Hydrological alterations by dams and climate change can reduce aquatic biodiversity by disrupting the life cycles of organisms. Here, we aimed to evaluate and compare the hydrological alterations caused by dams and climate change throughout the Omaru River catchment, Japan, using a distributed hydrological model (DHM). First, to assess the impacts of dam and climate change independently, we performed runoff analyses using either dam discharge or future climatic data (two future periods, 2031–2050 and 2081–2100 × three representative concentration pathways). Subsequently, we derived indicators of hydrologic alterations (IHA) to quantify changes in flow alterations by comparing them to IHA under natural conditions (i.e., without dam or climate change data). We found that dams altered IHAs more than climate change. However, on a catchment-scale standpoint, climate change induced wider ranges of flow alterations, such as a further decrease in low flow metrics along the tributaries and uppermost main stem, suggesting a catchment-level shrinkage in important corridors of aquatic organisms. We also observed that the altered flow by water withdrawals was ameliorated by the confluence of tributaries and downstream hydropower outflows. Our approach using a DHM captured the various patterns of flow alterations by dams and climate change.

  • The hydrologic model is used to assess flow alterations by dams and climate change.

  • Compared with climate change, dams altered most environmental flow metrics.

  • Climate change decreased low flow metrics in upland streams and tributaries.

  • Ameliorations of the natural flow regime by hydropower outflows remained insufficient.

  • Prediction of spatial patterns of flow regime alterations is useful to river managers.

Natural flow regimes in rivers worldwide have been altered through flood control, water resource development, and climate change (Poff et al. 1997; Nilsson et al. 2005). Researchers have found that changes in flow regimes can decrease riverine biodiversity through the disruption of life cycles and habitat degradation (Poff et al. 1997; Lytle & Poff 2004). Hence, it is necessary to understand the extent of the changes in the flow regimes resulting from anthropogenic factors. Dams are anthropogenic factors that greatly modify natural riverine flow with different alteration patterns according to their type (Richter et al. 1996; Nislow et al. 2002). For example, dams for flood control suppress seasonal fluctuations of natural flow by truncating peak discharges during potentially devastating flooding events, resulting in less variable flow patterns (Munn & Brusven 1991). In addition, water withdrawal by large dams (e.g., for hydropower generation) often creates a section where the river flow is dramatically decreased (Li et al. 2017; Nukazawa et al. 2020). On the other hand, due to changing climates, extreme and frequent rainfall events, which have been observed recently in Japan, have triggered unexpected magnitude of floods and led to changes in river flow regimes (Sato et al. 2012). Döll & Zhang (2010) demonstrated that climate change has a greater impact on ecologically relevant river flow characteristics than dams and water withdrawals. Therefore, quantifying the extent to which dams and climate change alter flow regimes is a central challenge for river managers to safeguard riverine environments.

There have been many attempts to evaluate changes in flow regimes, although most have been limited to specific sites where flow data are available, such as the outlet of dams or reaches with a gauging station (Richter et al. 1998; Larned et al. 2011; Nukazawa et al. 2020). However, at a certain spatial scale, such as a catchment scale, attempts which rely on local flow data provide spatially insufficient information on river flow alteration as the alteration can be exacerbated or mediated through additional abstractions or convergences of small-to-large tributaries, respectively. To fill this gap, the application of distributed hydrological models (DHMs) is a promising approach. DHMs reflect spatial information, including altitude and land use/land cover, to estimate hydrological processes throughout a catchment of interest. Therefore, by using DHMs to obtain longitudinal profiles of river flow data at a given period and catchment, we can infer the spatial patterns of flow regimes and their alterations caused by dams/weirs and climate change. To date, DHMs have been used to evaluate changes in the flow regimes by dams and weirs (Ryo et al. 2015; Jardim et al. 2020; Mineda et al. 2020). Although many studies have evaluated the impact of climate change on environmental flows (Mahmoodi et al. 2021), to the best of our knowledge, those evaluations have been limited to specific sites such as reaches with a gauging station and estuary. Schneider et al. (2013) assessed the impact of climate change on environmental flows across Europe. Since they used a global scale model (5 arc min grid size), they evaluated differences in impacts between climate zones but did not grasp differences in impacts along segments in a catchment (i.e., up- and downstream gradients). Fatichi et al. (2015) have assessed the impact of climate change on spatial distributions of streamflow regimes such as minimum and maximum streamflow while considering dam operations, although the authors did not focus on potential changes in a variety of environmental flow metrics typically assessed using indicators of hydrologic alterations (IHA) or its equivalents. Since IHA considers five critical components of the flow regime regulating ecological process in rivers: the magnitude, frequency, duration, timing, and rate of change of hydrologic conditions, IHA provides more comprehensive information which is useful for assessing the impact of alteration on the ecological status in river (Richter et al. 1996). However, no study has evaluated the impact of climate change on environmental flows in an entire catchment using a distributed hydrological model (DHM) and its spatial heterogeneity.

Previous studies evaluating dam-induced flow alteration have generally adopted an approach that compares flow data before and after dam construction and defines the alteration as the degree of change in flow regimes in the presence of the dam (Zuo & Liang 2015; Faye 2018). However, as the calculation periods typically cover several decades (Lu et al. 2018), this approach could be subject to the effects of climate change (Cui et al. 2020). Consequently, the extent of flow regime alterations by dams cannot be appropriately evaluated. Therefore, it is of primary importance to separately assess the impacts of such anticipated anthropogenic factors (i.e., flow regulations and climate change) on environmental flow for adequate water resource management and environmental conservation (Goldstein & Tarhule 2015). However, few previous studies have proposed frameworks to separate the impacts of dams and climate change when quantifying the impacts of dams (Lu et al. 2018; Cui et al. 2020). For example, Cui et al. (2020) estimated the flow alteration due to dam construction by comparing observed pre-impact (i.e., before dam construction) and post-impact flow regime indicators value while excluding the impact of climate change estimated based on differences between the observed pre-impact value and post-impact value simulated by a hydrological model. However, the authors stated that the influences of all dams located in the study catchment and other human activities such as land-use change on flow regimes were not always addressed. In addition, previous works comparing pre- and post- impact periods remain to estimate indirect measures of alteration since the two periods may involve distinctive flow events even without any other potential anthropogenic effects.

The present study aims to evaluate and compare the changes in flow regimes caused by dams and climate change throughout a river catchment. First, we will apply a DHM to the Omaru River catchment in southwest Japan, which contains multiple dams and intake weir. Then, we will perform runoff analyses using dam discharge data to adequately quantify the spatial patterns of dam-induced flow alteration. Similarly, using future climatic data acquired from general circulation models (GCMs) or not, we will run a DHM to evaluate the impacts of climate change on environmental flow regimes in the study catchment. Finally, we will compare the extent and patterns of flow alteration between dams and climate changes. Our approach will provide important environmental implications as spatial extents of flow regime alteration by such major anthropogenic impacts.

