Progress in Physical Geography 29, 1 (2005) pp. 129-130

Book reviews

Liang, S. 2004: Quantitative remote sensing of land surfaces. Hoboken
NJ: Wiley- Interscience. xxvi + 534 pp. + CD-ROM. 82.50 cloth. ISBN 0
471 28166 2.

By Paul M. Mather, The University of Nottingham

Shunlin Liang is an Associate Professor of Geography at the University
of Maryland. His research interests centre on the topic of radiative
transfer modelling, and this book represents his largely successful
attempt to describe the development and use of physically based models
to recover estimates of the values of Earth surface variables from
remotely sensed data. Reliable estimates of these variables, which are
used for the purposes of environmental prediction, can only be
achieved, the author believes, through the use of physical dynamic
models. He therefore focuses on two main themes  the theory underlying
the models and the algorithms required to write operational computer
code. He does not cover the use of remotely sensed images to generate
thematic maps of, for example, land cover; rather, he explains how
remotely sensed data can be calibrated to physical values (radiance),
how the effects of the atmosphere can be removed from the signal
received at the sensor, and how these calibrated and atmospheric

The first chapter is a concise and well illustrated introduction to
quantitative remote sensing, concentrating on the definition of
radiometric variables, and on the components of a remote sensing
modelling system (e.g., atmospheric radiative transfer models, sensor
models). Chapters 2  4 provide in-depth treatments of atmospheric
short-wave radiative transfer modelling canopy reflectance modelling,
and modelling snow and soil reflectance. Chapter 5 then deals with the
radiometric calibration of sensors, with specific details of Landsat TM
and NOAA AVHRR calibration procedures. A review of atmospheric
correction methods (Chapter 6) is followed by a detailed review of
topographic correction procedures (Chapter 7). Next, the estimation of
land surface biophysical variables from multispectral and hyperspectral
data sets is considered in some detail (Chapter 8). Chapters 9 and 10
cover the topic of estimation of the surface radiation budget, with
Chapters 8 and 9 covering broadband and longwave albedo estimation

A CD accompanies the book. Its contents are divided into two parts,
data and software, each of which is arranged in chapter order. Data
sets include extra-terrestrial solar irradiance derived using MODTRAN4,
and a set of sensor spectral response functions for Landsat TM and ETM,
AVIRIS and other sensors. The software provided on the CD is mainly
written in Fortran, with some C code. Source code is provided for a
number of atmospheric and canopy reflectance models, including DISORT
and GeoSAIL. The level of physics and mathematics required to
understand and appreciate the content of this book  described as a good
working knowledge of statistics, calculus and linear algebra on an
undergraduate level  is well above that achieved by the majority of
geographers, and it is therefore likely that the main readership of the
book will consist of postgraduate-level environmental physicists. It
will find its main use as a reference book rather than as a course
text. There are a number of unimportant typographic errors, plus some 
mistakes that should have been picked up at the proofreading stage (for
example, the use of the term generic algorithm rather than genetic
algorithm). Such quibbles apart, this is a well-researched book that the 
environmental modeling community will find an indispensable reference.