This is a nice booklet on well-selected aspects of spatial statistics, which can be recommended as a textbook for a course for graduate students and as an introduction for researchers. The reviewer is impressed how fast and direct the author comes to the point, avoiding long boring discussions. Perhaps, a price for this is that Bayesian statistics does not play a role.
The Introduction presents main notions and the fundamentals of time series theory. Chapter 2 gives the fundamentals of geostatistics in an excellent and concise manner, with outlooks to modern ideas. Chapter 3 discusses variograms and covariance functions, models and statistics. Chapter 4 then turns to spatial autoregressive models with discrete and continuous variables.
A speciality of the book is an extensive study of directional behaviour. Chapter 5 discusses anisotropy in random fields and Chapter 8 the same for point processes.
Chapter 6 considers space-time data in the context of regionalized variables. Chapter 7 is a very short but nevertheless useful introduction to point process statistics. Indeed, for understanding the main ideas it is sufficient to consider crude estimators. Chapter 9 generalizes to the case of multivariate data, where at every location/instant a series of variables is measured.
Many readers will like the last chapter on resampling for correlated observations, a frequently needed technique. And it is fine that the book recalls some central limit theorems that are often omitted in textbooks of that level.
The book is well written and well illustrated by examples.
In some minor points the book could be improved. Some variables and symbols are not sufficiently explained. The role of the Ising model is perhaps overestimated. It is rarely used as a direct model, but it is valuable as a contextuality prior in Bayesian statistics even for clustered data. It is not nice to read ''paired correlation function'' instead of ''pair-correlation function''. And the reviewer has doubts that tests for isotropy are really needed, since always only aspects of isotropy can be estimated (as it is also with stationarity). It may be sufficient for practical statistics to have diagnostic tools, which appear already in earlier books.
All in all, this book can be highly recommended.