Assessment of Impacts of River Flow Changes on Reservoir Operation for Water Supply
Impacts of river flow changes on reservoir operation for water supply are first assessed considering changes in reservoir storage volume. Figure 6.7 shows the trajectories of the storage volume of the Sameura Reservoir in the Yoshino River basin. It can be seen in this figure that the storage volume at each time step averaged for 25 years decreased from May to December under the future climate conditions compared to that under the present climate conditions. This could be because w'ater flow in the Yoshino River decreased in this period under the future climate conditions as shown in Figure 6.6a. The storage volume of the Sameura Reservoir decreased to zero (empty) in 9 years out of the 25 years of the simulation period under the future climate conditions, although this was not observed in the simulation using the results of the current climate experiment. While the number of years where such severe droughts occur may decrease if the reproducibility of the calculated river flow improves to mitigate the underestimation of river discharge in those seasons, operation of the Sameura Reservoir for water use would become more challenging with the decreased river flow' regime under the future climate conditions. The storage volume in the Sameura Reservoir did not recover to the initial storage volume, which is fixed for January 1st of each year in this case study,
Figure 6.7 Trajectories of water storage volume of the Sameura Reservoir in the Yoshino River basin (a) under the present climate conditions and (b) under the future climate conditions.
Figure 6.8 Trajectories of water storage volume of the Shirakawa Reservoir in the Mogami River basin (a) under the present climate conditions and (b) under the future climate conditions.
at the end of the year in 13 years out of the 25 years of the future climate experiment. This means that a low storage condition is carried over into the next year and that the risk of inter-annual drought is greater under the future climate conditions.
Figure 6.8 shows the trajectories of storage volume of the Shirakawa Reservoir in the Mogami River basin. It can be seen in this figure that water storage in the Shirakawa Reservoir maintains greater volumes in winter from January to March under the future climate conditions compared to that under the present climate conditions. This tendency continued also in April after the drawdown period of the Shirakawa Reservoir for floods due to snowmelt in this season. However, the mean storage volume is smaller under the future climate conditions than that under the current climate conditions. This could be because the amount of river discharge decreased in this period due to the shift in the timing of the snowmelt peak as shown in Figure 6.6.
However, the water use capacity of the Shirakawa Reservoir is decreased in June in order to enlarge the empty volume for flood management (flood control capacity) in the flood season (in summer). Due to this operational constraint of this reservoir, the shift in the peak of river discharge due to snowmelt did not affect the water use operation of this reservoir very much after June where the water storage capacity starts to be regulated. The decrease in river flow in the summer season may rather impact water use operation of the reservoir and cause late recovery of water storage after the flood season ends in October. This implies that reallocation of reservoir storage capacity between flood control and water use purposes would be needed for the Shirakawa Reservoir to mitigate the increase in drought risk in summer in the future.
Assessment of Adaptation Options
Assessment of adaptation options to river flow change under the future climate conditions is conducted for water use operation of the Sameura Reservoir in the Yoshino River basin in this case study. In order to evaluate the effectiveness of adaptation options, an evaluation index is first defined for the performance of reservoir operation for water supply. The drought damage function (DDF), which was proposed by
Ikebuchi et al. (1990), was employed as the evaluation index in this case study. The DDF can be defined using the following equation:
where со* is the drought damage at time step k, and and are, respectively, water demand and river discharge at a reference point at time step k. The annual drought damage can be calculated using the following equation:
where Fj is the total (annual) drought damage over year j, coj^ is the drought damage at time step к in year j, and К is the number of time steps in a year, respectively.
Evaluation of Impacts of Climate Change on Reservoir Operation for Water Supply
The simulation of water use operation of reservoirs in the Yoshino River basin was conducted for the present and future climate conditions in a similar way to that described above. The initial storage volume of reservoirs was, however, set to be identical to that at the end of the previous year to take inter-annual droughts into account when calculating drought damage. The distribution of annual drought damages for the 25 years of each climate experiment period estimated by using Equation 6.20 is shown as box plots in Figure 6.9. The annual drought damages observed in the simulation of reservoir operation with the observed hydrological data from 1979 to 2003 are also shown in Figure 6.9 for comparison. Values out of the 1.5 interquartile range were plotted as outliers in this figure. It can be seen in this figure that the annual drought damages were smaller in the simulation for the present climate experiment than that with the observed hydrological data in terms of mean and median values.
Figure 6.9 Distribution of annual drought damages estimated for the operation of reservoirs for water supply in the Yoshino River basin for the present and future climate experiments and for the observed flow regimes from 1979 to 2003.
