Predicting flood susceptibility using LSTM neural networks

Zhice Fang,Yi Wang,Ling Peng,Haoyuan Hong

DOI: https://doi.org/10.1016/j.jhydrol.2020.125734

IF: 6.4

2021-03-01

Journal of Hydrology

Abstract:Identifying floods and producing flood susceptibility maps are crucial steps for decision-makers to prevent and manage disasters. Plenty of studies have used machine learning models to produce reliable susceptibility maps. Nevertheless, most research ignores the importance of developing appropriate feature engineering methods. In this study, we propose a local spatial sequential long short-term memory neural network (LSS-LSTM) for flood susceptibility prediction in Shangyou County, China. The three main contributions of this study are summarized below. First of all, it is a new perspective to use the deep learning technique of LSTM for flood susceptibility prediction. Second, we integrate an appropriate feature engineering method with LSTM to predict flood susceptibility. Third, we implement two optimization techniques of data augmentation and batch normalization to further improve the performance of the proposed method. The LSS-LSTM method can not only capture the attribution information of flood conditioning factors and the local spatial information of flood data, but also has powerful sequential modelling capabilities to deal with the spatial relationship of floods. The experimental results demonstrate that the LSS-LSTM method achieves satisfactory prediction performance (93.75% and 0.965) in terms of accuracy and area under the receiver operating characteristic (ROC) curve.

geosciences, multidisciplinary,water resources,engineering, civil

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to explore the transcriptional regulatory mechanism of the ABCC6 gene and its tissue - specific expression pattern. Specifically, the researchers cloned the 2.6 - kb human ABCC6 promoter region and developed a series of 5' - deletion constructs linked to the luciferase reporter gene to identify the key regulatory elements that affect the expression level of this gene in different cell lines. In addition, they also investigated the effects of specific cytokines (such as transforming growth factor β (TGF - β), tumor necrosis factor α (TNF - α) and interferon γ (IFN - γ)) on ABCC6 promoter activity, as well as the role of Sp1/Sp3 transcription factors in this process.

### Summary of main research questions:
1. **Determine the tissue - specific expression regulatory mechanism of the ABCC6 gene**: The researchers aim to identify and characterize the cis - acting elements responsible for the high - expression of the ABCC6 gene in specific tissues such as the liver.

2. **Evaluate the effects of inflammation - related cytokines on ABCC6 gene expression**: Verify experimentally whether cytokines such as TGF - β, TNF - α and IFN - γ can regulate the activity of the ABCC6 promoter and explore the underlying molecular mechanisms.

3. **Analyze the functions of Sp1/Sp3 transcription factors in ABCC6 gene expression**: The researchers pay special attention to the role of Sp1/Sp3 binding sites to understand how these factors participate in the transcriptional regulation of the ABCC6 gene.

4. **Verify the function of the ABCC6 promoter in vivo**: Use the tail vein injection technique to introduce plasmids containing the ABCC6 promoter into mice to further confirm its performance in a living environment.

Through the study of the above problems, the authors hope to reveal the specific mechanism of ABCC6 gene expression regulation, which will not only help to understand the pathogenesis of hereditary diseases such as pseudoxanthoma elasticum (PXE), but may also provide new therapeutic targets for other diseases involving ABCC6 gene abnormalities.