Return period is one of the important indices of the natural hazards (eg. heavy rains), which does not only indicate the magnitude and the frequency of the largest hazards for decision making of flood protection plans but also plays several important roles in the mathematical theories to deal with an univariate extreme value and multivariate joint ones. Firstly, we introduce two manners to establish the definition of return period: 1) The exceedance probability for the annual maximum and 2) The occurrence rate in the Poisson process of extremes exceeding a threshold are possible to be regards as the fraction 1/Return period. There is a simple mathematical formula of exchange both quantities, and it will not really matter to employ the latter in place of the former which has been used conventionally for the simplicity. We will remark another simplicity in multivariate extremes, which employs unit Frechet distribution for the marginals. And we should struggle with the joint occurrence rate of Threshold EXcess Multivariate EXtremes (TEXTMEX) rather than the Componentwise Maxima (CM) of annual largest values which is not always given by the identical simultaneous meteorological factor.

Recent studies have further demonstrated that precipitation patterns have been affected by climate change, leading to severe flood. Dam operation in flood event is important as climate change adaptation. However, flood control operation during an extreme flood (hereafter, extreme flood operation) that is carried out in the case of the storage ratio of a dam exceed about 80 % causes a serious damage to the downstream area of the dam. Therefore, this study aims to estimate the impact of climate change on frequency of river flow regime and the probability of extreme flood operation using simple runoff model and dam operation model which is developed based on actual dam operation rule with d4PDF. In this study, we selected the Shimouke Dam, in which it has been carried out in July 2020, and Matsubara Dam, which is located directly below the Shimouke Dam as research fields. As the result, it was shown that low frequency discharge like flood discharge in the future (under +4K scenario) might be increased relative to the present, since extreme precipitation would be increased such as maximum 24 hour precipitation. On the other hand, it was suggested that the high frequency discharge like low discharge would be decreased, since the frequency of precipitation times might be decreased in the future. Furthermore, the results of the study revealed that the frequency of extreme flood operation by the Shimouke Dam in the future might increase relative to the present with the increase of flood discharge. On the other hand, the frequency of the operation of the Matsubara Dam in the future might be almost same as the present. Therefore, it was revealed that the connected two dams were able to operate effectively in order to reduce damage to the downstream area of the dams.  

In recent years, Japan has suffered from heavy rains almost every year, causing huge damage on people property. In 2019, the amount of damage was the highest ever, and the government  a urban planning policy "location optimization", which attempts to reduce damage by guiding residents to low-flood-risk areas, as alternative flood risk management. This measure cannot be evaluated with static models such as general equilibrium models because actual policy induces long-terms changes in the location structure. In previous studies, therefore, an agent-type location choice model was adopted as its evaluation method, which enabled to quantify the risk reduction effect associated with location guidance. However, the previous version of this model did not consider flood disaster occurrence, but flood risk affects the location choice of households only through the payment of flood insurance whose price depends on local flood risk derived from flood risk curves. In addition, house prices also fluctuated at the beginning of the simulation period, which was unrealistic in actual house market. Therefore, in this study, we improved this model by introducing asset update and flood disaster occurrence processes to make it possible to update and lose their assets by flood disaster events. As a result, the number of households living in high-risk areas ( and thereby the number of properties affected by the disaster) has increased due to increased priority in economic activities compared to flood risk avoidance. In addition, house market in the model became more realistic by suppressing price fluctuations of house prices.

In this study, a vertical rainwater harvesting facility that can collect rainwater through the wall was developed and used. The RHF was manufactured in a total of three types according to its height. The heights of the RHFs in this study are 1.00 m, 1.50 m and 2.00 m. This study estimated the amount of rainwater that could be harvested by the RHF and compared it with the measured amount of rainwater. In the process of estimation, the empirical equation for estimating the wind-driven-rain on building wall was applied. This study used BloomSky weather device to observe the wind speed and rainfall for every minute. BloomSky weather device had been installed on the rooftop of the Engineering Building in Korea University. In this study, a total of 16 rainfall events were observed from June 10, 2020 to September 7, 2020. Rainfall and wind speed were observed for every minute using BloomSky weather device. As a result, estimated rainwater in RHF was found to be similar to that was measured by the RHF. The average of estimated and measured rainwater in RHF for 16 rainfall events were 1.770 mm/hr∙m² and 1.878 mm/hr∙m², respectively. In the case of standard deviation, it was 1.269 mm/hr∙m² for estimated rainwater and 1.181 mm/hr∙m² for measured rainwater. Acknowledgements : This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT) (No. 2020R1A2C2008714).

