atmosphere. The physical interaction of the sunlight with the gases and particles in the
atmosphere with the land cover surface and its transmission along a different path
upward through the atmosphere affect the data recorded by the sensor in Earth’s orbit
(Adler-Golden et al., 1999; Gao et al., 2009; Song et al., 2001). On the other, the
instrumentation set-up, the local environment, and the target properties relative to the
sun’s illumination significantly interfere with the energy measured by the ground
sensor. In order to normalize the remote measurement discrepancy, these factors have
to be accounted for, suppressed, or standardized prior further land cover analysis. This
is of special importance considering the fact that many published remote sensing
studies strongly emphasize the image processing techniques while little attention is
given to the methods used for collecting field spectral data and information on the field
campaign.
The present chapter aims at discussing the contribution of field spectroradiometry
and hyperspectral remote sensing in extracting information related to land cover
mapping. The chapter starts by providing an overview on the use of field spectroradiometry in examining the spectral discrimination between different land cover
targets. Within this framework, the main factors affecting field spectroradiometric
measurements are discussed. Subsequently, the recent developments in spectral
libraries from field spectroradiometric measurements are touched upon. Following
this, an overview on the main statistical approaches employed in spectral separation of
different land cover targets is provided, linked with the most important scale factors.
In this framework, a critical review and some examples of recent related studies and
spectral libraries are furnished as well. The second part of the chapter is focused on the
use of hyperspectral remote sensing imagery from airborne or satellite platforms for
obtaining regional estimates of land use/cover. An overview of the different techniques employed in mapping land cover types from such data is first presented with
selective examples from case studies. Next, the importance of observation scale in
land cover extraction using remote sensing data is discussed. Finally, the main
conclusions are drawn and the challenges toward a more precise estimation of land
use/cover from field spectroradiometry and hyperspectral remote sensing with respect
to spectral information acquired at different spatial scales are highlighted.
15.2 FIELD SPECTRORADIOMETRY
The science of spectroradiometry began in the nineteenth century with the use of
spectrometry (the study of human vision only) and later with the interpretation of
aerial photography (Schaepman, 2007). Yet, the real scientific involvement of
spectroradiometry started with the development of airborne multispectral sensors
in the middle of the twentieth century (Short, 2009). Ever since, field spectroradiometry is an integral part of the remote sensing science since both use the sun’s
radiation as the primary light source. The key variable in spectroradiometry is the
spectral reflectance. This is emphasized as opposed to imaging spectroscopy, which
primarily analyzes the causes and exhibition of spectral absorbance from imagery and
scattering processes that occur when light strikes the target (Milton et al., 2009). The
FIELD SPECTRORADIOMETRY
287
atmosphere with the land cover surface and its transmission along a different path
upward through the atmosphere affect the data recorded by the sensor in Earth’s orbit
(Adler-Golden et al., 1999; Gao et al., 2009; Song et al., 2001). On the other, the
instrumentation set-up, the local environment, and the target properties relative to the
sun’s illumination significantly interfere with the energy measured by the ground
sensor. In order to normalize the remote measurement discrepancy, these factors have
to be accounted for, suppressed, or standardized prior further land cover analysis. This
is of special importance considering the fact that many published remote sensing
studies strongly emphasize the image processing techniques while little attention is
given to the methods used for collecting field spectral data and information on the field
campaign.
The present chapter aims at discussing the contribution of field spectroradiometry
and hyperspectral remote sensing in extracting information related to land cover
mapping. The chapter starts by providing an overview on the use of field spectroradiometry in examining the spectral discrimination between different land cover
targets. Within this framework, the main factors affecting field spectroradiometric
measurements are discussed. Subsequently, the recent developments in spectral
libraries from field spectroradiometric measurements are touched upon. Following
this, an overview on the main statistical approaches employed in spectral separation of
different land cover targets is provided, linked with the most important scale factors.
In this framework, a critical review and some examples of recent related studies and
spectral libraries are furnished as well. The second part of the chapter is focused on the
use of hyperspectral remote sensing imagery from airborne or satellite platforms for
obtaining regional estimates of land use/cover. An overview of the different techniques employed in mapping land cover types from such data is first presented with
selective examples from case studies. Next, the importance of observation scale in
land cover extraction using remote sensing data is discussed. Finally, the main
conclusions are drawn and the challenges toward a more precise estimation of land
use/cover from field spectroradiometry and hyperspectral remote sensing with respect
to spectral information acquired at different spatial scales are highlighted.
15.2 FIELD SPECTRORADIOMETRY
The science of spectroradiometry began in the nineteenth century with the use of
spectrometry (the study of human vision only) and later with the interpretation of
aerial photography (Schaepman, 2007). Yet, the real scientific involvement of
spectroradiometry started with the development of airborne multispectral sensors
in the middle of the twentieth century (Short, 2009). Ever since, field spectroradiometry is an integral part of the remote sensing science since both use the sun’s
radiation as the primary light source. The key variable in spectroradiometry is the
spectral reflectance. This is emphasized as opposed to imaging spectroscopy, which
primarily analyzes the causes and exhibition of spectral absorbance from imagery and
scattering processes that occur when light strikes the target (Milton et al., 2009). The
FIELD SPECTRORADIOMETRY
287