Study area

We investigated the Omaru River (catchment area: 474 km2), which originates in the Sampo Mountains, flows 75 km east, and drains into the Pacific Ocean (Figure 1). The mean annual temperature, annual precipitation, and elevation are approximately 14.9 °C, 3,100 mm, and 250 m at the Mikado meteorological station, and approximately 17.6 °C, 2,300 mm, and 57 m at the Takanabe meteorological station in the downstream terrain (http://www.qsr.mlit.go.jp/miyazaki/kasen/omaru/gaiyou/omaru_saigai.html). The major land uses/land covers in the catchment are forests (∼87%), agricultural fields (∼10%), and urban areas (∼3%) (Figure S1(a)). There are two dams for flood control and hydropower generation (Dogawa and Matsuo Dams), the two dams only for hydropower generation (Tozaki and Ishikawauti Dams), as well as a dam for irrigation and hydropower generation in the catchment (Table S1). A weir for abstraction (Kijino Weir) is installed in the uppermost stream to supply river water to the Dogawa Dam reservoir. As key characteristics of each dam, the ages, heights, and reservoir capacities are, respectively, 65 years, 62.5 m, and 143,000 m3 for the Dogawa Dam, 70 years, 68 m, and 168,000 m3 for the Matsuo Dam, 78 years, 25 m, and 25,000 m3 for the Tozaki Dam, 14 years, 47.5 m, and 134,000 m3 for the Ishikawauti Dam, and 82 years, 23.6 m, and 34,000 m3 for the Matsuo Dam.
Figure 1

The Omaru River catchment (catchment area: 474 km2) with the spatial distributions of the studied dams, weir, hydropower stations, and meteorological and gauging stations. The main stem and tributaries handled as channel parts in hydrologic modeling are depicted by white and blue colors, respectively. Arrows indicate water conveyance from the dams to the hydropower stations. The outflow data time series available at all weir, dams, and hydropower stations are used for the model input. Observed daily discharge data at the Dogawa Dam, Matsuo Dam, and Takajo gauging station are used for the model validation.

Figure 1

The Omaru River catchment (catchment area: 474 km2) with the spatial distributions of the studied dams, weir, hydropower stations, and meteorological and gauging stations. The main stem and tributaries handled as channel parts in hydrologic modeling are depicted by white and blue colors, respectively. Arrows indicate water conveyance from the dams to the hydropower stations. The outflow data time series available at all weir, dams, and hydropower stations are used for the model input. Observed daily discharge data at the Dogawa Dam, Matsuo Dam, and Takajo gauging station are used for the model validation.

Close modal

Geographic data

We acquired digital elevation model (DEM) data at a spatial resolution of 250 m (Figure S1(b)) from the Ministry of Land, Infrastructure, Transportation, and Tourism. We used the spatial distribution of land use/land cover at a resolution of 100 m and converted its spatial resolution to 250 m (Figure S1(a)). Based on the DEM, the slope and flow directions in the catchment were estimated while correcting the pans (a mesh with lower elevation than the surrounding eight meshes), so the direction was properly determined in the analyses. To do so, the altitudes of pans were slightly raised repeatedly, and finally, corrected DEM and flow direction maps were created for use in subsequent runoff analyses.

Meteorological data

Observed precipitation data were acquired from Automated Meteorological Data Acquisition System (AMeDAS) data at three meteorological stations (Takanabe, Mikado, and Tsuno) inside and outside the Omaru catchment, and from the Dogawa and Matsuo Dams (Figure 1). Data on air temperature, wind speed, and sunshine duration were acquired from the AMeDAS data at three meteorological stations (Takanabe, Nishimera, and Mikado). For the cloud cover, atmospheric pressure, and humidity observation data, we used data from the Miyazaki Local Meteorological Observatory, which is the local meteorological station closest to the Omaru River catchment. The temperature, wind speed, and precipitation were spatially interpolated by averaging the weights of geographical distance based on the point data at the meteorological stations and were input over the study catchment.

For future climate data, we used eight GCMs (Table S2) provided by the Earth System Grid Federation (http://esgf-node.llnl.gov/search/cmip5). We acquired monthly surface temperature and precipitation data from each of the eight GCMs targeting the two future periods, the near future (2031–2050) and the far future (2081–2100). Future emissions and radiative forcing were considered using three representative concentration pathways (RCP2.6, RCP4.5, and RCP8.5; following numerals indicate anticipated radiative forcing around 2100). To adequately estimate regional temperature and precipitation in the future from GCMs that only have low spatial resolution, we eliminated and corrected systematic errors (bias). This setting allows to evaluate changes in local climate and hydrology as demonstrated in previous studies (e.g., Sulis et al. 2011; Tsegaw et al. 2020; Moraga et al. 2021). For details, see Text S1.

Distributed hydrological model

We developed a DHM for the studied catchment by slightly modifying the DHM originally developed for the Natori River catchment in northeast Japan (Figure S3) (Kazama et al. 2007, 2021; Nukazawa et al. 2011, 2015). The model's spatial resolution is 250 m. In brief, DHM estimates the direct flow, base flow, river channel flow, snow cover, and evapotranspiration using the kinematic wave model (Lighthill & Whitham 1995a, 1995b), storage function method (Kimura 1961), dynamic wave model (Liggett 1975), degree-day method (Martinec 1960), and Modified Penman–Monteith equation (Allen et al. 1998), respectively, at the channel and hillslope parts separately. For details, see Text S2.

A one-dimensional dynamic wave model was used in the channel parts. The channel parts were manually designated, including the mainstem, the Do River, and 15 tributaries (Figure 1). The dynamic wave model was operated using the backward difference method at the downstream ends (i.e., the meshes upstream of the estuary, dams, and weir), the forward difference method at the upstream ends (i.e., the meshes at the upstream ends of the channels and outlets of the dams and weir), and the central difference method at the other channel meshes. We provided the observed discharges from the dams and hydropower plants at the outlet meshes. In addition, the observed discharge bypassed from the Kijino Weir to the Dogawa Dam reservoir was given to an upstream mesh of the Dogawa Dam.

Model evaluation

The DHM was run from 2008 to 2019. A warm-up period of 2 years was allocated (2008–2009) prior to the 10-years evaluation period. Subsequently, the model parameters were determined based on a comparison of observed and simulated daily discharge data at an inflow mesh of the Dogawa Dam (hereafter the Dogawa inflow mesh), which was not affected by the boundary conditions (i.e., the observed dam and hydropower discharges). The Nash–Sutcliffe efficiency (NSE) (Nash & Sutcliffe 1970), percent bias (PBIAS) (Moriasi et al. 2007), and reproducibility of the hydrograph (magnitudes and timings of peak discharges and falling limbs) were used to evaluate the model performance. For details, see Text S3–4.

We validated the final model at three points (the Dogawa and Matsuo inflow meshes and the Takajo gauging station released at http://www1.river.go.jp/) from 2010 to 2019 using NSE and PBIAS. The model performance can be considered satisfactory if NSE ≥ 0.7 and PBIAS ≤ ±15% (Moriasi et al. 2007; Matsubara et al. 2015). Negative/positive PBIAS indicates that the model is overestimated or underestimated.

Evaluation of flow alterations

Runoff analysis without the boundary conditions of dams and weir

To quantify flow alterations by dams and climate change, we performed runoff analyses without boundary conditions (e.g., intake or bypassed discharge by dams and weir) which is a natural flow scenario eliminating the influence of dams (no-dam scenario) as well as the above-mentioned analyses with boundary conditions (dam scenario) from 2010 to 2019. Because discharge depends on upstream boundary conditions (i.e., input observed data of discharge from weir, dams, and hydropower stations), under the dam scenario, if the simulated flow is over- or underestimated upstream of dams, it does not propagate downstream. On the other hand, under the no-dam scenario, if the discharge is over- or underestimated upstream, it propagates downstream and influences the analysis. Therefore, we confirmed the 10-year average discharge difference and the ratio between the scenarios at meshes not affected by the boundary conditions of the dams, weir, and hydropower plants (e.g., upstream of the Kijino Weir), an upstream mesh from the Matsuo Dam, and the Takajo gauging station.

Runoff analyses under climate change

Using bias-corrected future climate data of air temperature and precipitation instead of current climate data, we performed hydrologic analyses to simulate hydrological conditions under climate change (hereafter, climate change scenario). This scenario eliminated the boundary conditions (e.g., dam discharges) to individually represent the impacts of climate change on the flow regimes in the study catchment. In total, 48 patterns of future climate data (8 GCMs × 2 future periods × 3 RCP scenarios) were used to create 48 patterns of future flow regime projections. Subsequent flow alteration assessments were carried out using mean discharges among the eight GCMs for each future period and the RCP scenario (i.e., with six patterns of future flow regime outputs).

