Page 109 - Digital Analysis of Remotely Sensed Imagery

80 Cha pte r T w o

is an especially important factor to consider if the remote
sensing data are used to produce elevational information
such as DEMs. However, it is a less important consideration
in thematic mapping.
• Third, what does the study area look like geometrically?
When deciding which type of remote sensing data is the most
appropriate, the geometric properties of the study area need
to be considered. They include its size, shape, orientation,
and type of terrain (e.g., the proportion of land, if land is the
object of study). These factors govern how many scenes of
imagery have to be purchased. More images may be needed
if the area is not oriented in parallel to the orbital path or has
a highly irregular boundary. In the worst case a small portion
of the area can spread into several neighboring scenes. In this
case, a few half- or even quarter-scene images may be
purchased, instead of a full-scene image.
• The last factors to consider are related to quality, reliability,
and currency of the data source. For instance, what is the
revisit period and acquisition dependability of the data? This
quality is important in carrying out longitudinal studies of
ephemeral phenomena such as flooding and fires. The
Earth’s surface is in a state of constant change. Some ground
covers (e.g., forest) change faster than others (e.g., urban).
When purchasing the data, the user needs to know the
acceptable time frame of the data. If the features or
phenomenon under study do not change quickly over time,
then data recorded years back are still useable. If the temporal
resolution is too coarse, then these data cannot fulfill such
applications as fire monitoring. Recent data have to be
acquired at a higher cost.

2.8.2 Seasonal Factors
When considering what data are the optimal, the user must be
aware of the seasonal factor, even if the subject of study is not
directly related to phenology. Imaging should take place at a time
when the phenomenon under study is maximized, and thus most
easily distinguished from other features or phenomena. It is not a
good idea to obtain wintry images when the ground is likely to be
buried underneath snow. Another reason for considering seasonality
is because different seasons have different shadow lengths and
different chances of cloud cover. In general, summer images have
the shortest topographic shadow, whereas winter ones have the
longest shadow. Topographic shadow is generally not considered
desirable as it may obscure critical information on the phenomenon
under study. This is especially true in the mountainous setting.
Shadow also degrades the accuracy of mapping if the data are