In July 2020, the first class river Kumagawa flooded due to “the July 2020 heavy rain event”, this flood induced severe damage around Hitoyoshi city and Kuma village.To predict this disaster, we simulated the Japan Meteorology Agency Runoff Index model (JMA-RI model) with 1000 and 100 ensemble weather forecast data. The ensemble forecast was simulated by LETKF and JMA-NHM. The spinup period of the RI model was from 01 June 00:00 (JST) to 03 July 18:00 and the simulation started from 03 July 19:00 to 04 July 15:00. To evaluate the performance of the 1000 ensemble simulation, we also used 100 ensemble data. The RI model with the 1000 data well simulated the peak of the flood. The probability of exceeding the largest record flood is 56%. Meanwhile, in the RI model with the 100 data, the probability of exceeding the largest record flood is 51%. These results shows that large ensemble simulation improve accuracy of the flood prediction model.

Maki district, Ozu, Kumamoto, is located in the western region of the Aso caldera rim. The Yagogawa River and three springs located in this district have been used for irrigation. Additionally, since 2013, the springs have been subjected to winter flooding by the Kumamoto Groundwater Foundation. However, the water flow in the upper reaches of the Yagogawa River was cutoff in May 2016, which resulted in the drying up of one of the springs in February 2018. Under these circumstances, paddy fields of approximately 20 ha were abandoned and partial winter flooding was inhibited. Therefore, in this study, we investigated the causes of the cut-off and subsequent depletion of the water flow by the field survey with the aim of developing strategies to revive the river and spring. A closed-type sabo dam was constructed in the upper reaches of the river because of the debris flow caused by heavy rainfall in the northern Kyushu District in July 2012. The dam, however, resulted in decreased water flow downstream because six drainage holes became clogged. The rapid rise in groundwater levels after the 2016 Kumamoto earthquake may have caused the forceful drainage of mountainous groundwater and significantly reduced base runoff to the river upstream of the dam. Specifically, the earthquake and the sabo dam which was constructed because heavy rainfall were considered to represent the major factors for the river cut-off. Furthermore, river water flows underground in some regions upstream of the spring. The Ca-HCO₃ type constitution of the river water was the same as that of the spring. Based on this viewpoint, spring water can be considered as underflow water, which suggests that the recovery of the river is essential for the revival of the spring.

Recently, flood disasters are occurring frequently not only in Korea but also around the world due to climate change. In particular, the damage caused by torrential rains is increasing in urban areas. Accordingly, research on prevention and evacuation measures for flood disasters has been steadily increasing. However, climate change is difficult to predict even with advance prevention, so it is necessary to evacuate to a safe place. In addition, vulnerable people need special attention because they are more prone to disastrous situations as compared to ordinary people. Therefore, in this study, we analyzed Dongnae and Yeonje district, which are areas with a high risk of urban inundation and often suffered damages in Busan Metropolitan City. The evacuation time was analyzed from the perspective of the vulnerable people by determining the proper place of evacuation centers in consideration of the flood range in Dongnae and Yeonje district. Furthermore, it was not limited to each district, but also reviewed the effects of wide-area evacuation. The results of this study are considered to be useful for establishing an evacuation plan suitable for disaster vulnerable areas.

