Stochastic Modeling for Estimating Real-Time Inundation Depths at Roadside IoT Sensors Using the ANN-Derived Model

Abstract

This paper aims to develop a stochastic model (SM_EID_IOT) for estimating the inundation depths and associated 95% confidence intervals at the specific locations of the roadside water-level gauges, i.e., Internet of Things (IoT) sensors under the observed water levels/rainfalls and the precipitation forecasts given. The proposed SM_EID_IOT model is an ANN-derived one, a modified artificial neural network model (i.e., the ANN_GA-SA_MTF) in which the associated ANN weights are calibrated via a modified genetic algorithm with a variety of transfer functions considered. To enhance the reliability and accuracy of the proposed SM_EID_IOT model in the estimations of the inundation depths at the IoT sensors, a great number of the rainfall induced flood events as the training and validation datasets are simulated by the 2D hydraulic dynamic (SOBEK) model with the simulated rain fields via the stochastic generation model for the short-term gridded rainstorms. According to the results of model demonstration, Nankon catchment, located in northern Taiwan, the proposed SM_EID_IOT model can estimate the inundation depths at the various lead times with high reliability in capturing the validation datasets. Moreover, through the integrated real-time error correction method integrated with the proposed SM_EID_IOT model, the resulting corrected inundation-depth estimates exhibit a good agreement with the validated ones in time under an acceptable bias.

Modeling rainfall-induced 2D inundation simulation based on the ANN-derived models with precipitation and water-level measurements at roadside IoT sensors

“…Ultimately, the model validation could compare the gridded inundation depths and the corresponding flood extents estimated by the proposed SM_EID_2D model with those from the training datasets comprised of the simulated rainfall-induced flood events.

Figure 1 Schematic process of characterizing and generating event-based rainstorms 13 , 36 , 37 .

…”

“…However, the real-time practical flood-induced runoff and inundation depths, especially in urban areas, might be challenging to measure due to the limitation of measurement equipment, hindrances in data acquisition, accuracy in the parameters of the numerical model for processing and analysis, and spatial uncertainties in the digital elevation map. The above deprivation probably causes uncertainties in the delineated flooding zones and area 4 , 8 – 13 . Nevertheless, enhancing the computation power and capability, a group of numerical simulation models based on the rainfall-runoff analysis and flood dynamics routing are comprehensively applied in the 1D/2D inundation/flood simulation 9 – 11 .…”

“…Generally speaking, by employing the hydrodynamic numerical models, the at-site inundation depths and induced potential flooding zones, as well as the associated area, can be estimated under consideration of the various types of rainfalls, such as the design rainfall events of the different return periods and the precipitation forecasts as well as observations 13 – 18 ; for example, Ming et al 16 developed a 2D hydrodynamic simulation module to perform real-time flooding forecast with the precipitation forecasts provided via the weather numerical models. Although the hydrodynamic numerical models can reproduce the inundation depths at the specific locations and induced flooding areas in the catchment and urban spaces, their simulation performance and accuracy significantly positively rely on the resolutions of the resulting flood-related variates in time and space 10 , 19 .…”

Modeling the 2D Inundation Simulation Based on the ANN-Derived Model with Real-Time Measurements at Roadside IoT Sensors

“…When training the ANN-derived models, the initial conditions, including the number of hidden layers, the total number of neurons, and transfer functions of interest (see Table 2), should be given in advance. It is well-known that the 3-layer network structure: one input layer, one output layer, and one hidden layer, is frequently utilized in hydrological/hydraulic modeling [1,14]; thus, the proposed SM_EID_VIOT model is developed using the 3-layer ANN_GA-SA_MTF model. Additionally, the number of neurons can be estimated via the numerous formulae (see Table 1) with the number of model inputs and outputs.…”

“…With computation power increasing, the CNN-based and ANN-derived models are widely applied to the relevant hydrological/hydraulic analysis, especially in flood-related simulations/forecasts [1,3,4,[9][10][11][12][13][14]. In detail, the CNN-based model is derived via the neutral network, comprising the convolution, pooling, and fully connected layers, requiring extensive 2D data (i.e., gridded data) as datasets (e.g., the image and videos) for the model training and application [3,4,15]; accordingly, the CNN model can efficiently provide the single model output from the grid-format model inputs.…”