Tropical Cyclone Intensity Estimation Using a Deep Convolutional Neural Network

Tropical cyclone intensity estimation is a challenging task as it required domain knowledge while extracting features, significant pre-processing, various sets of parameters obtained from satellites, and human intervention for analysis. The inconsistency of results, significant pre-processing of data, complexity of the problem domain, and problems on generalizability are some of the issues related to intensity estimation. In this study, we design a deep convolutional neural network architecture for categorizing hurricanes based on intensity using graphics processing unit. Our model has achieved better accuracy and lower root-mean-square error by just using satellite images than 'state-of-the-art' techniques. Visualizations of learned features at various layers and their deconvolutions are also presented for understanding the learning process.

I. Introduction

Deep learning uses a deep architecture of multiple processing layers composed of linear or nonlinear transformations [1]–[6] while replacing handcrafted features with automated feature learning and hierarchical feature extraction [7]. Convolutional Neural Networks (CNNs) can be used to model spatial correlation with translation invariance making them suitable for image recognition [8], [9]. This study proposes a deep CNN architecture for estimating the hurricane intensity by learning features.¹

We use tropical cyclone, TC, cyclone, and hurricane interchangeably in this paper.