Urban flash floods have caused severe damage frequently. One of the challenges to address the urgent issue of flash floods in the tropical Asian urban areas is to create the management time required to obtain and analyze hydrometeorological data. This study attempted to develop an integrated real-time flood forecasting system at Ramkhamheng polder, Bangkok, Thailand. This area has recently been particularly vulnerable to flash floods, resulting in significant damages to municipal infrastructures. The forecasting system was based on predicted precipitations with radar images and a hydraulic model. This system consists of three main parts. The first part obtains regional radar images every 10 minutes from the C-band radar and 30 minutes of rainfall forecasting from TITAN software. The second part uses a hydraulic model, PCSWMM, with the current and forecasted precipitation obtained at the first part to simulate flash floods. The third part was modified from the PCSWMM real-time software. This part manages all the forecasting system tasks, including input data preparation, simulation order, and output data providing post-processing. The forecasting results with 116 flash floods' historical data showed that the system provided a pre-warning accurately for the 71 flash floods. The whole system's accuracy depended heavily on the rainfall forecasting in the first part, which currently had 69.1% accuracy in the 116 events.  As for the lead time, the present forecasting system provided potential flood notification 30 minutes earlier than the conventional flood model using radar and local-station data individually. It could allow management staff to make more spatially explicit flood control decisions, including pump and gate operations, and provide the forecast maps available on the internet and social media for public information. These results strongly support that the integrated flood forecasting system in this study could be a useful tool for real-time flash flood alert and management.

Evapotranspiration (ET) is one of the main components of the global and regional hydrological or water cycle. Satellite remote sensing of ET has been applied to monitoring regional and global droughts in recent decades and is probably the most practical and efficient approach to providing ET for the numerical weather, climate, and hydrological prediction models. Based on the Atmosphere-Land Exchange Inversion (ALEXI) model, NOAA-NESDIS has developed an operational GOES ET and Drought (GET-D) product system using thermal infrared (TIR) observations of the Baseline Imagers on GOES-13 and GOES-15 satellites. The Advanced Baseline Imagers (ABIs) on the new generation of NOAA geostationary operational satellites allow the GET-D system to be significantly enhanced with higher spatial resolution and better accuracy that is needed for a variety of applications. The GET-D system is therefore upgraded to generate ET data products using ABI observations at 2 km spatial resolution covering the continental United States (CONUS). Additionally, given that the current GET-D product is dependent on the availability of remotely sensed TIR observations, the ALEXI model can only be executed under clear-sky conditions. A big advance in the new system is to combine microwave observations with the TIR data using data mining technique to obtain all-weather land surface temperature (LST) that is then used to derive all-weather ET. This study exploited the combination of AMSR2 Ka-band brightness temperature and GOES LST to come up with a surface temperature map under all weather conditions. Preliminary results demonstrated that the all-weather ET product increases data availability by more than 160%, which would be extremely beneficial to the end-users for a variety of applications. This presentation will provide validation results of all-weather ET and clear-sky ET products by comparing them with the in-situ ET measurements from more than 50 stations of the AmeriFlux networks over CONUS.

The energy non-closure near the land surface has been a key topic in the land surface processes research.  The energy closure rate is still not high even after considering heat storage and photosynthesis energy consumption, while the contribution of advective energy to the closure rate needs to be considered further under the non-uniform underlying surface. In this paper, the advective energy caused by thermal heterogeneity of underlying surface is calculated by using the energy budget data collected from the Flower-Lake observation site in the Zoige Alpine Wetland in 2017, and the contribution of thermal advection to energy closure near the ground is estimated. The result shows: In summer of 2017, the maximum value of the advective heat flux was 23.8w/m2 at the Zoige alpine wetland. When the contribution of advective heat flux is introduced into the energy balance equation, the energy closure rate increases from 72.0% to 79.4%. With considering the contribution of horizontal heat transfer, it has a certain effect on improving energy closure rate for the flat terrain and thermal inhomogeneous underlying surface. The near surface thermal inhomogeneity leads to the accumulation of heat, which is the basic reason for the heat advection to affect the energy closure rate, and also an important reason for the difference between the wetland characteristics of water and heat exchange of the wetland with the other regions.

We use 40-year-long time series originating from over 13 000 stations around the globe for the identification of patterns related to the spatial variability of mean monthly temperature, total monthly precipitation and mean monthly river flow features. To formalize this identification, we also develop and apply a new hydroclimatic time series clustering methodology that can be perceived as a geographical location clustering methodology (since hydroclimatic time series correspond to geographical locations). This new methodology is based on Breiman’s random forests and on approximately 60 diverse features. The latter are mostly sourced from scientific fields beyond geoscience and environmental science (e.g., the fields of neuroscience, biology, biomedicine and forecasting), thereby constituting a new concept for our fields. The spatial coherence characterizing the clusters delivered by the new methodology build confidence in its future exploitation for the delivery of descriptive and exploratory insights. Given this spatial coherence and the scale-independent nature of the computed feature values (which makes them particularly useful in forecasting and simulation contexts), we believe that this methodology could also be beneficial within regionalization frameworks, in which knowledge on hydrological similarity is exploited in technical and operative terms.