Flow regime metrics

We used the IHAs to assess changes in the flow regimes caused by dams and climate change (Richter et al. 1996; iha, R ver. 3.6.3). The IHA is composed of 33 ecologically relevant indicators, which are classified into five groups: (1) magnitude of the monthly discharge; (2) magnitude and duration of the annual extreme flow; (3) timing of the annual extreme flow; (4) frequency and duration of low and high pulses; and (5) rate and frequency of flow changes (Table 1). The 33 IHA parameters provide a detailed representation of the flow regime and entail hydrologic statistics commonly employed in limnology studies because of their great ecological relevance (Richter et al. 1996). We derived the IHA from the daily average discharge for each calendar year of the study period. Because there was no zero-flow days in this catchment during the study period, we excluded zero-flow days from subsequent analyses. Using the following equations, we evaluated the extent of alteration of IHA in the Dam scenario and the climate change scenario compared to IHA in the no-dam scenario.

  • (i)
    given that the unit of IHA is m3 s−1, number, or days,
    (1)
  • (ii)
    given that the unit of IHA is day,
    (2)
    where both and are percent changes (% and day, respectively), is IHA in the no-dam scenario, is IHA in the dam scenario, and is IHA in the climate change scenario. We used the means of the percent change metrics of the 10 calendar years for subsequent evaluations of the flow regime alterations.
Table 1

The 32 indicators of hydrologic alterations (IHA) indicators used in this study

IHA parameters groupHydrologic parametersUnit
Group1 Magnitude of the monthly discharge Median monthly streamflow (m3 s−1
Group2 Magnitude and duration of the annual extreme flow 1, 3, 7, 30, 90 days min (m3 s−1
1, 3, 7, 30, 90 days max (m3 s−1
Baseflow index (m3 s−1
Group3 Timing of the annual extreme flow T-Min (day) 
T-Max (day) 
Group4 Frequency and duration of low and high pulse Low and high pulse number (Number) 
Low and high pulse duration (days) 
Group5 Rate and frequency of flow changes Rise and fall rate (m3 s−1
Reversals (Number) 
IHA parameters groupHydrologic parametersUnit
Group1 Magnitude of the monthly discharge Median monthly streamflow (m3 s−1
Group2 Magnitude and duration of the annual extreme flow 1, 3, 7, 30, 90 days min (m3 s−1
1, 3, 7, 30, 90 days max (m3 s−1
Baseflow index (m3 s−1
Group3 Timing of the annual extreme flow T-Min (day) 
T-Max (day) 
Group4 Frequency and duration of low and high pulse Low and high pulse number (Number) 
Low and high pulse duration (days) 
Group5 Rate and frequency of flow changes Rise and fall rate (m3 s−1
Reversals (Number) 

If the absolute value of percent change is greater than 20% or day, IHA is considered significantly altered (Richter et al. 2012; Yang et al. 2017). Furthermore, we counted the number of significantly altered IHA to evaluate the extent of the flow regime alterations; no flow alteration occurred when the number of significantly altered IHA was 0 (i.e., all IHA falls into 0–20 absolute value of percent change), while small, moderate, and large alterations occurred when the number of significantly altered IHA was 1–10, 11–20, and ≥21, respectively (Laizé et al. 2014; Yang et al. 2017). The percent changes of each IHA and the number of significantly altered IHA were derived for all river channel meshes (n = 555) and visualized throughout the catchment.

Model validation

The NSE and PBIAS for the 10 years evaluation period were 0.921 and 4.0% at the Dogawa Dam inflow mesh, 0.964 and 3.9% at the Matsuo Dam inflow mesh, and 0.957 and 4.0% at the Takajo gauging station (Table 2 and Figures 2 and S4), suggesting high accuracy of runoff modeling throughout the studied catchment while involving little error in the dam and hydropower outflow data.
Table 2

The Nash–Sutcliffe efficiency (NSE) and percent bias (PBIAS) for the Dogawa and Matsuo Dams and Takajo gauging station from 2010 to 2019

YearDogawa Dam
Matsuo Dam
Takajo
NSEPBIAS (%)NSEPBIAS (%)NSEPBIAS (%)
2010 0.906 3.1 0.968 4.2 0.975 4.9 
2011 0.938 −0.5 0.995 4.6 0.995 2.0 
2012 0.951 3.1 0.974 7.8 0.994 −3.5 
2013 0.883 −12.2 0.927 4.3 0.950 4.6 
2014 0.942 −3.1 0.958 9.8 0.982 11.0 
2015 0.870 6.9 0.975 −4.6 0.986 9.0 
2016 0.934 9.3 0.978 0.8 0.980 11.4 
2017 0.890 9.4 0.964 −1.1 0.995 7.9 
2018 0.945 13.1 0.949 5.8 0.997 2.4 
2019 0.782 6.3 0.933 3.6 0.989 −6.4 
10 years 0.921 4.0 0.964 3.9 0.957 4.0 
YearDogawa Dam
Matsuo Dam
Takajo
NSEPBIAS (%)NSEPBIAS (%)NSEPBIAS (%)
2010 0.906 3.1 0.968 4.2 0.975 4.9 
2011 0.938 −0.5 0.995 4.6 0.995 2.0 
2012 0.951 3.1 0.974 7.8 0.994 −3.5 
2013 0.883 −12.2 0.927 4.3 0.950 4.6 
2014 0.942 −3.1 0.958 9.8 0.982 11.0 
2015 0.870 6.9 0.975 −4.6 0.986 9.0 
2016 0.934 9.3 0.978 0.8 0.980 11.4 
2017 0.890 9.4 0.964 −1.1 0.995 7.9 
2018 0.945 13.1 0.949 5.8 0.997 2.4 
2019 0.782 6.3 0.933 3.6 0.989 −6.4 
10 years 0.921 4.0 0.964 3.9 0.957 4.0 
Figure 2

Hydrographs at the inflow meshes of (a) the Dogawa Dam and (b) the Matsuo Dam, and (c) Takajo gauging station from 2010 to 2019.

Figure 2

Hydrographs at the inflow meshes of (a) the Dogawa Dam and (b) the Matsuo Dam, and (c) Takajo gauging station from 2010 to 2019.

Close modal

At the mesh not affected by the boundary conditions (e.g., upstream of the Kijino Weir), there was no difference in the 10-year average discharges of the no-dam and dam scenarios. The discharges at the Matsuo Dam inflow mesh and Takajo gauging station under the no-dam scenario were 0.34 m3 s−1 (1.3%) and 0.63 m3 s−1 (1.9%) smaller than those under the dam scenario, respectively. Considering the small fractions of the differences in the average discharge, the hydrologic balance between the scenarios in the study catchment was mostly negligible.

