4 Applications to Water Resources
The main applications of remote sensing and GIS in water resources are data
collection, data management, analysis functions, physical and mathematical representation of hydrologic data, surface hydrology, and groundwater management.
These can be summarized as mapping, monitoring, and modeling.
4.1 Mapping and Monitoring
GIS and remote sensing have been commonly used to map and monitor floods,
floodplain, suspended sediment, turbidity, chlorophyll, and total phosphorus. An
important step in these analyses is the correct mapping of land-water interface. The
land-water interface is located in regions where both aquatic and terrestrial resource
systems co-exist in space and time.
4.1.1 Mapping Land-Water Interface
The ultimate objective of flood mapping is to delineate precisely where the landwater interface is located, preferably at multiple stages of inundation. A typical
procedure for mapping land-water interface using remote sensing and GIS techniques involves a number of steps shown in Fig. 5.3.
The land-water categorization begins with the identification of the spectral band
which can separate land and water bodies. The distinction between water and land
varies among the spectral bands, as can be seen in Fig. 5.4. The figure shows the
visible and near infrared bands of Ikonos satellite imagery of an area in Michigan,
USA. The visible bands (wavelength, 0.45–0.70 μm), as often the case, show
variation within water bodies but do not show sharp differentiation between water
and land surfaces. However, the reflected infrared bands (wavelength, 0.7–1.10 μm)
can detect water areas as solid bodies with little reflective variation and are ideal for
land-water categorization. The actual land-water categorization is usually performed
using slicing operation or supervised and unsupervised classification. The slicing
operation is based on a cutoff value which distinguishes between water and land. All
pixels with values below the cutoff value are categorized as water and those with
higher values as land.
Supervised classification and unsupervised classification are two important and
very basic methods in the processing of remote sensing image. In supervised
classification, the user indicates the characteristic spectral signatures of known
surface types such as water or land. The system then assigns each pixel in the
image to the surface type to which its signature is most similar. Unsupervised
classification is more computer-automated. It allows the specification of parameters
that the computer uses as guidelines to uncover statistical patterns in the data.
5 Geographic Information Systems and Remote Sensing Applications in Environmental. . . 205
The main applications of remote sensing and GIS in water resources are data
collection, data management, analysis functions, physical and mathematical representation of hydrologic data, surface hydrology, and groundwater management.
These can be summarized as mapping, monitoring, and modeling.
4.1 Mapping and Monitoring
GIS and remote sensing have been commonly used to map and monitor floods,
floodplain, suspended sediment, turbidity, chlorophyll, and total phosphorus. An
important step in these analyses is the correct mapping of land-water interface. The
land-water interface is located in regions where both aquatic and terrestrial resource
systems co-exist in space and time.
4.1.1 Mapping Land-Water Interface
The ultimate objective of flood mapping is to delineate precisely where the landwater interface is located, preferably at multiple stages of inundation. A typical
procedure for mapping land-water interface using remote sensing and GIS techniques involves a number of steps shown in Fig. 5.3.
The land-water categorization begins with the identification of the spectral band
which can separate land and water bodies. The distinction between water and land
varies among the spectral bands, as can be seen in Fig. 5.4. The figure shows the
visible and near infrared bands of Ikonos satellite imagery of an area in Michigan,
USA. The visible bands (wavelength, 0.45–0.70 μm), as often the case, show
variation within water bodies but do not show sharp differentiation between water
and land surfaces. However, the reflected infrared bands (wavelength, 0.7–1.10 μm)
can detect water areas as solid bodies with little reflective variation and are ideal for
land-water categorization. The actual land-water categorization is usually performed
using slicing operation or supervised and unsupervised classification. The slicing
operation is based on a cutoff value which distinguishes between water and land. All
pixels with values below the cutoff value are categorized as water and those with
higher values as land.
Supervised classification and unsupervised classification are two important and
very basic methods in the processing of remote sensing image. In supervised
classification, the user indicates the characteristic spectral signatures of known
surface types such as water or land. The system then assigns each pixel in the
image to the surface type to which its signature is most similar. Unsupervised
classification is more computer-automated. It allows the specification of parameters
that the computer uses as guidelines to uncover statistical patterns in the data.
5 Geographic Information Systems and Remote Sensing Applications in Environmental. . . 205