Information of evapotranspiration on a rice paddy is important as a basis for national water resource planning and design of hydraulic structures such as reservoir. Since the amount of paddy evapotranspiration is difficult to measure, most of it is estimated by the FAO Penman-Monteith equation. However, the Penman-Monteith equation is developed based on upland evapotranspiration, which does not reflect the characteristics of paddy fields. In this study, the evapotranspiration model in the Agricultural Policy and Environmental eXtender (APEX) model, which is a farmland and watershed model, was modified in consideration of rice cultivation characteristics, and the relevant parameters were corrected. H₂O flux at the paddy field measured for three years from 2015 to 2017 in Gimje, Jeollabuk-do, South Korea were used for the calibration and validation. Atmospheric gas fluxes were measured 10 times per second (10 Hz) with a CO₂ / H₂O gas analyzer (LI-7500, LI-COR Biosciences Inc. USA), and the 30 minute average physical flux was calculated by covariance with the 30 minute average. And it was organized by daily data. The Penman-Monteith equation and the Hargreaves equation were used as a potential evapotranspiration model. The soil evaporation rate was replaced to the water surface evaporation rate and the water surface evaporation coefficient was set to 1.0. As a result, R² were improved from 0.70 to 0.77 for the calibration period and 0.62 to 0.70 for the validation period, respectively. Financial Support: This work was carried out by the support of Cooperative Research Program for Agriculture Science & Technology Development (PJ014937), Rural Development Administration, Republic of Korea.

As a vital resource for survival, development and ecological demands, water is always the first issue of our consideration. With the climate change and socioeconomic development over the Pearl River Basin (PRB), considerable changes have occurred in the water cycle and the uncertainty of water availability and water usage increases. It is of great importance to comprehend the change of hydrological and sediment processes of the basin to enable sustainable water utilization. In this study, SWAT (Soil and Water Assessment Tool) is selected to simulate the terrestrial processes in the PRB at daily time step for the period from 2008 to 2018 and the observed streamflow data at daily time step from 9 streamflow gauge stations are used to calibrate and validate the model. This study will explore the challenge and difficulty in improving the model simulation results, and the sensitivity study of key model parameters is conducted. Meanwhile, the physical meaning of the key parameters is explored to evaluate the model performance, and then the improvement of model simulation accuracy is discussed. The results can improve our understanding of the features of those watersheds and therefore contribute to enhancement in planning, management, and operation.

 As the frequency and severity of droughts increases, drought countermeasures are being sought from various aspects. In Korea, precipitation is concentrated in summer, so water stress is very high at each season. In addition, depending on the region, there are areas with abundant water and with insufficient water, so it is very important how to properly distribute water. About 1.5 million groundwater wells have been developed in Korea, but it was not enough to solve this spatiotemporal inequality. However, most wells are not always used throughout the year, and the frequency of use is generally low, depending on the region. Therefore, research is being conducted to develop a system that properly allocates groundwater resources by linking these wells with each other. Several pumping wells are physically or virtually connected to extract groundwater, and an optimal groundwater supply plan is suggested through an algorithm. The pumping well can be controlled remotely, and the monitoring well checks whether excessive water is pumped. It is named as Well Network System, and the water level and water quality are automatically monitored through sensors in pumping wells and observation wells. The research area is the Yanggok-ri area of Hongseong-gun, and in order to apply the Well Network System, water balance analysis, groundwater impact modeling according to pumping, and seasonal use surveys have been conducted. Although it is a small area and there is a difference in supply and use of groundwater in the upstream and downstream areas, it is expected that water can be distributed rationally to areas with a lot of water and areas with less water with the system of the interconnecting existing wells. Acknowledgements : This work was supported by Korea Environment Industry & Technology Institute(KEITI) through Demand Responsive Water Supply Service Program(or Project), funded by Korea Ministry of Environment(MOE)