Flow alteration by dams and weir

Figures 3 and 4(a) illustrate the number of significantly altered IHA caused by dams and weir in the catchment. We found moderate to large alterations (the number of significantly altered IHA ranged from 12 to 21) in the river section from downstream of the Kijino Weir to the mesh upstream of the Do River confluence, large alterations (25–27) in the river section downstream of the Dogawa Dam to the mesh upstream of the confluence with the mainstem, and moderate to large alterations (14–27) in river sections downstream of the Matsuo Dam and the river mouth. Tables 3 and S3–4 show the percent changes of each IHA and the number of significantly altered IHA at the selected meshes, for example, outlets of the weir, dams, and hydropower plants. The general patterns among the assessed meshes were negatively altered rise rate (the percent change ranged from −99.5 to −64.7%) and fall rate (−99.7 to −27.7%), and no significant alteration in the date of annual maximum flow (−15.1 to −3.4 days).
Table 3

The percent changes of selected indicators of hydrologic alterations (IHA) and the number of significantly altered IHA (# altered IHA) at the selected meshes; outlets of the weir and dams

WeirDam
IHAKijinoMatsuoKawabaru
Dec − 75.7 − 98.6 − 76.4 (%) 
90-d. min − 76.0 − 97.3 − 83.1 (%) 
1-d. max − 35.5 −11.9 −11.8 (%) 
90-d. max −6.7 − 39.7 − 37.2 (%) 
T-Max −5.3 −3.5 −11.5 (day) 
High PN − 46.3 − 23.0 −2.9 (%) 
Low PD 8.8 30.9 − 27.8 (%) 
High PD 39.2 34.5 −9.6 (%) 
Rise rate − 91.2 − 99.5 − 97.3 (%) 
Fall rate − 99.7 − 99.5 − 98.2 (%) 
Reversals −3.0 −13.7 13.1 (%) 
# altered IHA 21 25 24  
WeirDam
IHAKijinoMatsuoKawabaru
Dec − 75.7 − 98.6 − 76.4 (%) 
90-d. min − 76.0 − 97.3 − 83.1 (%) 
1-d. max − 35.5 −11.9 −11.8 (%) 
90-d. max −6.7 − 39.7 − 37.2 (%) 
T-Max −5.3 −3.5 −11.5 (day) 
High PN − 46.3 − 23.0 −2.9 (%) 
Low PD 8.8 30.9 − 27.8 (%) 
High PD 39.2 34.5 −9.6 (%) 
Rise rate − 91.2 − 99.5 − 97.3 (%) 
Fall rate − 99.7 − 99.5 − 98.2 (%) 
Reversals −3.0 −13.7 13.1 (%) 
# altered IHA 21 25 24  

Values are boldfaced if absolute values are larger than 20. PN and PD represent pulse number and pulse duration, respectively.

Figure 3

The number of significantly altered IHA (# altered IHA) along the geographical distances from the upstream ends of the (a) Omaru River and (b) Do River under the dam scenario and climate change scenarios. The hydropower plants located on the Omaru River are, in order from upstream, Ishikawauti-Daiichi, Ishikawauti-Daini, and Kawabaru Power Plants.

Figure 3

The number of significantly altered IHA (# altered IHA) along the geographical distances from the upstream ends of the (a) Omaru River and (b) Do River under the dam scenario and climate change scenarios. The hydropower plants located on the Omaru River are, in order from upstream, Ishikawauti-Daiichi, Ishikawauti-Daini, and Kawabaru Power Plants.

Close modal
Figure 4

The spatial distributions for the number of significantly altered IHA under the (a) Dam scenario, (b) RCP4.5 in the near future, and (c) RCP8.5 in the far future. Arrows indicate water conveyance from the dams to the hydropower stations.

Figure 4

The spatial distributions for the number of significantly altered IHA under the (a) Dam scenario, (b) RCP4.5 in the near future, and (c) RCP8.5 in the far future. Arrows indicate water conveyance from the dams to the hydropower stations.

Close modal

Large alterations (number of significantly altered IHA ranged from 22 to 26) were detected along river sections where the river water was abstracted (Figures 3 and 4(a) and Tables 3 and S3). We found large negative alterations in all the median monthly streamflows (the percent change ranged from −98.6 to −48.6%), minimum flows (−98.7 to −42.3%), and maximum flows at longer time-windows (i.e., 30–90 days) (−51.7 to −25.1%) (Tables 3 and S3). The percent changes in the maximum flows at shorter time-windows (i.e., 3–7 days) showed significant negative alterations (−28.1 to −23.0%) at the Dogawa and Tozaki Dams. For the base flow index, the Dogawa, Matsuo, and Tozaki Dams had large negative alterations (−96.4 to −50.4%). The patterns of alterations in the pulse metrics differed depending on the dams. For example, while the high pulse number was negatively altered at the Matsuo and Tozaki Dams (−23.0 and −45.4%, respectively), the high pulse duration was positively altered at Matsuo Dam (34.5%). We found large negative alterations in the low flow metrics, such as the median monthly flows in autumn to winter (i.e., November to February) and the 90-day minimum downstream of the Kijino Weir (−81.0 to −74.2%). On the other hand, negative alterations in the high flow metrics, such as some median monthly flows in rainy seasons (i.e., May to October) and the maximum flows (i.e., 7–90 days) were suppressed (−16.6 to 6.1%). This result is consistent with that of a previous study in Taiwan; weir intake reduced low flows rather than high flows (Shiau & Wu 2004). This is ascribed to the smaller amount of water abstraction at the Kijino Weir than at the dams.

Despite the combined outflow from hydropower plants, moderate to large flow alterations were observed at the outlet meshes of the plants (# altered IHA ranged 17–27) (Figures 3 and 4(a) and Table S4). As a global trend, the median monthly streamflow in rainy seasons such as May and August (the percent change ranged from 40.3 to 299.1%), high pulse duration (36.5–110.8%), and the reversals (26.8–44.9%) exhibited positive alterations at these meshes, while the high pulse number showed negative alterations (−58.6 to −49.1%). The Dogawa hydropower station displayed distinct patterns of alterations, for example, positive alterations in the median flow in February, June, December, and 90-day minimum (26.2 to 56.2%). The discharge bypassed from the Kijino Weir increased the discharge in the Do River, resulting in positive alterations in the median monthly streamflow. The Kawabaru hydropower station also displayed different patterns of alterations, such as negative alterations in the date of annual minimum flow (−67.1 days) and 90-days minimum (−22.2%).

The combined outflow from the hydropower stations, however, ameliorated the severe negative alterations in the minimum and maximum flows that occurred in the sections where the river flow dramatically decreased downstream of the dams, and even caused positive alterations (Table S4). For instance, the positive alteration of the low pulse number at the Kawabaru hydropower station was ameliorated (from 46.2 to 15.7%), while negative alteration occurred at the Ishikawauti-Daiichi hydropower station (from 11.49 to −22.2%).

We found that the confluence of tributaries with varying catchment areas ameliorated flow alterations; that is, the percent changes of each IHA approached 0 and the number of significantly altered IHA decreased throughout the catchment (Figure 4(a) and Table S5). Downstream of the Kijino Weir, the confluences of the Matanoe and Mizushidani rivers ameliorated the negative alterations in the minimum flows and reduced the number of significantly altered IHA from 20 to 15. In the section between the Kawabaru hydropower station and the river mouth, the tributaries such as Kirihara River ameliorated alterations in flow metrics (e.g., the negative alteration in the 30-days maximum and the positive alteration in the low pulse number), which led to a decrease in the number of significantly altered IHA from 18 to 14. A previous study has successfully demonstrated mathematically that the alteration of a maximum flow metric (equivalent to the 1-day maximum flow) was ameliorated by a confluence of tributary (Volpi et al. 2018; Cipollini et al. 2022). On the one hand, the amelioration phenomena observed in our physically based model could be illustrated explicitly (Text S5), suggesting that the amelioration is a general phenomenon regardless of configurations of alterations and confluences (e.g., spatial allocations of withdrawal and tributaries), such observed in a previous study that focused on flow duration curve metrics (e.g., Q75; Mineda et al. 2020). However, we found the exacerbated flow alteration such as February median at the Kirihara River, low PN at the Matanoe River, and high pulse metrics at the Do River. These cases are explained by the timing at which the indicators are calculated under each scenario does not match or flow in the tributary is altered due to anthropogenic factors (i.e., climate change and water conveyance at the Do River in the present study) (Text S5).

Flow alteration under future climate change

Under all climate change scenarios (two future periods × three RCPs), a limited number of IHA were projected to be altered (0–10 IHA) (Figures 4(b) and 4(c) and S5). Under RCP2.6 and RCP4.5, fractional alteration was projected in the partial river meshes. A larger number of significantly altered IHA were projected at most river meshes under RCP8.5, compared to RCP2.6 and RCP4.5. Figure 5 shows the percent changes of selected IHA from upstream to downstream along the mainstem under each climate change scenario. Note that the variation in the percent changes, here climatic change-induced flow alteration, is prone to be large because the discharges at uppermost meshes are extremely small.
Figure 5

The percent changes of selected IHA from up- to downstream along the mainstem under each climate change scenario.

