In this study, NHRCM20 was used to assess the uncertainty of changes in hydrological quantities due to climate change. 50 different present and 90 different future climate scenarios were generated. Land surface model SiBUC was used to calculate surface energy and water balance, then distributed hydrological model Hydro-BEAM was used to calculate river discharge. In the future, evapotranspiration increases and drought flow rate decreases in many areas of Japan. There is a possibility that the amount of water resource in Japan may decrease.
The climate impact studies on flood and inundation used high resolution MRI-AGCM at 20 km and NHRCM at 5 km was developed for further understanding of the impact of extreme rainfall. The objective is to assess the climate change impact on flood and inundation in a humid tropical river basin.
Basin averaged daily rainfall of NHRCM (RCP 8.5) was corrected using quantile-quantile mapping method (QM). In addition to QM, the coefficient of variance of rainfall over the space was corrected a certain range of percentiles. The corrected rainfall was used to simulate flood and inundation volume using Rainfall-Runoff-Inundation model (RRI).
Comparison of reproducibility of inundation volume shows that QM combined with spatial bias correction method is more reasonable than QM. In future, the inundation volume will be increasing at basin scale. The inundation will be more frequent along the river and at the downstream, which potentially affects agriculture in peatland.
Bioenergy and bioenergy with carbon capture and storage (BECCS) technologies can achieve zero or even negative CO2 emissions, hence have been considered as one of the essential technologies in achieving the 2-degree climate target. With ambitious climate policy, the demand for bioenergy would be up to 200-300 EJ per year based on recent predictions. At this level, a large volume of biomass is needed to generate energy. In order to enhance the simulation performance in bioenergy crop yield, we modified the algorithm and adjusted parameters of a state-of-the-art global hydrological model termed H08. Overall, overestimation or underestimation seen in the original H08 have been largely suppressed in this enhanced H08, which performs much better in simulating yield of Miscanthus and Switchgrass. Results also showed that irrigation significantly increased the yield for both Miscanthus and Switchgrass especially under dry climate condition.
The global hydrological H08 model and a new generation of routing model, CaMa-Flood, were successfully coupled to represent the effect of dams on river discharge and hence inundation dynamics. While, at the end of the 21st century, change in flood frequency highly depends on geographical factors, implementing dams reduced flood frequency for the majority of locations. In addition, implementing dams also reduced maximum flooded areas in major basins, up to 39%, compared to the same scenario with no dam.
Water scarcity indicators have been used for global water resources assessments, as there is no adequate physical quantity to indicate water scarcity. The most widely used indicators are the Withdrawal to Availability (WTA) and the Availability per capita (APC). WTA is the ratio of annual water withdrawal to annual water resources, and empirically, a region is judged under water scarcity when it exceeds 0.2 and 0.4. APC is the annual volume of water resources per person, and a region is judged so when it falls below 1700 m3/year/person and 1000 m3/year/person. Both are widely accepted, but the concrete basis for these thresholds has never been presented. Here, we conducted a series of global simulations using the latest global water resource model H08, and investigated what kind of water resources state the WTA and APC thresholds represent. This contribution is a summary of recent publication in Water Resources Research in 2018.
Water use in Central Asia highly depends on meltwater of glaciers. In order to know the meltwater’s seasonal availability, seasonal change of glacial water/heat balance should be clarified. This research aims to demonstrate SWE’s behavior on a land-surface process model SiBUC, which requires 7 meteorological factors for input data, using observed data and re-analysis data JRA55 as the input data. Since the sensor failed to observe snowfall, snowfall amount was reproduced by using observed snow depth and estimated density. SWE calculation was done in two ways; by using observed data (with reproduced precipitation), and by replacing each factor with JRA55, which aims to see the discrepancy between JRA55 and observed data and how it leads to the error in SWE. The result showed that the accuracy of short wave and long wave radiation particularly in the snowmelt season is crucial to the correct demonstration of SWE.
The existing rainfall-runoff models require discharge data for their calibration even though there will be uncertainties in the discharge data resulting from errors in rating curves. The direct prediction of observed water level will reduce the model uncertainties which is often sufficient to make an early warning about the flooding.
In this study, therefore, we aim to propose a generalized storage function (GSF) model for the water level prediction from the rating curve relationship by considering the spatial distribution of rainfall over the basin and incorporating all the possible inflow and outflow components in order to reduce the uncertainties involved.
The results revealed that the GSF model performed well in reproducing the water level hydrograph.
In recent years, many studies on rainfall and runoff using AI have been conducted. However, the rivers and floods targeted by each study are different, and it is difficult to determine the superiority or inferiority of the model being handled. Therefore, in order to improve the performance of AI rainfall runoff, it is desirable to study using common data. With this background, we created an urban small and medium river actual basin data set that contributes to the AI rainfall runoff benchmark test.
