0.6 Observed 31.30 81.30 282.20 y c Historical simulation 9.57 76.55 371.99

n 0.5 e

u Future simulation 5.73 86.28 454.48 q e

r 0.4 f

e v i t

a 0.3 l e R 0.2

0.1

0 0-50 50-100 100-150 150-200 200-250 250-300 300-350 350-400 400-450 450-500 Annual maximum rainfall (mm/72h)

Figure 3. Relative frequency of AMR-72h in the Ishikari (a)

River basin for the observed rainfall data, and the historical 141o0’0”E 142o0’0”E 143o0’0”E and future simulations Legend

Ishikari River basin

N Uryu River basin ” 0 0”N ’ 0 0 ’

Future changes in extreme rainfall between the T20- o Chubetsu River basin o 44 T36-REs 44 Sorachi River basin 21.4% Ikushunbetsu River basin Tables SIII and SIV show the T20-T36-REs for the IRB Yubari River basin 14.1% and IM-SBs. In Figure 4(a), the AMR-72h is expected to Chitose River basin 21.2% increase in the IRB and all IM-SBs. The percentage differ‐ Toyohira River basin 15.5% ence in the median value of AMR-72h between the T20- River channel 29.9% N ” 0 0”N

T36-REs for the IRB and IM-SBs is shown in Figure 4(b). ’ 28.6% 0 0 ’ o o 43 18.7% 43 It is shown that the median 72-h value in the IRB is pro‐ 34.8% jected to be 21% higher in the future. Additionally, the extreme rainfall is expected to increase significantly in the 141o0’0”E 142o0’0”E 143o0’0”E Chitose, Ikushunbetsu, and Yubari river basins located in the southern part of the IRB, with estimated increases of Figure 4. (a) Boxplots of AMR-72h for the top 20 and top 35%, 30%, and 29%, respectively. Moreover, the spatial 36 rainfall events in the Ishikari River basin and in its main distributions of AMR-72h averaged for the T20-T36-REs, sub-basins, and (b) the percentage difference in the median and the percentage difference between the selected histori‐ of AMR-72h between the top 20 and top 36 rainfall events

—158— FLOOD RISK ASSESSMENT cal and future events are shown in Figure S2. These results depths. In particular, the Ikushunbetsu and Yubari river also indicate that the increase in extreme rainfall is basins located in the southern part of the IRB are expected expected to be particularly significant in the Ikushunbetsu to experience a significant increase in extreme rainfall and and Chitose river basins. river floods. However, the river floods from extreme rain‐ Future changes in river floods between the T20-T36- fall will increase to a greater extent. REs Time difference prediction Tables SV and SVI show the results of peak discharge The difference between the TOPD-RS in each tributary for the T20-T36-REs at 8 reference stations in the IRB and and the TOPWL-CP in the main river was estimated for the IM-SBs. Figure 5(a) shows boxplots of the peak runoff T20-T36-REs. Figure 6 shows the time difference averaged depth (mm/h) for the T20-T36-REs, and Figure 5(b) shows for the T20-T36-REs in each of the sub-basins. The results the percentage difference in the median of the peak runoff show a slight increase of 0.19 h and 0.70 h in the future depth between the T20-T36-REs in the IRB and IM-SBs. time difference in the Uryu and Sorachi river basins, These results indicate that the river floods were projected to respectively. Conversely, the time differences in the increase significantly in the IRB and IM-SBs. In particular, Ikushunbetsu, Yubari, Chitose, and Toyohira river basins the Ikushunbetsu, Yubari, and Toyohira river basins were are expected to decrease by 1.30, 2.14, 1.02, and 0.84 h, more likely to experience extremely large river flooding respectively. The time difference in the Chubetsu river events with peak runoff depths exceeding 25 mm/h. basin was almost unchanged with a slight decrease of The percentage difference in the median of the peak 0.05 h. The shorter the time difference, the greater the flood runoff depth between the T20-T36-REs indicated that the risk. These results can serve as a useful additional reference peak runoff depth is projected to increase by 33% in the for flood damage mitigation strategies in vulnerable areas. IRB. The Ikushunbetsu, Uryu, Sorachi, and Yubari river Particular attention should be paid to the Chitose River basins are expected to undergo remarkable increases of basin because it is a lowland area prone to flood damage. 56%, 54%, 54%, and 53%, respectively, in peak runoff Limitations Because our approach focused on natural hazards, we ignored future changes in land-use activities and popula‐ tions. Additionally, we did not consider the existing flood control facilities in the basin. However, we predict that short-term extreme rainfall events will have much greater effects on river flooding than the aforementioned factors.

