Various kinds of GCM output are provided in the SOUSEI program. Comparing from 20km model, 60km GCM has many number of ensemble members. Although 60km is relatively high resolution among GCMs, complex topography in Japan region is not well resolved. As the evaluation of snow water amount is crucial for assessing the impact of climate change on water resources, effect of the resolution of GCM output on the snow water equivalent (SWE) estimation is investigated by the land surface model’s output driven by original (20km) resolution and up-scaled (60km) resolution. Annual maximum SWE tends to be underestimated by up-scaled information, and this tendency is different with topographic condition and geographical location. Annual maximum SWE ratio by 60km-scale analysis and 20km-scale analysis is attempted to be explained by winter season’s precipitation, standard deviation of sub-grid scale elevation, and precipitation weighted average winter season’s temperature from multiple regression analysis with determination coefficient of 0.92 for middle (300-600m) elevation zone.

Despite a decade of research into climate change impacts on water resources, the scientific community has delivered relatively few practical methodological developments for integrating uncertainty into water resources system design. This study involves an application of the “decision scaling” methodology for assessing climate change impacts on water resources system performance and asks how such an approach might inform planning decisions. The decision scaling method reverses the conventional ethos of climate impact assessment by first establishing the climate conditions that would compel planners to intervene. Climate model projections are introduced at the end of the process to characterize climate risk in such a way that avoids the process of propagating those projections through hydrological models. Here 1000 multisite synthetic monthly streamflow traces were propagated through a model of the Melbourne bulk supply system to test the sensitivity of system performance to variations in streamflow characteristics. An empirical relation was derived to convert decision-critical flow statistics to climatic units, against which 138 alternative climate projections were plotted and compared. The decision threshold was defined using system yield metric constrained by multiple performance criteria. The approach allows for fast and simple incorporation of demand forecast uncertainty and demonstrates the reach of the decision scaling method through successful execution in a large and complex water resources system. The analysis provides insights to planners in understanding the level of risk exposure to climate change.

General Circulation Model (GCM) precipitation scenarios are often characterized by biases and coarse resolution that limit their direct application for basin level hydrological modeling. GCM outputs are usually inadequate to capture the spatial variability at regional or local scales necessary for hydrological applications. The correction of precipitation and temperature leads to satisfactory result in hydrological impact studies. Thus bias correction method applied to the output data of GCM is essential for the climate impact studies. In this study, we improved the quantile mapping bias correction by using a loop optimization approach. The GCM data under The Coupled Model Intercomparison Project Phase 5 (CMIP5) was employed to compare the quality of GCMs precipitation and temperature which includes BCC-CSM1.1, CanESM2, CNRM-CM5, CSIRO-Mk3.6, EC-Earth, GFDL-CM3, MIROC5 and NorESM1-M with different representative concentration pathway scenario (RCP). The performance of this method was evaluated by comparing observed and simulated GCMs output climate series in the term of spatial and temporal. Furthermore, the bias corrected parameters also could be used to correct the future climate GCM data for assessing the impact of climate change on hydrology. The bias corrected results were applied to simulate the runoff of Chi River Basin Thailand by using SWAT rainfall-runoff model. In the term of monthly mean and distribution of climate pattern, the improving bias correction method gave the result of incorporating the spatial heterogeneity in precipitation and temperature and thus they can obvious reduce the biases from raw GCM. In conclusion, our analysis results showed that spatial distributions of time-based statistics are significantly and consistently improved by intensity-based statistical bias correction methods. The results of this study revealed the performance of GCM data, thus it provided useful information in selecting of GCM data in hydrological impact studies in the basin scale.   

The climate of Myanmar is highly dominated by India monsoon and there are high precipitations from May to October (southwest monsoon) , in contrast, almost no rain from December to March (northwest monsoon). Although Myanmar has abundant water resources, there are major challenges for effective water resources management, such as uneven spatial and temporal distribution of those resources.

