Ensemble flash flood predictions using a high-resolution nationwide distributed rainfall-runoff model: case study of the heavy rain event of July 2018 and Typhoon Hagibis in 2019

AbstractThe heavy rain event of July 2018 and Typhoon Hagibis in October 2019 caused severe flash flood disasters in numerous parts of western and eastern Japan. Flash floods need to be predicted over a wide range with long forecasting lead time for effective evacuation. The predictability of flash floods caused by the two extreme events is investigated by using a high-resolution (~ 150 m) nationwide distributed rainfall-runoff model forced by ensemble precipitation forecasts with 39 h lead time. Results of the deterministic simulation at nowcasting mode with radar and gauge composite rainfall could reasonably simulate the storm runoff hydrographs at many dam reservoirs over western Japan for the case of heavy rainfall in 2018 (F18) with the default parameter setting. For the case of Typhoon Hagibis in 2019 (T19), a similar performance was obtained by incorporating unsaturated flow effect in the model applied to Kanto Region. The performance of the ensemble forecast was evaluated based on the bias ratios and the relative operating characteristic curves, which suggested the higher predictability in peak runoff for T19. For the F18, the uncertainty arises due to the difficulty in accurately forecasting the storm positions by the frontal zone; as a result, the actual distribution of the peak runoff could not be well forecasted. Overall, this study showed that the predictability of flash floods was different between the two extreme events. The ensemble spreads contain quantitative information of predictive uncertainty, which can be utilized for the decision making of emergency responses against flash floods.

Download Full-text

Ensemble Flash Flood Predictions Using a High-Resolution Nationwide Distributed Rainfall-Runoff Model: Case Study of the Heavy Rain Event of July 2018 and Typhoon Hagibis in 2019

10.21203/rs.3.rs-40714/v2 ◽

2020 ◽

Author(s):

Takahiro Sayama ◽

Masafumi Yamada ◽

Yoshito Sugawara ◽

Dai Yamazaki

Keyword(s):

Lead Time ◽

Flash Flood ◽

Heavy Rain ◽

Flash Floods ◽

Rain Event ◽

Rainfall Runoff ◽

Heavy Rain Event ◽

Peak Runoff ◽

Rainfall Runoff Model ◽

Runoff Model

Abstract The heavy rain event of July 2018 and Typhoon Hagibis in October 2019 caused severe flash flood disasters in numerous parts of western and eastern Japan. Flash floods need to be predicted over a wide range with long forecasting lead time for effective evacuation. The predictability of flash floods caused by the two extreme events are investigated by using a high-resolution (~150 m) nationwide distributed rainfall-runoff model forced by ensemble precipitation forecasts with 39-h lead time. Results of the deterministic simulation at nowcasting mode with radar and gauge composite rainfall could reasonably simulate the storm runoff hydrographs at many dam reservoirs over western Japan for the case of heavy rainfall in 2018 (F18) with the default parameter setting. For the case of Typhoon Hagibis in 2019 (T19), a similar performance was obtained by incorporating unsaturated flow effect in the model applied to Kanto region. The performance of the ensemble forecast was evaluated based on the bias ratios and the relative operating characteristic curves, which suggested the higher predictability in peak runoff for T19. For the F18, the uncertainty arises due to the difficulty in accurately forecasting the storm positions by the frontal zone; as a result, the actual distribution of the peak runoff could not be well forecasted. Overall, this study showed that the predictability of flash floods was different between the two extreme events. The ensemble spreads contain quantitative information of predictive uncertainty, which can be utilized for the decision making of emergency responses against flash floods.

Download Full-text

Ensemble Flash Flood Predictions Using a High-Resolution Nationwide Distributed Rainfall-Runoff Model: Case Study of the Heavy Rain Event of July 2018 and Typhoon Hagibis in 2019

10.21203/rs.3.rs-40714/v1 ◽

2020 ◽

Author(s):

Takahiro Sayama ◽

Masafumi Yamada ◽

Yoshito Sugawara ◽

Dai Yamazaki

Keyword(s):