Occurrence of long-term drought is mainly associated with long-range dependency (LRD) that are considered as one of the intrinsic characteristics of natural process. Since severe droughts tend to persist multiple years come more frequently under changing climate, robust hedging rules should be prepared for adaptation to those long-lasting droughts. In this study, a simple and efficient way for developing time-varying hedging rules which can improve resiliency of reservoir operation is introduced. First, sequent peak algorithm is used to derive multiple hedging rules that play a role of contingency reservoir operation under long-term drought. Next, to evaluate the performance of the developed hedging rules, a synthetic streamflow generation scheme-that can produce a wide range of LRD along with historical flow statistics-is proposed. Autoregressive fractionally integrated average (ARFIMA) model and periodic autoregressive (PAR) model are embedded in the synthetic generation model. As a result, the developed hedging rules enhanced the performance of reservoir operation in terms of resiliency and vulnerability indices. Notably, the developed hedging rules outperformed the benchmark hedging rules (not considering LRD), especially under long-term dry conditions. The proposed method based on the sequent peak algorithm can be easily applied to different reservoirs. Besides, it enables phased water rationing by providing simple discrete guidelines for reservoir operators. LRD impact on reservoir operation was successfully assessed by taking advantage of the stochastic streamflow generator, which reflects the Hurst coefficient in its annual series generation. Under the circumstance in which the frequency of long-term droughts and their severity are increasing, it is imperative that a wide range of scenarios that reflect LRD are considered to water resources management practices.

Cropland nutrients are discharged along the surface runoff, percolating water and eroded soil by rainfall. Nutrients which excessively input to agricultural lands can lead to pollution of aquatic ecosystems. The discharge of nutrients from crop land is referred to as non-point source (NPS) pollution which is difficult to control because it is not known exactly where they occur and discharged in low concentration and large volume. In this study, a test plot consisted of 9% slope, 24 m slope length, and 3 types of soils (sandy loam, loam, silty loam) was constructed to analyze the BMP effect of the field. The measurement period was 2018-2019, and the crops were corn - Chinese cabbage in 2018 and corn in 2019. Furrow vegetation mulching and furrow fabric mulching were applied as BMPs. Alfalfa was used for vegetation mulching. Runoff was measured using a 1/100 sampler, and water quality was analyzed in accordance with the Water Quality Test Act of South Korea. Results of Y2018 survey showed that the effect of vegetation mulching on the reduction of load of T-N and T-P was 6.4% to 24.5% and 4.5% to 27.0%, respectively. The effect of reducing the load of T-N and T-P by the fabric mulching was 1.0 to 29.7% and 10.7 to 47.4%, respectively. In Y2019, the survey showed that the effect of vegetation mulching on the reduction of load of T-N and T-P was 14.6% to 61.5% and 19.0% to 49.1%, respectively. The effect of reducing the load of T-N and T-P by the fabric mulching was 28.8 to 39.7% and 23.3 to 52.2%, respectively. Financial Support: This work was carried out by the support of Cooperative Research Program for Agriculture Science & Technology Development (PJ014932), Rural Development Administration, Republic of Korea.

There are many people lacking access to safe drinking water, especially low-income families living in remote areas. The characteristics of the water supply (type of sources, distribution system, treatment facility etc.) in remote rural areas are quite different from the water supply in urban area. It faces many challenges in terms of water quality and quantity. Most risk assessments in the past have focused on large and centralized water supply systems, while small-scale water supply systems or household water supply systems have not been well studied. Much less information is available for these systems or household water supply units due to the information is not accessible. The quantitative characterization of these water supplies and their internal relations, is the key issue to undertake a risk assessment. Therefore, on the basis of recognizing the complexity of the water supply system, fuzzy logic theory is introduced to characterize the subjective information, and the Analytic Hierarchy Process (AHP) is applied to determine the weight of the risk indices. In this paper, five types of small drinking water supply systems in five regions of Gansu Province, China, are selected as case studies. The risk assessment in this study uses 11 risk factors, and the weighted result shows that "the source of the water " is the most important factor. The extreme events include drought and flood increased the risk of safe drinking water supply in all the regions. The risks in summer and fall are higher than those in spring and winter. Flood has a greater impact on centralized water supply than all other supplies, while groundwater supply and rainwater supply are more sensitive to drought. This research provides a systematic and practical method for decision-makers to plan the most effective risk mitigation measures for water supply systems in remote areas with scientific support.