CONCLUSIONS

The following conclusions were drawn from this study: – Following validation, the IFAS model can provide reasonable simulations of river discharges in the IRB. – Severe rainfall and river flooding events are expected to increase significantly in the IRB and IM-SBs in the (a) future. Additionally, the river flooding resulting from 141o0’0”E 142o0’0”E 143o0’0”E

Legend

Ishikari River basin 140o0’0”E 141o0’0”E 142o0’0”E 143o0’0”E

N Uryu River basin

” Sub_basins Distance from the reference 0 0”N ’ 0 ’ 0 station to the confluence point

o Chubetsu River basin o

44 44 Toyohira 11.1 km Sorachi River basin N ” 0 0”N 53.8% ’ Chitose 15.0 km 0 0 ’ o Ikushunbetsu River basin o 44 Yubari 9.8 km Historical: 3.25 h 44 Yubari River basin 15.9% Ikushunbetsu 7.0 km Future: 3.44 h Historical: 3.55 h Chitose River basin 33.3% Sorachi 17.5 km Future: 3.50 h Toyohira River basin 53.5% Chubetsu 18.1 km Historical: 9.05 h Historical: 2.30 h River channel 55.9% Uryu 6.0 km Future: 7.75 h Future: 3.00 h

N Reference station ” N 0 0”N Historical: 9.40 h Historical: 8.20 h ” ’ 52.6% 0 0”N 0 ’ 0 ’ o o Confluence point 0 Future: 8.56 h Future: 6.06 h 0 ’ o o 43 43 43.7% ) 43 43 Ishikari Ohashi station s 37.0% / Historical: 5.55 h 3 Lake Shikotsu m (

Future: 4.53 h e (m) g r

Time difference: l 141o0’0”E 142o0’0”E 143o0’0”E hour (h) = t_confluence point - t_reference station lev e

disch a

3 te r

Peak discharge at the reference station (m /s) a ive r Figure 5. (a) Boxplots of the peak runoff depth (mm/h) for Time (hour) W Peak water level at the confluence point (m) R the top 20 and top 36 rainfall events in the Ishikari River 140o0’0”E 141o0’0”E 142o0’0”E 143o0’0”E basin and in its main sub-basins, and (b) the percentage dif‐ ference in the median of peak runoff depth between the top Figure 6. Time difference averaged for the top 20 and top 20 and top 36 rainfall events 36 rainfall events in each of the sub-basins

