The book looks more like an encyclopedia than a handbook or textbook. The chapter titles mention climate issues that are typically discussed in the last subsections of the given chapter. Each chapter systematically goes through the formal computer science and mathematical foundations. This material is not for the faint-hearted reader. The descriptions of methods, techniques, and algorithms require a thorough grounding in mathematics. The Internet of Things (IoT) and sensors play a role only in the subsections dealing with climate issues. The source data collection is data that has been collected from all over the world for decades. The accurate data, from 1950, could be used for analysis.

The focus of each chapter is to apply graph theories--in the book, network theories--to climate issues, that is, to yield predictions about climate change. Chapter 1 introduces issues related to climate data analytics: how big data analytics is used, how a large number of dimensions can be handled through downscaling, and how the optimal estimate can be created for state variables. The chapter concludes by describing how cloud computing is used for spatiotemporal big data related to climate analysis.

Chapter 2 discusses various methods for feature extraction. Climate data contains both qualitative and quantitative data. In the first cut, disparate regression methods are apt for qualitative data. Data clustering fits quantitative data. The chapter closes with techniques for rainfall, flood, and crop estimation.

Chapter 3 overviews the fashionable methods categorized as deep learning, namely the various versions of artificial neural networks. Downscaling means transforming the global climate models into local meteorological variables. The neural network approach showcases better performance than traditional multiple regression models.

Chapter 4, “Climate Networks,” looks at the representation of climate models in directed and undirected graphs. The chapter shows the foundation of network (graph) analysis through linear algebraic and numeric calculations, as well as the use of random walks and similarity measures based on probability theory. Utilization of the climate network model is demonstrated on the El Niño and North Atlantic oscillations.

Chapter 5 depicts the mathematics of random networks and entropy. It requires broad mathematical knowledge of information theory, probability theory, and the application of graphs that includes small-world and Barabási-Albert networks. Chapter 6, “Spectra of Climate Networks,” lays out the spectral theories of matrices and graphs--which are intimately interconnected in this viewpoint. Chapter 7 outlines the Monte Carlo methods and the subtleties of their various approaches, for example, the sampling methods. Chapter 8, “Sparse Representation of Big Climate Data,” is very important from an information technology (IT) point of view, covering storage, processing, and dissemination. A crucial question is how to compress the data without information loss. There are appropriate methods where the loss of information is not significant, so compressed information is an important subject in big data analytics.

Chapters 9, 10, and 11 discuss the reduction of carbon emission, how big data can be used for low carbon management, and arctic maritime transportation, respectively. The changing climate makes it possible to use the Northwest Passage (NWP) and Northeast Passage (NEP) in the 21st century. The utilization of these arctic maritime routes based on big data analytics will exploit various data sources as economic data; IoT data acquisition; measurement, climate data, and climate computer models; and so on.

Although the authors concentrate on climate issues and apply significant mathematics and computer science know-how to this domain, the content of the book can be used in other fields, too. How specific methods and techniques are applied to climate modeling is interesting in itself. Because of the included mathematical approaches and real-world examples, it is a useful book for researchers and practitioners involved in modeling complex systems.