Figure 5

The percent changes of selected IHA from up- to downstream along the mainstem under each climate change scenario.

Close modal

Decreasing trends were identified in the January, April, August, November, and December streamflow throughout the catchment, specifically upstream, under all climate change scenarios (Figures 5 and S6). This was most apparent under RCP8.5 in the far future; the percent changes of the January, April, November, and December flows reached around −20% throughout the catchment. The negative alterations for the low flow metrics such as minimum flows and base flow index were more prominent upstream and, in the tributaries where the discharges were smaller (Figures 5 and S6–7).

Low flows such as minimum flows at longer time-windows (i.e., 90 days) were projected to decrease (−17.5 to −7.0%) the most throughout the mainstem under RCP8.5 in the far future (Figure S7), whereas such a decline was rarely observed in the other scenarios. High flows such as September and maximum flows were projected to increase upstream and in the tributaries under many climate change scenarios. Specifically, under RCP8.5 in the near future, high flows increased markedly, characterized by positive alterations in the 1-day maximum (22–30%) upstream (Figure S8). However, the 1-day maximum under RCP8.5 in the near future would be overestimated due to an outlier projected with the CSIRO-MK3.6 (Figure S9). In contrast, the maximum flow metrics were projected to decrease throughout the catchment under RCP2.6 in the far future.

In the river section downstream of the Kawabaru hydropower station to the river mouth, which is surrounded by relatively populated cities, high flows at shorter time-windows, such as 1-day max, increased by 9.4–11.6% under RCP8.5 in the near future, while it declined (−7.6 to 5.3% under RCP2.6 in the far future and RCP4.5 in the near future) or was unchanged in the other scenarios (Figures S6 and S8).

For the date of annual minimum and maximum flow, little change was projected throughout the catchment under all climate change scenarios (Figure S6). However, in the Do River, the annual minimum flow was significantly altered at 41 meshes under all climate change scenarios with varying percent changes from −273 to 197 days (Figure S10). The low and high pulse numbers were projected to vary spatially under all climate change scenarios (Figures 5 and S6). The percent changes were approximately ±10% for the low pulse number and ±15% for the high pulse number. The low and high pulse durations were significantly altered under many climate change scenarios with varying percent changes depending on the climate change scenario and location. The highest percent changes were projected for the high (75.0%) and low pulse duration (50.0%) under RCP8.5 in the far future. The percent change of rise rate tended to vary under RCP8.5 in the far future; reaching a highest value of 46.5% around the uppermost meshes and lowest value of −24.9% near the Do River confluence. On the other hand, the percent changes in fall rate and reversals tended to fall negative under most climate change scenarios (Figures 5 and S6).

Evaluation of environmental flow alterations using DHM

Our catchment-scale evaluations of the environmental flow alterations based on a distributed hydrologic model enable us to understand the overall picture of such alterations due to various water intakes/supplies as well as climate change. Previous studies have attempted to separate the impacts of complex factors (e.g., dam and climate change) and estimate contributions of a given factor to flow alteration (e.g., Cui et al. 2020). However, because the dam-induced impact on flow regimes was estimated by subtracting those induced by climate change from overall impacts, it involves other withdrawal or anthropogenic factors, and thereby, is difficult to identify the impacts of dam and climate change. Moreover, since the cumulative impacts of many small dams on flow regimes are significant or cannot be ignored (Deitch et al. 2013; Kibler & Tullos 2013; Lu et al. 2018), all dams located in a study catchment should be considered as possible for accurately quantifying the impact of dam. Therefore, earlier works have not yet sufficiently considered a separation of these impacts, highlighting a uniqueness or novelty of our approach, which inputs the impacts of dams and climate change on the model independently and compares the resulting environmental flow alterations. The independent analyses can directly quantify each impact of dam and climate change without other impact factors. In addition, our approach rigorously considered the impacts of major water withdrawals and hydropower discharges in the study catchment, which enables us to quantify and visualize spatial patterns of the impacts of these manipulations. Therefore, our approach may contribute to seek effective counter measures of climate change (Haddeland et al. 2014).

Our approach evaluated the spatial heterogeneity of changes in the flow regimes by dam and climate change in the entire catchment, which is novel results and useful for understanding the river environmental management. We demonstrated that the flow alterations caused by water withdrawals were ameliorated by the confluence of tributaries and hydropower outflows downstream (e.g., downstream sections of the Kijino Weir and Kawabaru Dam; Figure 4(a)). Furthermore, we detected the localized impacts of climate change on upstream flow alterations (Figures 5 and S6). In particular, the impacts of climate change on the low and high pulse metrics and the rise/fall rates were highly spatially variable; hence, attention should be given to local alteration patterns of these flow metrics in cases of catchment management and future studies. In this sense, developing our model at a finer spatial resolution would be a promising future work to better capture localized patterns of flow regime alteration in uppermost small streams.

Although our hydrologic model showed high accuracy throughout the study period, the flow peak extremes tended to be underestimated at all gauging sites (Figure 2). This is presumably because the limited information on rainfall inputs (i.e., point data from the five meteorological stations) could not reflect the spatial and temporal heterogeneity of heavy rainfall, such as during typhoons and torrential rainfall events. Consequently, some IHA in terms of high flows might contain relatively larger uncertainties than others.

Dam-induced flow alteration

Large negative alterations were detected in the median monthly streamflow, the maximum flows, especially at the longer time-windows (30–90 days maximum), and the minimum flows at all the sections where the river water was abstracted (Tables 3 and S3). Such decreases in streamflow could affect the community structures of many organisms, including fish and benthic animals (e.g., Tipulidae, Baetidae, and Heptageniidae) inhabiting the studied catchment, by decreasing the habitable area downstream of dams and weirs (Nukazawa et al. 2020). All studied dams were used for water abstractions (mainly for power generation), while the Dogawa and Matsuo Dams were also used for flood control. The decreases in low flows (e.g., minimum flows) and maximum flows are characterized by dams used for hydropower generation and flood control, respectively (Lu et al. 2018). Because the effects of flood control at the Dogawa and Matsuo Dams propagated downstream, both characteristics were observed in the downstream sections of all dams.

In the section downstream of the Tozaki Dam, the high pulse duration exhibited no clear change and the high pulse number decreased (Tables 3 and S3), which led to a reduction in the connectivity between river channels and floodplains and a decline in biodiversity (Lu et al. 2018). In the section downstream of the Matsuo Dam, positive alterations occurred in the low and high pulse durations. An increase in low pulse duration has negative impacts on ecosystems as it decreases habitable area, the detachment opportunity for attached algae, and the connectivity with lentic habitats. On the other hand, an increase in the high pulse duration promotes more connections among lotic-lentic habitats and provides opportunities for plant seed establishment (Riis et al. 2008).

We found that the date of maximum flow only showed insignificant changes downstream of the weir, dams, and hydropower stations (Tables 3 and S3–4). This result indicates that the timing of the peak discharge is not affected by the dam operation regardless of its purpose.

The rise and fall rates decreased significantly downstream of the weir, dams, and hydropower stations (Tables 3 and S3–4). In the sections downstream of the dams and weir, the daily flow fluctuations and small peak discharges typically seen in the natural flow regime were suppressed, and the residual flows were characterized by prolonged small constant flow, except during extreme rainfall events, resulting in decreasing rise and fall rates. In the meshes downstream of the hydropower stations, the smaller rise and fall rates were presumably ascribed to the decreased number of small pulse discharges and constantly kept daily discharges due to the electricity supply and demand. Decreases in the rise and fall rates could have negative impacts on emergent vegetation because they can reduce or even eliminate the patch size and often facilitate the colonization of invasive species (Small et al. 2009). In addition, less frequent small floods may reduce the chances of algal detachment and regrowth, resulting in a dominant mature algal riverbed, which is generally not favored by algal feeders (e.g., Plecoglossus altivelis) and aquatic insects.