Within Myanmar’s four major rivers, river discharge data is available only in Sittaung River. A hydrologic circulation model is conducted in this study by using BASIN-HSPF model, developed by EPA, United States and satellite estimated precipitation data of TRMM/GSMaP, ESA land use data and the collected dams and reservoir operation information. The discharges of Sittaung River were reproduced for four years (2010-2013) and the simulated outputs were validated with the observed discharge to identify the hydrologic model parameters which will be used for the future hydrologic projections of Sittaung River.

In CMIP5, the historical simulation precipitation data and the future projected precipitation data are available from various models. The likely reliable models were analyzed and validated by GSMaP precipitation data (hourly, 0.1 degree space resolution). Then, it is found that the output of MIROC4h model is acceptable for rainfall distribution rather than of the other models. Precipitation of MIROC4h is employed to estimate the future hydrologic projection in Sittaung river basin for RCP scenarios.

The historical precipitation data of MIROC4h model were validated by using TRMM/GSMaP and then calculated the percentage errors for 10 years (2004-2013). The factor numbers (f_n) are calculated from those percentage error to get the closet value of MIROC4h data compared to TRMM/GSMaP data.

The experiments for future available water resources of Sittaung watershed were conducted by using local scenario of land use and land cover changes under RCP4.5 scenario.

There is an energy exchange between atmosphere and land through water cycle. Koster et al. (2004) indicated the influence of the soil moisture anomaly on precipitation over the semi-arid regions such as Central U. S., Pakistan-India and Eastern China by using 12 atmospheric general circulation models (AGCMs). These regions are the area of high irrigation intensity. Human activities in the semi-arid region such as withdrawing water from rivers and groundwater in order to grow  crops make the soil moisture tend to be wet which may increase the amount of precipitation. Therefore, human activities influence the amount of precipitation over the semi-arid regions. However, the current AGCMs do not consider the influence of the human activities. Thus, the objective of this study is to integrate the irrigation scheme proposed by Pokhrel et al. (2012) and the dynamic groundwater representation proposed by Koirala et al. (2014) into an AGCM. Then, we attempt to quantify the influence of the human activities on water cycle by comparing the results between the integrated model and the current model. Because of the lack of observation data such as soil moisture, river discharge and amount of groundwater, we used the results obtained by Pokhrel et al. (2012) which showed realistic results compared with observation data.By considering the effects of the human activities, the simulated annual basin water budget of the Mississippi river and the Chang-Jiang river is close to that of obtained by Pokhrel et al. (2012). And, the annual variation of 2m temperature and precipitation over those regions decreased which is caused by the decreasing of the annual variation of soil moisture due to human activities.

Reservoirs are important infrastructure in that they occupy a major fraction of water resources in our society. In Japan, more than 50% of domestic water is covered by reservoirs. However, because of their characteristics of closed water body, influences of water quality deterioration tend to accumulate. Therefore, improvement of water quality is generally takes a long time once a reservoir is polluted. In order to project potential influences from climate change on water quality in those water resources reservoirs, we tried to make assessments by the use of a vertical 1-dimensional numerical model. 

Thirty three reservoirs were chosen among multi-purpose reservoirs from all over Japan to make projections of the future conditions of stored water. Stratification intensity of surface layer of reservoirs were used as an index of algal (especially cyanobacteria) growth potential. Future meteorological conditions such as solar radiation and atmospheric temperature were given using the output of several GCMs. Our results predicted that eutrophication potential would be higher in the north part of Japan. 

Climate change may have a serious impact on water resources as well on hydropower generation in the future. The purpose of this study is to analyze the influence of climate change on Taiwan’s potential hydroelectric generating capacity based on the river discharge derived from precipitations of scenarios of RCP2.6, RCP4.5, RCP6.0 and RCP8.5 by rainfall-Runoff Coefficient.

In this study, under AR5 scenarios, we adopted twelve GCMs’: BCC-CSM1.1, CCSM4, CESM1-CAM5.1, CSIRO-Mk3.6.0, GFDL-CM3, GISS-E2-H, HadGEM2-AO, HadGEM2-ES, IPSL-CM5A-MR, MIROC5, MRI-CGCM3, and NorESM1-M. The outputs are applied to derive the trends of rainfalls and river discharges of future climate scenarios. Most GCMs show that potential hydropower in the wet season increases significantly, and decreases in the dry season.

