Spatial-Temporal Data Mining for Ocean Science: Data, Methodologies, and Opportunities

Hanchen Yang,Wengen Li,Shuyu Wang,Hui Li,Jihong Guan,Shuigeng Zhou,Jiannong Cao

2023-08-03

Abstract:With the rapid amassing of spatial-temporal (ST) ocean data, many spatial-temporal data mining (STDM) studies have been conducted to address various oceanic issues, including climate forecasting and disaster warning. Compared with typical ST data (e.g., traffic data), ST ocean data is more complicated but with unique characteristics, e.g., diverse regionality and high sparsity. These characteristics make it difficult to design and train STDM models on ST ocean data. To the best of our knowledge, a comprehensive survey of existing studies remains missing in the literature, which hinders not only computer scientists from identifying the research issues in ocean data mining but also ocean scientists to apply advanced STDM techniques. In this paper, we provide a comprehensive survey of existing STDM studies for ocean science. Concretely, we first review the widely-used ST ocean datasets and highlight their unique characteristics. Then, typical ST ocean data quality enhancement techniques are explored. Next, we classify existing STDM studies in ocean science into four types of tasks, i.e., prediction, event detection, pattern mining, and anomaly detection, and elaborate on the techniques for these tasks. Finally, promising research opportunities are discussed. This survey can help scientists from both computer science and ocean science better understand the fundamental concepts, key techniques, and open challenges of STDM for ocean science.

Machine Learning,Atmospheric and Oceanic Physics

What problem does this paper attempt to address?

This paper provides a comprehensive review of the issues related to Spatial-Temporal Data Mining (STDM) in marine science. Specifically, the paper aims to address the following key issues:

1. **Analysis of the Unique Characteristics of Marine Spatial-Temporal Data**: The paper points out that marine spatial-temporal data is more complex compared to typical spatial-temporal data such as traffic data. It has characteristics such as diverse regionality, high sparsity, inherent uncertainty, and deep spatial-temporal dependencies. These characteristics make it very difficult to design and train STDM models on marine data.

2. **Lack of Systematic Review of Existing Research**: Although there has been a lot of research on STDM in marine science, there is a lack of a comprehensive literature review to help computer scientists identify research issues in marine data analysis, and it also fails to help marine scientists apply advanced STDM techniques. Therefore, the paper aims to fill this gap.

3. **Categorical Summary of STDM Methods and Techniques in Marine Science**: The paper categorizes existing STDM research into four major tasks—prediction, event detection, pattern mining, and anomaly detection—and provides a detailed introduction and discussion of the techniques under each task.

4. **Discussion of Future Research Directions**: The paper also discusses the challenges and future research opportunities in STDM in marine science, such as developing end-to-end STDM models, integrating multi-source marine data, and improving the interpretability of deep STDM methods.

Through the above analysis, it can be seen that the main goal of this paper is to provide researchers in the fields of computer science and marine science with a comprehensive understanding framework to better understand the basic concepts, key technologies, and open challenges of STDM in marine science.