The percent changes of the reversals downstream of the dams were small but they were larger downstream of the hydropower stations (Tables 3 and S3–4). Because the daily discharge from hydropower plants varied depending on the electricity supply and demand, the fractional variations contributed to higher flow reversals, while the daily discharge of residual flow (i.e., downstream of the dams) was rigorously controlled to constant values. Although the outflow from the hydropower stations ameliorated the alterations (Figures 3 and 4(a) and Table S4), especially decreased monthly and minimum flow metrics, in the sections downstream of the dams, the alterations of some metrics such as high pulse metrics and rise/fall rates were observed (Table S4). This suggests that ameliorations of the natural flow regime by hydropower outflows in the catchment remained insufficient in light of potential environmental impacts.

The median monthly streamflow showed positive alterations downstream of the hydropower stations (Table S4). This may be attributed to the typical operation of hydropower generation that supplies stable daily electricity (i.e., daily discharge). In such a case, the median daily discharge in a certain period (here monthly) is probably larger than that under the natural condition (the no-dam scenario).

Flow alteration consequences of climate change

Under most climate change scenarios, the low flow metrics were projected to change slightly throughout the catchment, while the high flow metrics were projected to increase in the upstream and tributary meshes with smaller discharges (Figures 5, S6–7, and S11). The increased/decreased maximum/minimum flows typically observed under RCP8.5, in the far future, may play important roles in regulating patchy habitats of rivers, including floodplains, to accommodate some species of plants (Chen 2012). The maximum flows downstream of the Kawabaru hydropower station surrounded by relatively populated cities were projected to increase under RCP 8.5, while they were projected to decrease or remain unchanged under the other climate change scenarios. Therefore, if high radiative forcing is maintained in the future, further engineering works using climate change-based flood design should be considered. In addition, augmented maximum flows at shorter time-windows (e.g., 1-day maximum) observed over a wide range of upstream may trigger a greater riverbed disturbance in this region (Figures 5, S6, and S8). This will promote riverbed erosion (de Mello et al. 2015) as well as the passive migration of benthic organisms (Gibbins et al. 2007), resulting in widespread changes in the upstream environment.

Comparison of the impacts of dams and climate change

In the Omaru River and a major tributary (the Do River), the percent changes of most IHA were greater under the dam scenario than under the climate change scenario and a greater number of IHA were significantly altered (Figures 35 and S5–6 and Tables 3 and S3–4). This finding is consistent with a previous study that compared the impacts of dam and future climate change on hydrological regimes, specifically focusing on extreme flow events (e.g., maximum, and minimum flows) in the Rhone River catchment (Fatichi et al. 2015). Conversely, flow alteration in the Upper Yellow River can be mainly attributed to climate change than dam although the impact of dam would be underestimated (Cui et al. 2020). However, evaluation relying on the number of IHA may fall into a pitfall; partial IHA potentially behaves similarly as the others, resulting in an overestimation of the number of altered IHA. Indeed, according to the cross correlation among IHAs at up- and down-streams, some IHAs, especially monthly streamflow and low flow metrics, were highly correlated (Table S6). These results suggest that future studies are required to understand how such redundancy of metrics contributes to alteration levels over a certain geographical scale such as a catchment.

In a catchment-scale standpoint, widespread flow alterations were projected in the tributaries and uppermost main stem, where unaffected by the dams under the climate change scenario. Researchers have pointed out that even small alterations can trigger potential population losses for species dependent on the hydraulic conditions (e.g., rheophilic species), resulting in the loss of biodiversity (Schneider et al. 2013; Yang et al. 2017). Therefore, adequate environmental countermeasures for climate change should be implemented to safeguard biodiversity in tributaries because these marginal corridors account for most of the total streamflow length in river systems. Our results are consistent with a previous study that found dams further affected the streamflow regime such as minimum and maximum streamflow than climate change in the upper Rhone catchment. Since processes behind these results may not be comparable (e.g., different snowmelt due to climate change), further studies targeting catchments with different climates will provide insights into factors that govern the extent of alterations by dams and climate change.

Under the dam scenario, decreases in the low flow metrics were greater only in the sections downstream of the dams and weir (Tables 3 and S3–4). Whereas under the climate change scenario, decreases in low flow metrics were greater in the tributaries and upstream and were unaffected by the dams and weir (Figures 5 and S5–6). The reduced low flow, which typically involves a lower flow velocity, leads to poor water quality due to the deposition of pollutants (Smakhtin 2001). Although the anthropogenic pollutant loads were limited in the uppermost streams and tributaries, decreases in the low flow metrics can reduce the habitable area of riverine organisms through decreasing the water level and streamflow length. These negative ecological consequences of climate change can propagate downstream and impact catchment biodiversity.

Under all climate change scenarios, the median flow in January, April, August, November, and December showed a decreasing trend throughout the catchment (Figures 5 and S5). Therefore, for a sustainable power supply, dam managers should be required to operate dams considering the extent and timing of flow reduction expected under climate change in the future.

The 1-day maximum declined by 18.9 and 11.9% downstream of the Dogawa and Matsuo Dams, respectively, under the dam scenario, while it was projected to increase by 28.8 and 19.0% under RCP8.5, respectively (Tables 3 and S3 and Figure S8). Therefore, the flood control operations of the Dogawa and Matsuo Dams would need to be reconsidered, given that radiative forcing continues to increase in the near future.

In this study, a distributed hydrological model was applied to the Omaru River catchment located in the southwest Japan to evaluate the spatial extent of flow alteration throughout the catchment. The model predicted discharges with high accuracy along up- to down-streams for 10 years (NSE = 0.921–0.964, PBAIS = 3.9–4.0%). The flow alterations, represented by the indicator of hydrologic alterations (IHA) by the weir and dams were projected to be moderate to large; however, the alterations were ameliorated slightly by the discharge from hydropower plants. Such mitigation effects were also observed at the confluences of downstream tributaries. The flow alterations due to climate change were projected to be fractional in partial river sections under RCP2.6 and RCP4.5, while the decrease in low flow metrics was observed throughout the catchment under RCP8.5, suggesting a catchment-level shrinkage in important corridors of aquatic organisms through decreases in upstream length and water level. In the main stem Omaru River and the major tributary, the greater numbers of IHA were altered by dams than by climate change.

Future studies should test this approach in different catchments with contrasting climates, geologies, and water uses to highlight the differences in flow alterations between dams and climate change depending on background parameters. In addition, it is of great importance to examine the predictive abilities of our model for environmental and biological forecasting (e.g., environmental suitability or ecological niche models). The proposed framework will be useful for river managers when redesigning a management plan, for example, revising dam operations, retrofitting, or constructing new flood control structures, because the extent of flow alterations downstream, which have important implications for environmental protection, can be predicted. Furthermore, our study can provide reliable criteria for flood design considering the anticipated impacts of climate change scenarios.