The variation of potential hydropower derived from various GCMs in dry season is larger than those in wet season. The results showed that 66% GCMs indicate potential hydropower generation will increase in short-term future and the variation of potential hydropower generation is –30% ~ +40%; 75% will increase in mid-term future and the variation is –30%~+45%; 83% will increase in long-term future and the variation is –40% ~ +55%. The results of this study may offer for energy-related decision making in the future.

Water is an increasingly limited and highly essential resource for many countries where agriculture is the main income of the economy corresponding with ensures the well-being of the people. The proper planning of water resource availability based on uncertainty climate change impact is very necessary. The five large scale dam reservoirs located in the northern, central and western parts of Thailand which consist of the Ping River basin having Bhumibol Dam, the Nan River basin having the Sirikit Dam, the Pasak River basin having the Pasak Jolasid Dam and the Mae Klong River basin having the Srinagarind and Vajiralongkorn dams, were selected to study as dominant source for downstream water utilization. The proposed reservoirs were also chosen to evaluate water scarcity and potential for harmful effects of climate change. The observed temperature, precipitation data and inflow information were prepared from twenty-five meteorological stations and observed hydrological station of each dam site. The objective of this study is to assess future water resources availability in term of inflow through large scale reservoirs in Thailand by using General Circulation Model (MRI-AGCM3.2S) which is provided by Meteorological Research Institute of Japan (MRI). MRI-AGCM3.2S was chosen to investigate the effect of climate change on temperature, precipitation and evapotranspiration upstream of each dam. The MRI-AGCM3.2S simulation results were compared with observed data from meteorological stations from 1979-2003 and the projection of climate data was selected for 2075-2099 with worst scenario cases (RCP8.5).Lastly, the 1K-DHM hydrological model was used for daily dam inflow prediction. Results of this finding estimated the dam inflow based on lately MRI-AGCM scenario that can be used for optimized water resources management in the downstream areas and verified the effective reservoir operation for appropriate adaptability in the future.

  A daily mean gridded temperature dataset of monsoon Asia (15S-55N, 60E-150E) for the period of 1961-2007 is created and open to public (Yasutomi et al., 2011) as a part of APHRODITE project(Yatagai et al., 2012).
 The number of stations used the interpolation analysis is up to 2 times the number of stations based on the Global Telecommunication System (GTS), which have been used to obtain other gridded temperature products. We obtained daily surface observation in collabollation with the local agency.

Comparison between monthly mean temperature datasets (CRU_TS3.0 and Univ. of Delaware) and APRHODITE daily mean temperature dataset (AphroTemp V1204R1) is made to estimate the effect of the increase of surface observation input.
Significant difference is not shown over coastal and plain region over Monsoon Asia. However, differences of 2-5 degC are shown in mountainous region of Tibetan Plateau and Central Asia.

Another product (AphroTemp_V1204R1g), using on-line available surface observaion data and adapting same interpolation algorithm, is derived to estimate the differences attributed to the increase of input data.
Positive anomaly is shown in Himalayas and Tibet and negative anomaly is shown along the highlands such as Pamir, Tianshan and Altai. Similar differences are found in comparison with other monthly datasets (CRU_TS3.0 and Univ. of Delaware). Those differences is more dominant in
winter than in summer.

Increase of inputs in high altitudes contributes more on improvements in climatological mean temperature than plain or oceanic area. Estimation of the mean temperature of mountanous regions in South and Central Asia is important because the decrease of glacier is one of the most urgent issues on global environmental changes.

Recently, localized heavy rainfalls due to frontal rain-bands (BAIU front) are increasing in summer of Japan, which can cause severe flood disaster in small river channels. It is important to understand the behavior of these rainfalls not only under the present climate but also the future climate. However, due to its small scale, localized heavy rainfalls are difficult to simulate precisely using low-resolution climate models, and it is more difficult to project the future behavior under the changing climate.

