Abstract:Coastal regions in North America face major threats from storm surges caused by hurricanes and nor'easters. Traditional numerical models, while accurate, are computationally expensive, limiting their practicality for real-time predictions. Recently, deep learning techniques have been developed for efficient simulation of time-dependent storm surge. To resolve the small scales of storm surge in both time and space over a long duration and a large area, these simulations typically need to employ oversized neural networks that struggle with the accumulation of prediction errors over successive time steps. To address these challenges, this study introduces a hierarchical deep neural network (HDNN) combined with a convolutional autoencoder (CAE) to accurately and efficiently predict storm surge time series. The CAE reduces the dimensionality of storm surge data, streamlining the learning process. HDNNs then map storm parameters to the low-dimensional representation of storm surge, allowing for sequential predictions across different time scales. Specifically, the current-level neural network is utilized to predict future states with a relatively large time step, which are passed as inputs to the next-level neural network for smaller time-step predictions. This process continues sequentially for all time steps. The results from different-level neural networks across various time steps are then stacked to acquire the entire time series of storm surge. The simulated low-dimensional representations are finally decoded back into storm surge time series. The proposed model was trained and tested using synthetic data from the North Atlantic Comprehensive Coastal Study. Results demonstrate its excellent performance to effectively handle high-dimensional surge data while mitigating the accumulation of prediction errors over time, making it a promising tool for advancing storm surge prediction.

Prediction of hierarchical time series using structured regularization and its application to artificial neural networks

A machine learning approach for forecasting hierarchical time series

Hierarchically Regularized Deep Forecasting

Enhancing Time Series Forecasting: A Hierarchical Transformer with Probabilistic Decomposition Representation

Forecasting Hierarchical Time Series

End-to-End Modeling Hierarchical Time Series Using Autoregressive Transformer and Conditional Normalizing Flow based Reconciliation

SLOTH: Structured Learning and Task-Based Optimization for Time Series Forecasting on Hierarchies

Hierarchical Time Series Forecasting with Bayesian Modeling

Design of Hierarchical Neural Networks Using Deep LSTM and Self-organizing Dynamical Fuzzy-Neural Network Architecture

A Bayesian Model for Forecasting Hierarchically Structured Time Series

N-HiTS: Neural Hierarchical Interpolation for Time Series Forecasting

Hierarchical forecasting with a top-down alignment of independent level forecasts

Efficient Forecasting of Large Scale Hierarchical Time Series via Multilevel Clustering

Graph-based Time Series Clustering for End-to-End Hierarchical Forecasting

Hierarchical Classification Auxiliary Network for Time Series Forecasting

Advancing Spatio-temporal Storm Surge Prediction with Hierarchical Deep Neural Networks

DeepHGNN: Study of Graph Neural Network based Forecasting Methods for Hierarchically Related Multivariate Time Series

Toward a Digital Twin: Time Series Prediction Based on a Hybrid Ensemble Empirical Mode Decomposition and BO-LSTM Neural Networks

Neural Predicting Higher-order Patterns in Temporal Networks

TPRNN: A Top-Down Pyramidal Recurrent Neural Network for Time Series Forecasting

Hierarchical learning, forecasting coherent spatio-temporal individual and aggregated building loads