Lead Time ◽

Flash Flood ◽

Heavy Rain ◽

Flash Floods ◽

Rain Event ◽

Rainfall Runoff ◽

Heavy Rain Event ◽

Peak Runoff ◽

Rainfall Runoff Model ◽

Runoff Model

Abstract The heavy rain event of July 2018 and Typhoon Hagibis in October 2019 caused severe flash flood disasters in numerous parts of western and eastern Japan. Flash floods need to be predicted over a wide range with long forecasting lead time for effective evacuation. The predictability of flash floods caused by the two extreme events are investigated by using a high-resolution (~ 150 m) nationwide distributed rainfall-runoff model forced by ensemble precipitation forecasts with 39-h lead time. Results of the deterministic simulation at nowcasting mode with radar and gauge composite rainfall could reasonably simulate the storm runoff hydrographs at many dam reservoirs over western Japan for the case of heavy rainfall in 2018 (F18) with the default parameter setting. For the case of Typhoon Hagibis in 2019 (T19), a similar performance was obtained by incorporating unsaturated flow effect in the model applied to Kanto region. The performance of the ensemble forecast was evaluated based on the bias scores and the relative operating characteristic curves, which suggested the higher predictability in peak runoff for T19. For the F18, the uncertainty arises due to the difficulty in accurately forecasting the storm positions by the frontal zone; as a result, the actual distribution of the peak runoff could not be well forecasted. Overall, this study showed that the predictability of flash floods was different between the two extreme events. The ensemble spreads contain quantitative information of predictive uncertainty, which can be utilized for the decision making of emergency responses against flash floods.

Download Full-text

The benefit of high-resolution operational weather forecasts for flash flood warning

Hydrology and Earth System Sciences ◽

10.5194/hess-12-1039-2008 ◽

2008 ◽

Vol 12 (4) ◽

pp. 1039-1051 ◽

Cited By ~ 54

Author(s):

J. Younis ◽

S. Anquetin ◽

J. Thielen

Keyword(s):

High Resolution ◽

Short Range ◽

Weather Forecasting ◽

Flash Flood ◽

Flash Floods ◽

Rainfall Runoff ◽

Weather Forecasts ◽

Numerical Weather ◽

Rainfall Runoff Model ◽

Runoff Model

Abstract. In Mediterranean Europe, flash flooding is one of the most devastating hazards in terms of loss of human life and infrastructures. Over the last two decades, flash floods have caused damage costing a billion Euros in France alone. One of the problems of flash floods is that warning times are very short, leaving typically only a few hours for civil protection services to act. This study investigates if operationally available short-range numerical weather forecasts together with a rainfall-runoff model can be used for early indication of the occurrence of flash floods. One of the challenges in flash flood forecasting is that the watersheds are typically small, and good observational networks of both rainfall and discharge are rare. Therefore, hydrological models are difficult to calibrate and the simulated river discharges cannot always be compared with ground measurements. The lack of observations in most flash flood prone basins, therefore, necessitates the development of a method where the excess of the simulated discharge above a critical threshold can provide the forecaster with an indication of potential flood hazard in the area, with lead times of the order of weather forecasts. This study is focused on the Cévennes-Vivarais region in the Southeast of the Massif Central in France, a region known for devastating flash floods. This paper describes the main aspects of using numerical weather forecasting for flash flood forecasting, together with a threshold – exceedance. As a case study the severe flash flood event which took place on 8–9 September 2002 has been chosen. Short-range weather forecasts, from the Lokalmodell of the German national weather service, are used as input for the LISFLOOD model, a hybrid between a conceptual and physically based rainfall-runoff model. Results of the study indicate that high resolution operational weather forecasting combined with a rainfall-runoff model could be useful to determine flash floods more than 24 h in advance.

Download Full-text

The benefit of high-resolution operational weather forecasts for flash flood warning

Hydrology and Earth System Sciences Discussions ◽

10.5194/hessd-5-345-2008 ◽

2008 ◽

Vol 5 (1) ◽

pp. 345-377 ◽

Cited By ~ 12

Author(s):

J. Younis ◽

S. Anquetin ◽

J. Thielen

Keyword(s):

