Geography Reference
In-Depth Information
sensed data, because it was very difficult to obtain the additional non-remotely
sensed data (the ground reference data in particular), and the gathered ground
reference data would only be partly suitable for some purposes. The study area was
large with some locations inaccessible during field-work.
7.2 Concluding Remarks
The kernel of this study was whether, how and to what extent applying the various
remotely sensed data that were used here, would be an effective approach to
classify the historical and current land use/land cover, to monitor the dynamics of
land use/land cover during the last four decades, to map the development of the
irrigation areas, and to classify the major strategic winter- and summer-irrigated
agricultural crops in the study area of the ERB.
It is true that the development of remote sensing techniques focuses greatly on
construction of new sensors with higher spatial and spectral resolution, but it is not
possible to ignore the data of the older sensors (especially, the LANDSAT-mis-
sion) when the historical mapping of land use/land cover and monitoring of their
dynamics are needed, although their low spatial and spectral resolution in com-
parison to the new sensors launched in the last decade (e.g., IKONOS) needs to be
taken into consideration. These older sensors are still precious. To maintain the
advantages of these sensors, researchers during the last five decades have devel-
oped new and more effective digital image processing and interpretation methods,
to harvest more accurate results. Therefore, it is important to focus on the
development of new and enhanced techniques that can translate the relationship
between the general characteristics of the old acquired data and the specific
characteristics of each individual environment such as the arid and semi-arid lands.
Regarding to field-work, this remains very important as a basis in most remote
sensing applications, offering the training samples for supervised classification. It
provides for evaluation the results of classification using accuracy assessment
techniques. It is also useful to understand the specific characteristics of the envi-
ronment of the study area.
The application of the various remote sensing techniques, which were adopted
in this study, was not only related to the location of the study area, but also to
various types of remotely sensed data (see
Chap. 4.1
).
These techniques were: extraction of the borders of the study area using the
SRTM-data and ArcGIS-extensions (see
Chap. 5.1
); geometric correction based
on GCPs, and/or geometric registration based on image to image method (see
Chap. 5.2.1
)
; atmospheric correction using the ATCOR-2 program (see
Chap. 5.2.2
)
; relative radiometric normalization using the MAD-concept (see
Chap. 5.2.3
)
; enhancing the spatial resolution of LANDSAT-ETM+ data from 30
to 15 m using the Gram Schmidt Sharpening Technique to increase the spectral
resolution of ASTER-data using the fusion-technique (see
Chap. 5.2.4
)
; mosaic-
ing, subsetting and masking (see
Chap. 5.2.5
)
; training samples selection and
Search WWH ::
Custom Search