We applied the SCE-UA method to the parameter estimation of distributed runoff model. Target river basin is Toyokawa River, which catchment area is 724 km2. The model is an integrated model of a distributed runoff model and a one-dimensional unsteady flow model. The distributed runoff model incorporates saturation and unsaturated flow, and the mesh size is 100m. We input real river cross section shape into the one-dimensional unsteady flow model at the interval of 200 m. We use DioVISTA Flood for the modeling. We selected 4 parameters to be estimated by SCE-UA. Three flood events in recent years is selected as the target. Parameter estimated took 7 hours 30 minutes by using a PC. The error of the peak discharge was less than 9%, and the error of peak time was less than 5 minutes for the three flood events. A good parameter estimation was achieved by the proposed method.
For the paradigm shift from flood protection to floodplain risk management, it is critical to seek the effective application of point-based and probabilistic risk/hazard information, such as pin-point risk/hazard curves. In this study we illustrated the geo-spatial structure of probabilistic flood hazard, by ensemble flood simulation with d4PDF and RRI, and clustering analysis of discretised hazard curves.
From the case study in Yoshino river basin, Shikoku, Japan, we revealed that in the valley-bottom area, probabilistic hazard characteristics change from upstream to downstream; in contrast, in the meandering plain area, probabilistic hazard characteristics change parallelly with the main river and tributary channels. This difference is coming from both the basin-scale geological condition and small-scale topological condition. Further, the natural landforms such as natural levees or old channels could be extracted from only the hazard curve clustering. This result underlines the importance of natural landform classification to the floodplain risk management.
Since the latter half of the 1990s, nutrient salt in Harima-nada in Seto Inland Sea has been decreasing, a possible cause of which is the reduction of nutrient supply from flowing rivers. In order to seek countermeasures of the problem, it is necessary to calculate the nutrient inflow load from the rivers and estimate the nutrient concentration in the sea based on the inflow load to the sea. In this study, we constructed hydrological and water quality models and analyzed total nitrogen dynamics in the Kako river, which has the largest catchment area among the inflowing rivers, during rainfall and normal stage of water. Total nitrogen sources included point sources and rainfall-runoff driven nonpoint sources. The models well reproduced total nitrogen loads during both of rainfall and normal stage of water. In the future, we will expand the calculation area to basins of other rivers flowing into Harima-nada.
In this study, we examined flood mitigation effect by water release from irrigation ponds in Awaji district, Hyogo prefecture and proposed a method for selecting irrigation ponds with large flood mitigation effect. Firstly, we assumed that the water release is introduced in renovated irrigation ponds and estimated spillway width which can drain design flood safely in 1,902 irrigation ponds. Secondly, we conducted flood runoff analysis by the 10-year probability rainfall in all irrigation ponds and examined flood mitigation effect by the decrease of pond outflow in percentage terms for water release of 10% and 30% of total storage. Thirdly, we showed that the pond with large retention in mm (the space in pond by water release divided by catchment area) shows large reduction rate of peak discharge. Therefore, the retention is useful as an index of large flood mitigation effect by water release.
Integrated flood risk management (IFRM) implementation in Metro Manila is a challenging task to heavy reliance on traditional structural measures and because of the “barriers” that may hamper IFRM. This study presents for the first time The application for the Interpretive Structural Modeling (ISM) method to barrier analysis related to IFRM. This method is a systematic approach that analyzes the other The ISM model clearly showed that the barriers on the governance aspect are the most influential barriers that may dictate the movement of all other barriers.The ISM model produced in this study shows the interconnections of each barrier that can aid the decision makers and practitioners in Metro Manila, Philippines.
This study explores the potential use of a disaggregated flow duration curve (FDC) to estimate runoff in island catchments under humid conditions. The study disaggregates the FDC into three sections (top, middle and low) and attempts to estimate runoff in each section independently using simple hydrologic models. The results show the Curve Number method and the mean monthly flow (MMF) are able to make proper runoff estimations in the top and middle components respectively. For the low flow section, in perennial catchments the MMF or a process based Tank model is able to make proper estimations but not in ephemeral catchments. The ephemeral catchments low flows are estimated using the precipitation index. The study shows that in island catchments, climate is possibly the main control of the hydrologic nature.
Many researchers have been working on the physical mechanisms of hydrology and water resources from river basin scales to global scales. And the research using multiple kinds of research fields is becoming more necessary to clarify more complex and multi-aspect mechanisms such as combined cycle of water and materials, or socio-related issues.
Based upon the statement above, as a starting point, we have focused on the precipitation observation data which have been used as basic data in various research fields. The effect of climate change on snowfall and snowmelt intensity is hypothesized to alter the seasonality of river runoff patterns and material transportation and/or circulation in river basins. Therefore, in this study, we compared precipitation by radar and satellite observation during heavy snowfall event, in order to understand the characteristics and biases coming from the observation methods themselves.