High Resolution ◽

Weather Forecasting ◽

Flash Flood ◽

Flood Forecasting ◽

Flash Floods ◽

Rainfall Runoff ◽

Weather Forecasts ◽

Numerical Weather ◽

Rainfall Runoff Model ◽

Runoff Model

Abstract. In Mediterranean Europe, flash flooding is one of the most devastating hazards in terms of human life loss and infrastructures. Over the last two decades, flash floods brought losses of a billion Euros of damage in France alone. One of the problems of flash floods is that warning times are very short, leaving typically only a few hours for civil protection services to act. This study investigates if operationally available shortrange numerical weather forecasts together with a rainfall-runoff model can be used as early indication for the occurrence of flash floods. One of the challenges in flash flood forecasting is that the watersheds are typically small and good observational networks of both rainfall and discharge are rare. Therefore, hydrological models are difficult to calibrate and the simulated river discharges cannot always be compared with ground "truth". The lack of observations in most flash flood prone basins, therefore, lead to develop a method where the excess of the simulated discharge above a critical threshold can provide the forecaster with an indication of potential flood hazard in the area with leadtimes of the order of the weather forecasts. This study is focused on the Cévennes-Vivarais region in the Southeast of the Massif Central in France, a region known for devastating flash floods. The critical aspects of using numerical weather forecasting for flash flood forecasting are being described together with a threshold – exceedance. As case study the severe flash flood event which took place on 8–9 September 2002 has been chosen. The short-range weather forecasts, from the Lokalmodell of the German national weather service, are driving the LISFLOOD model, a hybrid between conceptual and physically based rainfall-runoff model. Results of the study indicate that high resolution operational weather forecasting combined with a rainfall-runoff model could be useful to determine flash floods more than 24 hours in advance.

Download Full-text

Uncertainty Evaluation in a Flood Forecasting Model Using JMA Numerical Weather Prediction

Journal of Disaster Research ◽

10.20965/jdr.2009.p0272 ◽

2009 ◽

Vol 4 (4) ◽

pp. 600-605 ◽

Cited By ~ 5

Author(s):

Hadi Kardhana ◽

◽

Akira Mano ◽

Keyword(s):

Numerical Weather Prediction ◽

Lead Time ◽

Weather Prediction ◽

Japan Meteorological Agency ◽

Rainfall Runoff ◽

Flood Prediction ◽

Numerical Weather ◽

Forecast Lead Time ◽

Rainfall Runoff Model ◽

Runoff Model

Numerical weather prediction (NWP) is useful in flood prediction using a rainfall-runoff model. Uncertainty occurring in the forecast, however, adversely affects flood prediction accuracy, in addition to uncertainty inherent in the rainfall-runoff model. Clarifying this uncertainty and its magnitude is expected to lead to wider forecast applications. Taking the case of Japan’s Shichikashuku Dam, 6 flood events between 2002 and 2007 were analyzed. NWP was based on short-range forecasts by the Japan Meteorological Agency (JMA). The rainfall-runoff model is based on a distributed tank model. This research calculates uncertainty by identifying and quantifying the relative error of forecasts by a) NWP and b) the runoff model. Results showed that NAP is the main cause of flood forecast uncertainty. They also showed the correlation between forecast lead time and uncertainty. Uncertainty rises with longer lead time, corresponding to the magnitude of observed discharge and precipitation.

Download Full-text

The challenge of forecasting impacts of flash floods: test of a simplified hydraulic approach and validation based on insurance claim data

10.5194/hess-2017-344 ◽

2017 ◽

Author(s):

Guillaume Le Bihan ◽

Olivier Payrastre ◽

Eric Gaume ◽

David Moncoulon ◽

Frederic Pons

Keyword(s):

Cross Sections ◽

Flash Flood ◽

Claim Data ◽

Flash Floods ◽

Water Levels ◽

Claim Database ◽

Rainfall Runoff ◽

Insurance Claim ◽

Local Reference ◽

Rainfall Runoff Model

Abstract. Up to now, flash flood monitoring and forecasting systems, based on rainfall radar measurements and distributed rainfall-runoff models, generally aimed at estimating flood magnitudes - typically discharges or return periods – at selected river cross-sections. The approach presented here goes one step ahead by proposing an integrated forecasting chain for the direct assessment of flash flood possible impacts on inhabited areas (number of buildings at risk in the presented case studies). The proposed approach includes, in addition to a distributed rainfall-runoff model, an automatic hydraulic method suited for the computation of flood extent maps on a dense river network and over large territories. The resulting catalogue of flood extent maps is then combined with land use data to build a flood impact curve for each considered river reach: i.e. number of inundated buildings versus discharge. Theses curves are finally used to compute estimated impacts based on forecasted discharges. The approach has been extensively tested in the regions of Alès and Draguignan, located in the South of France, where well documented major flash floods recently occurred. The article presents two types of validation results. First, the automatically computed flood extent maps and corresponding water levels are tested against rating curves at available river gauging stations as well as against local reference or observed flood extent maps. Second, a rich and comprehensive insurance claim database is used to evaluate the relevance of the estimated impacts for some recent major floods.