We gratefully acknowledge Kyusyu Electric Power Co. and Miyazaki prefecture, which provided all the hourly outflow discharge data from the dams. This study also received financial support (Grants-in-Aid for Scientific Research) from the Japan Society for the Promotion of Science (19K15101).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Allen
R. G.
,
Pereira
L. S.
,
Raes
D.
&
Smith
M.
1998
Crop evapotranspiration – guidelines for computing crop water requirements – FAO irrigation and drainage paper 56
.
Irrigation and Drainage
.
https://doi.org/10.1016/j.eja.2010.12.001
.
Chen
H.
2012
Assessment of hydrological alterations from 1961 to 2000 in the Yarlung Zangbo River, Tibet
.
Ecohydrology and Hydrobiology
12
(
2
).
https://doi.org/10.2478/v10104-012-0009-z
.
Cui
T.
,
Tian
F.
,
Yang
T.
,
Wen
J.
&
Khan
M. Y. A.
2020
Development of a comprehensive framework for assessing the impacts of climate change and dam construction on flow regimes
.
Journal of Hydrology
590
.
https://doi.org/10.1016/j.jhydrol.2020.125358
.
Deitch
M. J.
,
Merenlender
A. M.
&
Feirer
S.
2013
Cumulative effects of small reservoirs on streamflow in northern coastal California catchments
.
Water Resources Management
27
(
15
).
https://doi.org/10.1007/s11269-013-0455-4
.
de Mello
C. R.
,
Ávila
L. F.
,
Viola
M. R.
,
Curi
N.
&
Norton
L. D.
2015
Assessing the climate change impacts on the rainfall erosivity throughout the twenty-first century in the Grande River Basin (GRB) headwaters, Southeastern Brazil
.
Environmental Earth Sciences
73
(
12
).
https://doi.org/10.1007/s12665-015-4033-3
.
Döll
P.
&
Zhang
J.
2010
Impact of climate change on freshwater ecosystems: A global-scale analysis of ecologically relevant river flow alterations
.
Hydrology and Earth System Sciences
14
(
5
).
https://doi.org/10.5194/hess-14-783-2010
.
Ehsani
N.
,
Vörösmarty
C. J.
,
Fekete
B. M.
&
Stakhiv
E. Z.
2017
Reservoir operations under climate change: Storage capacity options to mitigate risk
.
Journal of Hydrology
555
.
https://doi.org/10.1016/j.jhydrol.2017.09.008
.
Fatichi
S.
,
Rimkus
S.
,
Burlando
P.
,
Bordoy
R.
&
Molnar
P.
2015
High-resolution distributed analysis of climate and anthropogenic changes on the hydrology of an Alpine catchment
.
Journal of Hydrology
525
.
https://doi.org/10.1016/j.jhydrol.2015.03.036
.
Faye
C.
2018
Characterization of the effects of Manantali Dam on the hydrological regime of the Senegal river by the IHA/RVA method
.
Open Access Journal of Science
2
(
6
).
https://doi.org/10.15406/oajs.2018.02.00116
.
Gibbins
C.
,
Vericat
D.
&
Batalla
R. J.
2007
When is stream invertebrate drift catastrophic? The role of hydraulics and sediment transport in initiating drift during flood events
.
Freshwater Biology
52
(
12
).
https://doi.org/10.1111/j.1365-2427.2007.01858.x
.
Goldstein
J. C.
&
Tarhule
A.
2015
Evaluating the impacts of climate change and switchgrass production on a semiarid basin
.
Hydrological Processes
29
(
5
).
https://doi.org/10.1002/hyp.10159
.
Haddeland
I.
,
Heinke
J.
,
Biemans
H.
,
Eisner
S.
,
Flörke
M.
,
Hanasaki
N.
,
Konzmann
M.
,
Ludwig
F.
,
Masaki
Y.
,
Schewe
J.
,
Stacke
T.
,
Tessler
Z.
,
Wada
Y.
&
Wisser
D.
2014
Global water resources affected by human interventions and climate change
.
Proceedings of the National Academy of Sciences of the United States of America
111
(
9
).
https://doi.org/10.1073/pnas.1222475110
.
Jardim
P. F.
,
Melo
M. M. M.
,
de Ribeiro
L. C.
,
Collischonn
W.
&
da Paz
A. R.
2020
A modeling assessment of large-scale hydrologic alteration in South American Pantanal due to upstream dam operation
.
Frontiers in Environmental Science
8
.
https://doi.org/10.3389/fenvs.2020.567450
.
Kazama
S.
,
Matsumoto
S.
,
Ranjan
S. P.
,
Hamamoto
H.
&
Sawamoto
M.
2007
Characterization of firefly habitat using a geographical information system with hydrological simulation
.
Ecological Modelling
209
(
2–4
).
https://doi.org/10.1016/j.ecolmodel.2007.06.029
.
Kazama
S.
,
Sakamoto
K.
,
Salem
G. S. A.
&
Kashiwa
S.
2021
Improving the accuracy of snow and hydrological models using assimilation by snow depth
.
Journal of Hydrologic Engineering
26
(
1
).
https://doi.org/10.1061/(asce)he.1943-5584.0002019
.
Kibler
K. M.
&
Tullos
D. D.
2013
Cumulative biophysical impact of small and large hydropower development in Nu River, China
.
Water Resources Research
49
(
6
).
https://doi.org/10.1002/wrcr.20243
.
Kimura
T.
1961
The Food Runoff Analysis Method by the Storage Function Model
.
Laizé
C. L. R.
,
Acreman
M. C.
,
Schneider
C.
,
Dunbar
M. J.
,
Houghton-Carr
H. A.
,
Flörke
M.
&
Hannah
D. M.
2014
Projected flow alteration and ecological risk for pan-European rivers
.
River Research and Applications
30
(
3
).
https://doi.org/10.1002/rra.2645
.
Larned
S. T.
,
Schmidt
J.
,
Datry
T.
,
Konrad
C. P.
,
Dumas
J. K.
&
Diettrich
J. C.
2011
Longitudinal river ecohydrology: Flow variation down the lengths of alluvial rivers
.
Ecohydrology
4
(
4
).
https://doi.org/10.1002/eco.126
.
Li
D.
,
Long
D.
,
Zhao
J.
,
Lu
H.
&
Hong
Y.
2017
Observed changes in flow regimes in the Mekong River basin
.
Journal of Hydrology
551
.
https://doi.org/10.1016/j.jhydrol.2017.05.061
.
Liggett
J. A.
1975
Basic equations of unsteady flow
. In:
Unsteady Flow in Open Channels
(
Mahmood
K.
&
Yevjevich
V.
, eds).
Water Resources Publications
,
Littleton, CO, USA
, pp.
29
62
.
Lighthill
M. J.
&
Whitham
G. B.
1995a
On kinematic waves I. Flood movement in long rivers
.
Proceedings of the Royal Society of London, Series A, Mathematical and Physical Sciences
229
(
1178
),
281
316
.
https://doi.org/10.1098/rspa.1955.0088
.
Lighthill
M. J.
&
Whitham
G. B.
1995b
On kinematic waves II. A theory of traffic flow on long crowded roads
.
Proceedings of the Royal Society of London, Series A, Mathematical and Physical Sciences
229
(
1178
).
https://doi.org/10.1098/rspa.1955.0089
.
Lu
W.
,
Lei
H.
,
Yang
D.
,
Tang
L.
&
Miao
Q.
2018
Quantifying the impacts of small dam construction on hydrological alterations in the Jiulong River basin of Southeast China
.
Journal of Hydrology
567
.
https://doi.org/10.1016/j.jhydrol.2018.10.034
.
Lytle
D. A.
&
Poff
N. L. R.
2004
Adaptation to natural flow regimes
.
Trends in Ecology and Evolution
.
https://doi.org/10.1016/j.tree.2003.10.002
.
Mahmoodi
N.
,
Kiesel
J.
,
Wagner
P. D.
&
Fohrer
N.
2021
Spatially distributed impacts of climate change and groundwater demand on the water resources in a Wadi system
.
Hydrology and Earth System Sciences