The greatest value of annual drought damage observed in the simulation with the observed hydrological data was not seen in that using data from the present climate experiment. It is therefore considered that extreme drought conditions are not perfectly simulated for the present climate experiment data used in this case study. This can also be considered to contribute to the underestimation in estimating the mean and median values of annual drought damage in the simulation for the present climate conditions. Further efforts are needed for improving the reproducibility of extreme drought conditions for impact assessment of climate change.
Comparing the results of the simulations for the present and future climate conditions in Figure 6.9, the annual drought damage distribution shifted to greater values for the future climate than that for the present climate. The median and third quartile values of annual drought damage in the future climate simulation, respectively, became close to the third quartile and maximum values in the present climate simulation. It can therefore be considered that the occurrence of large-scale drought events could increase in the future climate.
Identifying Potential Suitable Adaptation Options
Adaptation options to changes in river discharge were then analyzed considering the expected impacts on reservoir operation for water supply identified in the previous processes. Although various adaptation options can be considered including integrated river basin management approaches, adaptation options for the Sameura Reservoir were considered in this case study. First, statistic reverse storage mass- curves (SRSMCs) (Kyoshi and Shimoda, 1995) were drawn to identify potential effective adaptation options. The SRSMC is an extension of mass curve approaches (Klemes, 1979) to include order statistics on drought severity and persistence that are considered in drought duration curves (Takeuchi, 1988).
The estimation results of SRSMCs are shown in Figure 6.10. In this figure, the 77th reverse storage mass-curve from the top represents the required storage volumes when the /7th smallest river discharge is given at each time step. The required storage capacity can be estimated by taking a SRSMC with the same return period (or non-exceedance probability) with the designed level for water supply for the target reservoir. As the designed level of the current management by the Sameura Reservoir for water supply is for droughts with a ten-year return period, the averaged values of the second and third greatest curves were considered as the required storage capacity for water use at the designed level (denoted as RSCD hereafter) for this reservoir. From Figure 6.10, it can be seen that the RSCD of the Sameura Reservoir did not exceed the current storage capacity for water use, which is 173 million cubic meters (MCM), in the present climate simulation.
On the other hand, it largely exceeded the current storage capacity for water use as well as the capacity for water use and power generation (199-209 MCM) in the future climate simulation. The RSCD at the first time step, in January, was around 350 MCM, which is greater than the total storage capacity of this reservoir (289 MCM), in the future climate simulation. This result implies that severe low flow conditions, which cannot be managed by the reservoir even though all the storage capacity is allocated for water supply, could occur once in three years under the future climate.
Figure 6.10 Statistic reverse storage mass-curves and the required storage capacity at the designed level (for a low flow regime of a ten-year return period) (a) for the present climate simulation and (b) for the future climate simulation.
Although this tendency could be mitigated by integrated operation or by enlarging the storage capacity for water supply by storage capacity reallocation, countermeasures on the demand side, such as water demand control across the river basin, would be needed to manage these drought situations under the future climate. Sedimentation can lead to worsening of this situation by reducing the storage capacity, although it is currently not severe in the Sameura Reservoir. It is, however, needed to be carefully considered for better understanding of climate change impacts, because an increase in heavy rainfall may cause more sediment yield from the upper basin and more sedimentation in the reservoir.
Considering the results described above, three adaptation options are considered as a fundamental analysis in this case study, which are as follows: (i) enabling the use of the power generation capacity (260-360 MCM) for water supply purpose additionally (denoted as Option A), (ii) enabling the use of the flood control capacity (800-900 MCM) for water supply purpose additionally assuming that prior release operation can be conducted effectively by considering accurate real-time hydrological predictions in the future (denoted as Option B), and (iii) reducing water demand at Ikeda by 15% from the current values by advancing water saving technologies or by restructuring water rights (denoted as Option C). The results of drought damage distributions performed by the simulation of water use operation of the Sameura Reservoir considering each adaptation option are shown in Figure 6.11. It can be seen here that the distribution of annual drought damage was not improved to the level in the present climate simulation even though any adaptation option among Options A, B, and C was applied. This could be because those adaptation options could not mitigate the impacts of river discharge decrease in some years of the future climate experiment as it was extremely severe in those years. Among the three options, Option C showed the greatest mitigation of drought damages including those for the most severe drought year (outlier), while Options A and В did not improve drought damage very much for the most severe drought year. Adaptation options that control water demand in the river basin are therefore considered to be somewhat effective in the river basin where river discharge decreases significantly once in several years under the future climate like the Yoshino River basin.
Figure 6.11 Improvements in the distribution of annual drought damages estimated for the operation of reservoirs for water supply in the Yoshino River basin with each adaptation option.