Download Full-text

Multistage stochastic linear programming model for daily coordinated multi-reservoir operation

Journal of Hydroinformatics ◽

10.2166/hydro.2008.007 ◽

2008 ◽

Vol 10 (1) ◽

pp. 23-41 ◽

Cited By ~ 9

Author(s):

Yongdae Lee ◽

Sheung-Kown Kim ◽

Ick Hwan Ko

Keyword(s):

Stochastic Model ◽

Reservoir Operation ◽

Lead Time ◽

Weather Prediction ◽

Prediction System ◽

Operation Planning ◽

Rainfall Runoff ◽

Multi Stage ◽

Rainfall Runoff Model ◽

Runoff Model

Operation planning for a coordinated multi-reservoir is a complex and challenging task due to the inherent uncertainty in inflow. In this study, we suggest the use of a new, multi-stage and scenario-based stochastic linear program with a recourse model incorporating the meteorological weather prediction information for daily, coordinated, multi-reservoir operation planning. Stages are defined as prediction lead-time spans of the weather prediction system. The multi-stage scenarios of the stochastic model are formed considering the reliability of rainfall prediction for each lead-time span. Future inflow scenarios are generated by a rainfall–runoff model based on the rainfall forecast. For short-term stage (2 days) scenarios, the regional data assimilation and prediction system (RDAPS) information is employed, and for mid-term stage (more than 2 days) scenarios, precipitation from the global data assimilation and prediction system (GDAPS) is used as an input for the rainfall–runoff model. After the 10th day (third stage), the daily historical rainfall data are used following the ensemble streamflow prediction (ESP) procedure. The model is applied to simulate the daily reservoir operation of the Nakdong River basin in Korea in a real-time operational environment. The expected benefit of the stochastic model is markedly superior to that of the deterministic model with average rainfall information. Our study results confirm the effectiveness of the stochastic model in real-time operation with meteorological forecasts and the presence of inflow uncertainty.

Download Full-text

Hydrologic-hydraulic coupling for flash flood real-time simulation: Application to the October 2015 French Riviera floods

10.5194/egusphere-egu2020-20502 ◽

2020 ◽

Author(s):

Maryse Charpentier-Noyer ◽

François Bourgin ◽

Geoffroy Kirstetter ◽

Olivier Delestre ◽

Pierre Brigode

Keyword(s):

Water Level ◽

Real Time ◽

Flash Floods ◽

Rainfall Runoff ◽

Calculation Time ◽

Time Simulation ◽

Hydraulic Coupling ◽

Rainfall Runoff Model ◽

Runoff Model ◽

French Riviera

The vulnerability of the French Riviera to hydro meteorological hazards has been dramatically illustrated by the flash floods of October 3, 2015: 20 people were killed and the cost of the direct damages were higher than 600 million euros. Due to their fast dynamics, flash floods are difficult to predict and leave little time for forecasting. In this context, it is needed to improve real-time simulations to enable a short-range anticipation of the consequences of these phenomena. The main goal of this work was to test a hydrologic-hydraulic coupling in order to assess whether this coupling can be used for real-time forecasting purposes. The coupling is composed for the hydrological part of the event-based spatially distributed rainfall-runoff model Cinecar and for the hydraulic part of the Basilisk software, which is based on 2D hydraulic modelling (finite volume methods for shallow water equations) with adaptive grid refinement. The main interest of this coupling method is the compromise obtained between calculation time and precision. The rainfall-runoff model is run on the upstream part of the domain and feeds the hydraulic model applied in the downstream part. The rainfall-runoff model makes it possible to estimate very quickly the streamflow temporal evolution, while the hydraulic model, although much slower when applied at high spatial resolution (up to 4m), makes it possible to have water level and velocity at any point of the downstream area. The application of this coupling approach is presented for three basins severely affected by the October 2015 flash floods: the Brague (68 km&#178;), the Fray&#232;re (22 km&#178;) and the Riou de l&#8217;Argenti&#232;re (48 km&#178;) catchments. The results obtained for the three basins are compared with information gathered from post-event surveys, particularly the high water level marks. A particular attention is also put on computation times in order to evaluate the possibilities of real-time simulation. The results show promising performances both in terms of calculation time but also in terms of accuracy of the simulated flood areas and water levels.