25
(
9
).
https://doi.org/10.5194/hess-25-5065-2021
.
Martinec
J.
1960
The degree-day factor for snowmelt-run-off forecasting
.
IAHS Publications
51
,
468
477
.
Matsubara
T.
,
Tsuchida
K.
&
Hibiya
M.
2015
Study on parameter identification of distributed runoff model using SCE-UA method
.
Journal of Japan Society of Civil Engineers, Ser. B1 (Hydraulic Engineering)
71
(
4
).
https://doi.org/10.2208/jscejhe.71.i_265
.
Mineda
H.
,
Nukazawa
K.
,
Nakao
S.
&
Suzuki
Y.
2020
Longitudinal evaluation of altered flow regimes in the Omaru river network using a distributed hydrological model
.
Journal of Japan Society of Civil Engineers, Ser. G (Environmental Research)
76
(
5
).
https://doi.org/10.2208/jscejer.76.5_i_65
.
Moraga
J. S.
,
Peleg
N.
,
Fatichi
S.
,
Molnar
P.
&
Burlando
P.
2021
Revealing the impacts of climate change on mountainous catchments through high-resolution modelling
.
Journal of Hydrology
603
.
https://doi.org/10.1016/j.jhydrol.2021.126806
.
Moriasi
D. N.
,
Arnold
J. G.
,
Van Liew
M. W.
,
Bingner
R. L.
,
Harmel
R. D.
&
Veith
T. L.
2007
Model evaluation guidelines for systematic quantification of accuracy in watershed simulations
.
Transactions of the ASABE
50
(
3
),
885
900
.
Munn
M. D.
&
Brusven
M. A.
1991
Benthic macroinvertebrate communities in nonregulated and regulated waters of the clearwater river, Idaho, U.S.A
.
Regulated Rivers: Research & Management
6
(
1
).
https://doi.org/10.1002/rrr.3450060102
.
Nash
J. E.
&
Sutcliffe
J. V.
1970
River flow forecasting through conceptual models part I – a discussion of principles
.
Journal of Hydrology
10
(
3
).
https://doi.org/10.1016/0022-1694(70)90255-6
.
Nilsson
C.
,
Reidy
C. A.
,
Dynesius
M.
&
Revenga
C.
2005
Fragmentation and flow regulation of the world's large river systems
.
Science
.
https://doi.org/10.1126/science.1107887
.
Nislow
K. H.
,
Magilligan
F. J.
,
Fassnacht
H.
,
Bechtel
D.
&
Ruesink
A.
2002
Effects of dam impoundment on the flood regime of natural floodplain communities in the Upper Connecticut River
.
Journal of the American Water Resources Association
38
(
6
).
https://doi.org/10.1111/j.1752-1688.2002.tb04363.x
.
Nukazawa
K.
,
Shiraiwa
J. I.
&
Kazama
S.
2011
Evaluations of seasonal habitat variations of freshwater fishes, fireflies, and frogs using a habitat suitability index model that includes river water temperature
.
Ecological Modelling
222
(
20–22
).
https://doi.org/10.1016/j.ecolmodel.2011.09.005
.
Nukazawa
K.
,
Kazama
S.
&
Watanabe
K.
2015
A hydrothermal simulation approach to modelling spatial patterns of adaptive genetic variation in four stream insects
.
Journal of Biogeography
42
(
1
).
https://doi.org/10.1111/jbi.12392
.
Nukazawa
K.
,
Shirasaka
K.
,
Kajiwara
S.
,
Saito
T.
,
Irie
M.
&
Suzuki
Y.
2020
Gradients of flow regulation shape community structures of stream fishes and insects within a catchment subject to typhoon events
.
Science of the Total Environment
748
.
https://doi.org/10.1016/j.scitotenv.2020.141398
.
Poff
N. L.
,
Allan
J. D.
,
Bain
M. B.
,
Karr
J. R.
,
Prestegaard
K. L.
,
Richter
B. D.
,
Sparks
R.
&
Stromberg
J.
1997
The natural flow regime
.
BioScience
47
(
11
).
https://doi.org/10.2307/1313099
.
Richter
B. D.
,
Baumgartner
J. V.
,
Powell
J.
&
Braun
D. P.
1996
A method for assessing hydrologic alteration within ecosystems
.
Conservation Biology
10
(
4
).
https://doi.org/10.1046/j.1523-1739.1996.10041163.x
.
Richter
B. D.
,
Baumgartner
J. V.
,
Braun
D. P.
&
Powell
J.
1998
A spatial assessment of hydrologic alteration within a river network
.
River Research and Applications
14
(
4
).
https://doi.org/10.1002/(sici)1099-1646(199807/08)14:4<329::aid-rrr505>3.0.co;2-e
.
Richter
B. D.
,
Davis
M. M.
,
Apse
C.
&
Konrad
C.
2012
A presumptive standard for environmental flow protection
.
River Research and Applications
28
(
8
).
https://doi.org/10.1002/rra.1511
.
Riis
T.
,
Suren
A. M.
,
Clausen
B.
&
Sand-Jensen
K.
2008
Vegetation and flow regime in lowland streams
.
Freshwater Biology
53
(
8
).
https://doi.org/10.1111/j.1365-2427.2008.01987.x
.
Ryo
M.
,
Iwasaki
Y.
,
Yoshimura
C.
,
Saavedra
V.
&
C
O.
2015
Evaluation of spatial pattern of altered flow regimes on a river network using a distributed hydrological model
.
PLoS ONE
10
(
7
).
https://doi.org/10.1371/journal.pone.0133833
.
Sato
Y.
,
Kojiri
T.
,
Michihiro
Y.
,
Suzuki
Y.
&
Nakakita
E.
2012
Estimates of climate change impact on river discharge in Japan based on a super-high-resolution climate model
.
Terrestrial, Atmospheric and Oceanic Sciences
23
(
5
).
https://doi.org/10.3319/TAO.2012.05.03.02(WMH)
.
Schneider
C.
,
Laizé
C. L. R.
,
Acreman
M. C.
&
Flörke
M.
2013
How will climate change modify river flow regimes in Europe?
Hydrology and Earth System Sciences
17
(
1
).
https://doi.org/10.5194/hess-17-325-2013
.
Smakhtin
V. U.
2001
Low flow hydrology: A review
.
Journal of Hydrology
.
https://doi.org/10.1016/S0022-1694(00)00340-1
.
Small
M. F.
,
Bonner
T. H.
&
Baccus
J. T.
2009
Hydrologic alteration of the lower Rio Grande terminus: A quantitive assessment
.
River Research and Applications
25
(
3
).
https://doi.org/10.1002/rra.1151
.
Sulis
M.
,
Paniconi
C.
,
Rivard
C.
,
Harvey
R.
&
Chaumont
D.
2011
Assessment of climate change impacts at the catchment scale with a detailed hydrological model of surface-subsurface interactions and comparison with a land surface model
.
Water Resources Research
47
(
1
).
https://doi.org/10.1029/2010WR009167
.
Tsegaw
A. T.
,
Pontoppidan
M.
,
Kristvik
E.
,
Alfredsen
K.
&
Muthanna
T. M.
2020
Hydrological impacts of climate change on small ungauged catchments – results from a global climate model-regional climate model-hydrologic model chain
.
Natural Hazards and Earth System Sciences
20
(
8
).
https://doi.org/10.5194/nhess-20-2133-2020
.
Volpi
E.
,
Di Lazzaro
M.
,
Bertola
M.
,
Viglione
A.
&
Fiori
A.
2018
Reservoir effects on flood peak discharge at the catchment scale
.
Water Resources Research
54
(
11
).
https://doi.org/10.1029/2018WR023866
.
Yang
T.
,
Cui
T.
,
Xu
C. Y.
,
Ciais
P.
&
Shi
P.
2017
Development of a new IHA method for impact assessment of climate change on flow regime
.
Global and Planetary Change
156
.
https://doi.org/10.1016/j.gloplacha.2017.07.006
.
Zuo
Q.
&
Liang
S.
2015
Effects of dams on river flow regime based on IHA/RVA
.
IAHS-AISH Proceedings and Reports
368
.
https://doi.org/10.5194/piahs-368-275-2015
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).

Supplementary data