—159— T.T. NGUYEN ET AL. extreme rainfall would increase to a greater extent than the Hoshino T, Yamada TJ. 2017. Assessment of climate change increase in extreme rainfall in the IRB and IM-SBs. The effect on precipitation amount within river domains of effect of climate change is significant in the sub-basins Hokkaido. Proceedings of 2017 Annual Conference, Japan located in the southern part of the IRB, where extreme rain‐ Society of Hydrology and Water Resources 12. DOI: fall is expected to increase by 29%–35%, whereas the river 10.11520/jshwr.30.0_12. discharge is likely to increase drastically by 37%–56%. The Hoshino T, Yamada TJ, Kawase H. 2020. Evaluation for charac‐ difference between the TOPD-RS and TOPWL-CP is teristics of tropical cyclone induced heavy rainfall over the expected to decrease by 1.02 h to 2.14 h in these regions. sub-basins in the central Hokkaido, northern Japan by 5-km Special attention should be paid to the Chitose River large ensemble experiments. Atmosphere 11: 435. DOI: basin, as it is located in a lowland area of the IRB that is 10.3390/atmos11050435. prone to flood damage. These results will provide addi‐ International Centre for Water Hazard and Risk Management tional information about the effect of climate change on the (ICHARM). 2014. IFAS ver 2.0 Technical manual. http:// risk of river flood for establishing climate change adapta‐ www.icharm.pwri.go.jp/research/ifas/ifas_2.0_top.html. Last tion and flood damage mitigation strategies in vulnerable access October 20, 2018. areas. Intergovernmental Panel on Climate Change (IPCC). 2012. Man‐ aging the Risks of Extreme Events and Disasters to Advance ACKNOWLEDGMENTS Climate Change Adaptation; A Special Report of Working Groups I and II of the IPCC; Field CB, Barros V, Stocker TF, Qin D, Dokken DJ, Ebi KL, Mastrandrea MD, Mach KJ, This study was supported by MEXT/SICAT. We utilized Plattner GK, Allen SK (eds). Cambridge University Press, the “policy Decision making for Future climate change” Cambridge, UK, New York, USA. (d4PDF) database. Intergovernmental Panel on Climate Change (IPCC). 2013. The physical science basis in the fifth assessment report of the SUPPLEMENTS intergovernmental panel on climate change. IPCC. Cam‐ bridge University Press, Cambridge, UK; 1535. Japan Institute of Construction Engineering (JICE). 2011. Outflow Figure S1. Schematic of the IFAS model analysis system, analytical technique manual Ver 2.3. http:// Figure S2. Spatial distribution of AMR-72h averaged for www.jice.or.jp/. Last access April 15, 2019. the (a) top 20 rainfall events and (b) top 36 rainfall Japan River Association. 2003. The Ishikari River. http://www. events; (c) percentage difference between the selected japanriver.or.jp/EnglishDocument/DB/file/002%20Hokkaido historical and future events %2003.pdf. Last access May 20, 2019. Table SI. Parameter value settings in the IFAS model Japan Society of Civil Engineers (JSCE). 2017. Onsite Field Table SII. Performance of the IFAS model in the IRB Investigation Team, Site-Investigation Report for the Table SIII. Top 20 rainfall events in the IRB and IM-SBs Hokkaido Heavy Rain Disasters in August 2016, JSCE Table SIV. Top 36 rainfall events in the IRB and IM-SBs Report, JSCE: Tokyo, Japan (in Japanese; Title was trans‐ Table SV. Peak discharge results for the top 20 rainfall lated by Kimura N). events in the IRB and IM-SBs Kim S, Nakakita E, Tachikawa Y, Takara K. 2010. Precipitation Table SVI. Peak discharge results for the top 36 rainfall changes in Japan under the A1B climate change scenario. events in the IRB and IM-SBs Annual Journal of Hydraulic Engineering, JSCE 54: 127– 132. REFERENCES Kimura N, Chiang S, Wei H, Su Y, Chu J, Cheng C, Jiou J, Chen Y, Lin L. 2014. Tsengwen reservoir watershed hydrological flood simulation under global climate change using the Aziz A, Tanaka S. 2010. Regional parameterization and applica‐ 20 km mesh meteorological research institute atmospheric bility of Integrated Flood Analysis System (IFAS) for flood general circulation model (MRI-AGCM). Terrestrial, Atmo‐ forecasting of upper-middle Indus River. Pakistan Journal of spheric and Oceanic Sciences Journal 25: 449–461. DOI: Meteorology 8: 21–38. 10.3319/TAO.2014.01.02.01(Hy). Doocy S, Daniels A, Murray S, Kirsch TD. 2013. The human Kitoh A, Ose T, Kurihara K, Kusunoki S, Sugi M, KAKUSHIN impact of floods: A historical review of events 1980–2009 Team-3 Modeling Group. 2009. Projection of changes in and systematic literature review. PLoS currents 5. DOI: future weather extremes using super-high-resolution global 10.1371/currents.dis.f4deb457904936b07c09daa98ee8171a. and regional atmospheric models in the KAKUSHIN Pro‐ Higashino M, Stefan HG. 2019. Variability and change of precipi‐ gram: Results of preliminary experiments. Hydrological tation and flood discharge in a Japanese river basin. Journal Research Letters 3: 49–53. DOI: 10.3178/hrl.3.49. of Hydrology: Regional Studies 21: 68–79. DOI: 10.1016/ Ministry of Land, Infrastructure, Transport and Tourism (MLIT). j.ejrh.2018.12.003. 2004. Ishikari River System Basic Policy for River Improve‐ Hirabayashi Y, Kanae S. 2009. First estimate of the future global ment (in Japanese; Title was translated by Thu NT). https:// population at risk of flooding. Hydrological Research www.hkd.mlit.go.jp/sp/kasen_keikaku/kluhh4000000dbmr-att/ Letters 3: 6–9. DOI: 10.3178/hrl.3.6. kluhh4000000dd33.pdf. Last access September 8, 2020. Hirabayashi Y, Kanae S, Emori S, Oki T, Kimoto M. 2008. Ministry of Land, Infrastructure, Transport and Tourism (MLIT). Global projections of changing risks of floods and droughts 2020. Water Information System. http://www1.river.go.jp/. in a changing climate. Hydrological Sciences Journal 53: Last access September 20, 2020. 754–772. DOI: 10.1623/hysj.53.4.754. Mizuta R, Murata A, Ishii M, Shiogama H, Hibino K, Mori N,

