Research Methodology - Historical Land Use/Land Cover Classification Using Remote Sensing - page 64

data (raw-data) to GRID-format; ArcToolBox/Spatial Analyst Tools/Hydrology:
(Fill/Flow direction/Flow accumulation/Conditional-Con/Stream to feature/Add
one point—.shp file-/Watershed); Conversation Tools: (from raster—Watershed-/
Raster to polygon); and Analysis Tools: (Extract/Clip). Throughout the proposed
results, the spatial distribution layer of the natural borders of river-basin was
obtained from the SRTM-data. Concerning ASTER-data, there has been an
unwillingness to depict the river basin edges because of their higher spatial res-
olution rather than the SRTM-data. Unfortunately, dealing with this data proved to
be exhausting and full of errors. Therefore, a return to the SRTM-data ensued.
There has been no accredited map issued by the Ministry of Irrigation that draws
the borders of the ERB. The majority of Syria's irrigation projects lie within the
natural boundaries of the ERB, except for some projects in the north and the south
of the city of Aleppo, where waters have been extracted from the Euphrates River
for the past five years. This means that many of these projects are not introduced in
this study, as they occurred after the date of the last remote sensing data used (i.e.,
2007).
5.2 Pre-Processing of the Satellite Data
''A good player never makes more effort than he needs to win''—old Arabic
wisdom.
Remote sensing data may have two common types of distortions (systematic
and non-systematic). This is because the method act of the Earth observation
system and the characteristics of Earth's surface (Richards and Jia 2003 ). There
are a variety of preprocessing procedures that could be applied on satellite data:
finding and replacement of damaging lines of pixels; geographical registration of
image and geometric rectification; radiometric calibration and atmospheric cor-
rection; and correction the topographical effects. According to Mather ( 2004 ), pre-
processing procedures used to correct the generated deficiencies of geometric and
radiometric formation of a remotely sensed image, and then it used to remove the
errors of data. These deficiencies and errors have to be removed or at least
manipulated, if it is achievable, before the starting with imagery classification.
Which method would be applied, is dependent upon the goal of study. The most
availability of preprocessing procedures or programs—automatic, is for coarse and
medium spatial resolution data (e.g., LANDSAT-TM) and for high temporal
resolution data (e.g., NOAA-AVHRR).
A good optimization in presentation of an individual object in the dataset of
remote sensing data, is a result of a suitable selection of digital image prepro-
cessing procedures. This goodness can be confirmed using a visual interpretation
(Liu and Mason 2009 ). There are many digital methods to better enhancement of
an image. These methods have the benefit of increasing the visual interpretability
of used data and thus the thematic information of interest could be easily derived.
The common three methods of image-enhancement are: (1) enhancement of