Download Full-text

The real-time stochastic flow forecast

Soil and Water Research ◽

10.17221/13/2009-swr ◽

2010 ◽

Vol 5 (No. 2) ◽

pp. 49-57 ◽

Cited By ~ 4

Author(s):

L. Březková ◽

M. Starý ◽

P. Doležal

Keyword(s):

Lead Time ◽

Meteorological Data ◽

Rainfall Runoff ◽

Hydrological Simulation ◽

The Czech Republic ◽

Precipitation Prediction ◽

The Monte Carlo Method ◽

Rainfall Runoff Model ◽

River Profiles ◽

Runoff Model

In the Czech Republic, deterministic flow forecasts with the lead time of 48 hours, calculated by rainfall-runoff models for basins of a size of several hundreds to thousands square kilometers, are nowadays a common part of the operational hydrological service. The Czech Hydrometeorological Institute (CHMI) issues daily the discharge forecast for more than one hundred river profiles. However, the causal rainfall is a random process more than a deterministic one, therefore the deterministic discharge forecast based on one precipitation prediction is a significant simplification of the reality. Since important decisions must be done during the floods, it is necessary to take into account the indeterminity of the input meteorological data and to express the uncertainty of the resulting discharge forecast. In the paper, a solution of this problem is proposed. The time series of the input precipitation prediction data have been generated repeatedly (by the Monte Carlo method) and, subsequently, the set of discharge forecasts based on the repeated hydrological model simulations has been obtained and statistically evaluated. The resulting output can be, for example, the range of predicted peak discharges, the peak discharge exceeding curve or the outflow volume exceeding curve. The properties of the proposed generator have been tested with acceptable results on several flood events which occurred over the last years in the upper part of the Dyje catchment (Podhradí closing profile). The rainfall-runoff model HYDROG, which has been in operation in CHMI since 2003, was used for hydrological simulation.

Download Full-text

A Study on Parameters of Rainfall Runoff Model and The Prediction Method of Landslide Disaster in Mountainous Area

10.5194/egusphere-egu2020-18893 ◽

2020 ◽

Author(s):

Huimyeong yoo ◽

Naoki koyama ◽

Tadashi yamada

Keyword(s):

Prediction Method ◽

Heavy Rain ◽

Mountainous Area ◽

Rainfall Runoff ◽

Mountainous Areas ◽

Average Amount ◽

Tank Model ◽

Rainfall Runoff Model ◽

Runoff Model ◽

Landslide Disaster

This study is to analyze the evacuation behavior of residents living in the mountainous area and predict landslide disasters during heavy rain. 70% of Japan has are mountainous areas, and landslide disasters have occurred due to heavy rains caused by typhoons and heavy rainfall, etc. the annual average amount of damage caused by landslide disasters is 1000 in recent years. Also, landslide disaster warning information and evacuation information are important, it is difficult to predict landslide disasters, however, if we issued the evacuation advisory when the disasters already happened, there will be not enough time for the evacuation. In order to protect residents from such disasters, it is important to clarify "what information is effective for evacuation" and "when should those information be released?" Therefore, we conducted a survey on the residents in the mountainous areas which suffered from the heavy rain disaster in 2017 and analyzed the answers.As a result, some residents evacuated before the evacuation information was issued. Because some landslide disasters occurred even before the first evacuation information was transmitted, and they felt danger. This result shows that the early information based on the prediction of the disasters is important in mountainous areas.Therefore, we suggested a method for predicting landslide disasters, the method uses a rainfall and runoff tank model with high reproducibility and robustness of geological characteristics and uses the cumulative rainfall at the time of disaster occurrence as an index. As a result, this model predicted the occurrence of the landslide disaster 3 hours earlier by using forecasted rainfall. it is an effective method.

Download Full-text