Developing flood damage estimation model for calculating flood damage reduction due to weather control
In this research item, we develop a methodology for estimating the effect of weather control on reducing economic damage. Specifically, we are; 1) developing a flood and inundation simulation model for the entire Japan for calculating the inundation area and its depth in cases with and without weather control; 2) developing an exposure asset database based on various statistical data and flood economic damage estimation model based on inundation depth and exposure information. By utilizing the developed flood and inundation simulation model and the flood economic damage estimation model, we aim to quantify the effect of weather control on reducing economic damage.
Item 8-1Quantifying flood risk
Principal investigator: Masafumi Yamada
Outline
This study focuses on quantifying the changes in flood risk associated with weather control through the reproduction analysis of heavy rain disasters. We develop a sluice gate model capable of handling data for all of Japan, which will estimate inundation depth using meteorological information such as precipitation as input. We start with well-known flooding caused by overflow and gradually expand our focus to include more complex disasters, such as dike break and other events that require probabilistic prediction.
Methods
We establish a flood inundation depth model for large areas based on the Rainfall-Runoff-Inundation model, a distributed sluice gate model capable of integrated analysis while maintaining consistency among variables related to rainfall, runoff, and inundation. In this process, we apply a parameter estimation method developed by the researches to avoid combinatorial explosion problem, optimizing parameter settings based on regional runoff characteristics. Additionally, we comprehensively incorporate models of artificial structures such as sluice, which have been previously studied at the watershed scale, but this study extends to cover all of Japan. Furthermore, to verify the reliability of flood risk assessments, we conduct accuracy evaluations based on reproduction computation of past large-scale inundation events.
Importance
To quantify the effects of heavy rain disaster mitigation that this project aims to achieve, it is essential to establish a reliable method for estimating economic damage from floods. The development of a model to determine the spatial distribution of flood inundation depths using weather predictive information across all of Japan is central of estimating economic damages. Accurate runoff inundation analysis requires the appropriate reflection of regional runoff characteristics in model parameters, as well as comprehensive model of artificial structures like dike break and sluice gates that influence flooding patterns. We focus on quantifying flood risk through these elements.
Expected problems and solutions
Challenges in reflecting regional runoff characteristics in model parameters include combinatorial explosion due to high-dimensional search spaces during large-area optimization and the variability in optimal parameters caused by heavy rainfall events. We address these issues by applying a likelihood-based parameter estimation method developed by the researchers, which is robust and operates in polynomial time for multiple events, across all of Japan. Additionally, when introducing models for artificial structures, a challenge is the creation of a wide-ranging and homogeneous database. This study addresses this by constructing a consistent national database using various open data sources.
Members
Item 8-2Developing flood damage estimation model
Principal investigator: Shinji Yamada
Outline
This study develops a damage estimation model capable of estimating flood damage across all of Japan based on inundation depth data. By combining the inundation depth data computed by the sluice gate model with the damage estimation model, we are able to estimate flood damage costs across all of Japan. The inundation depth data will be based on the outcomes of Item 8-1. We begin by focusing on general damages, such as those to homes and buildings, which have the highest economic impact among flood damages in Japan, and subsequently expand the estimation to include public and community damages.
Methods
We develop an economic damage estimation model for flood disasters across Japan, based on the analytical methods of the CAT (Catastrophe) model, which is commonly used in the insurance industry for quantitative assessment of natural disaster risks.
We focus on building an exposure asset database and developing flood damage functions to estimate the economic damages based on inundation depth data computed by the sluice gate model. The inundation depth data will be sourced from rainfall-runoff-inundation model developed in Item 8-1, which enables the estimation of inundation depth across all of Japan.
Importance
The economic damage estimation model developed in this study will quantitatively assess the economic losses caused by flood disasters across all of Japan. This capability is essential for demonstrating the damage reduction effects of the project’s goal to mitigate heavy rain damage through maritime rain generation and reduction of precipitation on land.
