Astronomy began with observations of things visible in the sky. In the Modern Era, observations of radio waves as well as visible light further advanced human knowledge. In solar astronomy we observe images, events such as solar flares, and magnetic fields active on the surface of the sun. The primary reasons for observing the sun are to better understand the workings of the nuclear furnace that provides all of our energy, and also to forecast events that could affect us.

Two primary forms of solar observation are still images and time series. An image of the solar surface shows artifacts generated by solar activity, such as sunspots. A solar flare, for example, occurs over a time duration; it is thus represented by a time series of data captures.

A challenge for astronomers is to pick out a relatively small number of meaningful events from a massive volume of input data. Various algorithms have been developed during the Information Age to analyze the input data and pick out events of interest. The present book explores the use of artificial intelligence (AI), specifically machine learning systems, to see if these systems can improve the efficiency and accuracy of data analysis.

Two basic types of AI systems are used: machine learning and adversarial generation. The former are pattern recognition systems; after being trained on a set of patterns that are to be recognized, the system is then exposed to novel patterns that are classified according to the training data. The latter, which are based on machine learning, take an input image and then generate an output image that resembles the input. The generated images are then compared with the originals to enhance the image or increase the accuracy of its classification.

The book consists of six chapters. The first discusses the development of AI from its inception. It goes on to describe deep learning and several types of systems. Of relevance here are convolutional neural networks (CNNs), autoencoders, generative adversarial networks (GANs), and long short-term memory (LSTM) networks. Subsequent chapters describe how these tools are used to classify solar images, detect solar events, and process images in order to improve feature definition, recover from overexposure, generate magnetic field images from ultraviolet radiation, and forecast future events such as solar flares. In each case, the authors describe the type of AI applicable to the process and how they employed the tool, and compare the result with traditional older tools. Where relevant they discuss the mathematics used to optimize processing.

At just around 100 pages this work provides an overview of tools and methods versus an in-depth treatise. Each application is described with sufficient detail to give the reader an understanding of how AI is used and how its use compares with older tools used for the same purposes. The writing is clear, although there are occasional typographical errors, for example, “starts” when “stars” was clearly intended, and an excessive use of acronyms. Relevant images and tables enhance the reader’s understanding; many references accompany each chapter. This book should appeal to those interested in either AI or the field of solar astronomy.