Russell G. Congalton, Kass Green, Assessing the Accuracy of Remotely Sensed Data - Principles and Practices Second edition (2009) CRC Press, Taylor & Francis Group, Boca Raton, FL 978-1-4200-5512-2 183 pp., Price: $99.95

Assessing the Accuracy of Remotely Sensed Data is the second edition of a very successful handbook. The first edition of this book was published in 1999. The original 8 chapters have been expanded to 11. New in the second edition is the treatment of positional accuracy of a map in addition to thematic accuracy. Moreover, the use of fuzzy accuracy assessment and the accuracy of a change map have been extended significantly in comparison to the first edition. Chapter 1 provides an introduction to the theory and principles of quantitative accuracy assessment of maps. Both the positional and thematic accuracy assessments are introduced, emphasizing the critical role of the planning of the sampling to be done for the accuracy assessment. Chapter 2 starts with a review of the history of map making and the role of accuracy assessment, ending with the introduction of the error matrix together with the overall, producer's and user's accuracy. Chapter 3 introduces the positional accuracy assessment further. It describes causes of positional error, describes common US standards for positional accuracy and basic statistics. Then sample design and collection are described and, finally, the analysis for estimating positional accuracy is described for various standards. Strong point of this chapter is to bring descriptions of various standards back to one language (terminology). Difference between precision and accuracy is emphasized. The accepted measure of positional accuracy is the root-mean-square error (RMSE). Both the proper way for determining both vertical (1D) and horizontal (2D) positional accuracy are treated in detail, describing the correct statistics with their confidence interval to be applied. An important point made is that positional accuracy of a map or data set should be assessed for an independent set of sample points, not for those used in a registration process. Chapter 4 introduces the thematic accuracy assessment, the major focus of this book. This is a short chapter just introducing the error matrix as a measure for representing thematic accuracy, including errors of omission and commission, and the producer's and user's accuracy. Since the design of samples is critical for assessing the thematic accuracy, chapter 5 discusses the thematic map classes to be assessed, their definition and their distribution, the appropriate sample unit, the number of samples required and how they should be selected. Potential pitfalls are discussed. Special attention is paid to taking care of spatial autocorrelation, stressing that a polygon of pixels should be …