Expected problems and solutions
It is well known that the flood disasters analyzed in this study exhibit significant local variability in terms of both the occurrence and severity of damage. Therefore, if the spatial resolution of the exposure asset database is poor, there is a risk of increased uncertainty in the estimation results or (a significant divergence from actual conditions). In order to enhance the spatial resolution of the exposure asset database as much as possible, we utilize not only publicly available statistical data, such as census and building stock statistic, but also, when necessary, more spatially detailed paid data to advance the construction of the database.
Members
Item 8-3Evaluating landslide risk and addressing neighboring countries
Principal investigator: So Kazama
Outline
In order to address concerns in surrounding areas related to weather control, this study focuses on the impacts of weather changes induced by interventions on disaster risks in neighboring countries. By developing a landslide risk estimation model based on latest findings, we assess how weather interventions influence landslide risks in these areas. We estimate the hazards of landslide in Japan associated with changes in extreme rainfall using a hybrid model that combines occurrence probability models based on historical data with mechanistic models. Additionally, we quantitatively assess the risks of human due to heavy rainfall using a hybrid model that combines the probability of occurrence model, which is based on past experience, and a mechanical model. At the same time, we will quantitatively evaluate the human casualties and economic losses. Once the methods are established, we apply them to the East Asia or Southeast Asia regions and evaluate their impacts.
Methods
The SLIP (Shallow Landslides Instability Prediction) model, widely used internationally for the stability analysis of infinite slopes, is compared with a logistic function-based landslides hazard model that has demonstrated success domestically. This study aims to understand the difference between the latest mechanical models and probabilistic models in terms of parameters and identification methods. By providing rainfall as a probabilistic precipitation, we can derive the expected safety factor from the mechanical model and the expected annual occurrence probability from the probabilistic model in the context of slope hazard assessment. By integrating these models in the flow direction, we can develop a highly accurate landslide hazard estimation model. The combination methods and parameter settings will be determined based on domestic performance data, allowing us to select the most optimal approach. Subsequently, we conduct numerical experiments on weather control across a wide area, including neighboring countries. By applying developing landslide risk estimation model, we qualitatively assess the impact of these interventions on landslide risk (including human losses and damage costs) in the surrounding countries.
Importance
For the successful implementation of weather interventions, it is essential to address concerns from the regions affected and neighboring countries. Particularly, the key is the evaluation of how weather interventions impact disaster risk in neighboring countries. Additionally, if the application of weather intervention technologies developed thorough this study can quantitatively demonstrate a reduction in disaster risk in neighboring countries, it is believed that this could help relieve psychological resistance from those countries towards the project.
Implementing weather intervention experiments in real-world settings requires substantial costs. If we can express the mitigating effects of landslide disaster in terms of damage cost, it will enable cost-benefit analysis, allowing us to quantify the effectiveness of the interventions.
The selection of target countries will depend on the extent of the impact that the proposed weather interventions have on the surrounding weather. Therefore, it will not be clearly defined at the start of the research. We accumulate knowledge regarding the impacts of weather intervention operations on surrounding weather, taking into account the results from Item 5-1 and 5-3. Based on this information, we aim to select the target countries and regions for landslide risk assessment by the end of FY 2024. In order to achieve the goal of this project, it is essential to fully evaluate the effectiveness of an accurate landslide hazard and risk models that uses changes in heavy rainfall as input.
Expected problems and solutions
The combined model of mechanical and probabilistic approaches is unprecedented, representing a research and development initiative that integrates scientific advancements with practical applicability. While there are numerous performance data for estimating hazards and risks in Japan when weather interventions are implemented, it is known that collecting high-quality data in other countries, particularly in developing nations, poses significant challenges. The Asian Institute of Technology has data on landslide disaster areas obtained through remote sensing, and we plan to utilize this data for validation purposes. In addition, we have several contacts in Taiwan and Vietnam, which should facilitate access to local data.
Members