—160— FLOOD RISK ASSESSMENT

Arakawa O, Imada Y, Yoshida K, Aoyagi T, Kawase H, under climate change scenarios in the Yang River basin, Mori M, Okada Y, Shimura T, Nagatomo T, Ikeda M, Endo Thailand. International Journal of Sustainable Built Envi‐ H, Nosaka M, Arai M, Takahashi C, Tanaka K, Takemi T, ronment 6: 285–298. DOI: 10.1016/j.ijsbe.2016.09.006. Tachikawa Y, Temur K, Kamae Y, Watanabe M, Sasaki H, Tachikawa Y, Takino S, Ichikawa Y, Shiiba M. 2009. Study on Kitoh A, Takayabu I, Nakakita E, Kimoto M. 2016. Over the impact of climate change on river flow regime in the 5,000 years of ensemble future climate simulations by 60- Mogami and Yoshino River basins. Annual Journal of km global and 20-km regional atmospheric models. Bulletin Hydraulic Engineering 53: 475–480 (in Japanese). of the American Meteorological Society 98: 1383–1398. Try S, Tanaka S, Tanaka K, Sayama T, Hu M, Sok T, Oeurng C. DOI: 10.1175/BAMS-D-16-0099.1. 2020. Projection of extreme flood inundation in the Mekong Nash J, Sutcliffe J. 1970. River flow forecasting through concep‐ River basin under 4K increasing scenario using large ensem‐ tual models part I-A discussion of principles. Journal of ble climate data. Hydrological Processes 34: 4350–4364. Hydrology 10: 282–290. DOI: 10.1016/0022-1694(70) DOI: 10.1002/hyp.13859. 90255-6. Yamada TJ. 2019. Adaptation measures for extreme floods using River Council for Social Infrastructure Development. 2018. Coun‐ huge ensemble of high-resolution climate model simulation cil Report Flood Risk Management for Wide-area and Long- in Japan. Summary report on the eleventh meeting of the lasting Rainfall − Multi-layered countermeasures for complex research dialogue. United Nations Framework Convention disasters. https://www.mlit.go.jp/river/mizubousaivision/ on Climate Change (UNFCCC), Bonn, Germany: 28–29. toushin_e/1812_flood_risk_management_e.pdf. Last access https://unfccc.int/sites/default/files/resource/RD11Summary July 1, 2020. Report.pdf. Last access April 8, 2020. Sasaki H, Murada A, Hanafusa M, Oh’izumi M, Kurihara K. Yamada TJ, Hoshino T, Masuya S, Uemura F, Yoshida T, Omura 2011. Reproducibility of Present Climate in a Non- N, Yamamoto T, Chiba M, Tomura S, Tokioka S, Sasaki H, Hydrostatic Regional Climate Model Nested within an Hamada Y, Nakatsugawa M. 2018. The influence of climate Atmosphere General Circulation Model. SOLA 7: 173–176. change on flood risk in Hokkaido. Advances in River Engi‐ DOI: 10.2151/sola.2011-044. neering 24: 391–396 (in Japanese). Sato Y, Kojiri T, Michihiro Y, Suzuki Y, Nakakita E. 2012. Esti‐ Yoshino F, Yoshitani J, Sugiura M. 1990. A conceptual distributed mates of climate change impact on river discharge in Japan model for large-scale mountainous basins. In Lang H, Musy based on a super-high-resolution climate model. Terrestrial, A (eds). Hydrology in Mountainous Regions. I − Hydrologi‐ Atmospheric and Oceanic Sciences 23: 527–540. DOI: cal Measurements; the Water Cycle; IAHS, Lausanne, 10.3319/TAO.2012.05.03.02(WMH). Switzerland 193: 685–692. Shrestha S, Lohpaisankrit W. 2017. Flood hazard assessment

—161—