  So we have been focusing on atmospheric characteristics having larger spatial scale from GCM outputs and attempting to distinguish a specific atmospheric condition that can trigger localized heavy rainfalls. In our previous study, we distinguished the specific condition with visual judgment, which took us much time and could not be free from our subjective viewpoints. In this study, we have utilized Self-organizing map (SOM) to classify atmospheric conditions automatically and to distinguish the specific condition objectively. SOM is a technique to classify multidimensional data and to rearrange the data on two-dimensional surface according to clustered features, which also can cope with huge number of ensemble outputs effectively.

  We have analyzed frequency of the specific atmospheric condition under the future climate as well as the present climate, and evaluated the statistical significance of frequency changes under the future climate. In detail, present (1979~2003) and future (2075~2099) climate conditions were analyzed based on ensemble experiments output with 60km resolution GCM (MRI-AGCM3.2H). By counting the number of times for the specific atmospheric condition, it is able to estimate possible changes of localized heavy rainfall pattern in the future. Our approach shows one of example that even small scale of localized heavy rainfall can be captured by considering characterized atmospheric conditions in large scale.)H)H)

In the Aral Sea Basin, serious water scarcity was induced and the Aral Sea has drastically shrunk due to huge-scale irrigation project in the 20^th century. For basic information of sustainable water management, water resource and demand have to be clarified considering climate change. In fact, several tendencies of climate change were recently reported using local meteorological data and satellite analysis. And in the future, many GCMs projected that both precipitation and temperature will increase especially in mountainous region. As for basin water balance, water resource projection requires comparison between increases of precipitation and evapotranspiration, while water demand is expected to increase due to high evapotranspiration rate on farm.

The authors have attempted to develop terrestrial water circulation model in the Central Asia to estimate basin water balance. It mainly consists of land surface model SiBUC and its irrigation scheme was improved based on in situ measurement in local irrigated farms. Since it is physical model, not only past reproducing but also impacts of climate change and future irrigation projects can be quantitatively analyzed. In this study, impacts of climate change were projected using GCM outputs. Utilized GCM was MRI-AGCM3.2S based on A1B scenario which spatial resolution is 20km.

As a result of the analysis, water resource of whole basin will not change. That is because actual evapotranspiration is much less than potential one in arid area, where increment in precipitation can easily evaporate. On the other hand, water demand was projected to increase as expected. Irrigation times per year will increase and irrigation efficiency on farm will get worse especially at local furrow irrigation. Therefore, drought risk was projected to increase in this region and rational management will be furthermore required in the future.

     Makio dam in the Kiso River is a multipurpose dam which generates hydroelectric power and supplies water to the Aichi Irrigation Project. Reservoir sedimentation increased in this Dam because of Nagano West Earthquake in 1984. It cost about 30 billion yen to recover active storage capacity in 1996 by excavating deposited sediments. In the future, flow regime is going to change due to global warming which will have another impact on water resources management. In many dams in Japan including Makio dam, it is important to assess long-term vulnerability of water utilization by coupling impacts of reservoir sedimentation and climate change.

     In order to analyze this problem, we firstly made the dam operation model which is based on the actual operational rules. Secondly, we modified this model on some reservoir sedimentation scenarios based on annual rate of reservoir sedimentation and the quantity of the excavation in 1996. Additionally, we calculated future inflow changes by GCM model and distributed hydrological model (Hydro-BEAM). Finally, we evaluated economic effect in future (2093-2110) in terms of hydropower generation and water supply to the Aichi Irrigation Project.

     Regarding hydroelectric power generation, annual total generation can be almost maintained by modifying seasonal dam operation rule. Regardless of the sedimentation scenarios and flow-regime changes, annual total hydroelectric power generation can be maintained. However, water supply to the Aichi Irrigation Project will be seriously damaged due to a loss of active storage volume by reservoir sedimentation and changes in seasonal availability of discharge.