To conserve water environment and ecosystem services in current river basin, it is necessary to understand circulation of material (e.g. carbon, nitrogen, phosphorus) that is nutrient for river and estuary ecosystem in river basin. And also, it is important that the change of water and material circulation in river basin from past to current is understood, to evaluate impacts of climate change on future. However, long-term weather observed data such as precipitation is difficult to be obtained in high elevation. In generally, reanalysis data has been used in a data scarce area. However, despite reanalysis data has been applied data assimilation method based on observation data, reanalysis data does not match the real weather characteristics (e.g. precipitation) in a local scale since it is targeted at a global scale. Therefore, this study aims to evaluate the pseudo-observational data creation method that is constructed based on previous study by focusing on weather data in mountainous area.
Recently, it is become increasingly important to adapt for the impact of climate change on various fields by establishment of the Climate Change Adaptation Plan. It is necessary to handle the climate prediction information such as General Circulation Models (GCMs) in order to discuss the impact and the adaptation on climate change. However, it is hard to beginner (e.g. administrator officer) use GCMs, due to GCMs have some bias relative to observational data since it is estimated by computational model, and these biases should be corrected by statistical method. Therefore, in order to examine the method that climate prediction information is widely utilized among industry-government-academia, this study aims to reveal the necessary conditions for using of prediction information by investigation of the approach of climate change impact assessment on water environment in lake and river basin based on collaboration and each point of view of industry-government-academia.
Reduction of hazards caused by stormflow is discussed from the interdisciplinary and transdisciplinary points of view. The interdisciplinary studies are important to evaluate the effects of watershed properties on the stormflow responses. In addition to this, the transdisciplinary studies are needed because various social problems such as environmental issues are involved in the decision making for the flood management. This study suggests that low-key strategies keeping the status quo about the prevention level of the flood may be important from the transdisciplinary point of view.
Recently, the need for social system design that takes into consideration climate change is increasing.In Japan, the Climate Change Adaptation Act was promulgated, and framework for promoting by industry-academia-government collaboration was established.Adaptation to climate change is a business opportunity from "industry" to "government", but their specific application has not been established yet."Academia" researchers are familiar with technology development while not with applications.
So, in this research, we aim to develop and present to the government "Social implementation model case for climate change adaptation method".As one of the concrete applications, we are starting to work on "Estimation of future flood risk change by flood emulator development".This joint research is not an analysis request from one side, but a technology development that benefits both industry and academia.Starting from this initiative, the goal is to solve problems that are more widely recognized in the field.
In this study, we analyze relationships between buildings location and flood damages in Mabi town, Okayama prefecture, and Ozu town, Ehime prefecture, Japan, where serious damage occurred due to the Heavy Rainfall in July 2018. We develop a historical building points database and apply the database to the damaged areas in order to compare building age and flood depth. The result shows that the 70 % of high damaged buildings (inundation depth: 2.0-2.5) in Mabi were built after 1979, and the buildings in both towns have moved into the high flood risk areas thorough their history. The “levee effect” are observed in the both towns, however, different development processes were existed in each towns.
The 2017 Kamaishi forest fire occurred for 14 days from 8th till 22nd May 2017 and the total burned area was 413ha which is greater than the total burnt area for the whole Japan in 2016. The burned area was estimated based on burned and unburned area. However, in the burned area itself, there were differences of fire severity observed. The objective of this research was to estimate the fire severity in this burned area using Sentinel 2A Normalized Difference Vegetation Index (NDVI) and post fire observation of scorch crown height, hs and relative scorch crown height, hsr. The results shows NDVI and hsr has stronger relationship than NDVI and hs, suggesting NDVI is more sensitive towards hsr.
In this study, the characteristics of runoff generated by land surface models SiBUC and MRI-SiB were examined from river discharge view point. From the analysis result by using NHRCM 5km output data, it was found that the amount of runoff in SiBUC was higher compared to MRI-SiB. In MRI-SiB, the base runoff was more dominant than surface runoff, while in SiBUC was the opposite. The generated runoff from both model was given to a flow routing model 1K-FRM to conduct river discharge simulation. The 10-years-mean monthly river discharge analysis showed the different of discharge volume and timing of peak discharge from both models, which was due to different runoff amount and their characteristics.
To enhance the accuracy of runoff estimation by GCMs (General Circulation Models), effects of land surface parameters on runoff estimation were investigated by using MRI-AGCM3.2H. In this study, two experiments were conducted. One was simulation with default saturated hydraulic conductivity, the other one was simulation with low value saturated hydraulic conductivity. As a result of setting low saturated hydraulic conductivity, surface runoff was increasing and subsurface runoff was decreasing. In some area, the magnitude of increasing surface runoff exceeded that of decreasing subsurface runoff and vice versa.