Integration of Remotely Sensed Data into GIS

Remote sensing can be incorporated into the system in a variety of ways: as a measure of land use, impervious surfaces, for providing initial conditions for flood forecasting, and for monitoring flooded areas (Neumann et al., 1990). The GIS allows for the combining of other spatial data forms such as topography, soils maps as hydrologic variables such as rainfall distributions, or soil moisture. This approach was demonstrated by Kouwen et al. (1993) where their grouped response unit (GRU) included satellite-based land use and lies within a computational element that may be either a sub-basin or and area of uniform meteorological forcing. In HYDROTEL, Fortin and Bernier (1991) propose combining SPOT DEM (digital elevation model) data with satellite-derived land use and soils mapping data to define homogeneous hydrologic units (HHU). In a study of the impact of land-use change on the Mosel River Basin, Ott et al. (1991) and Schultz (1993) have defined hydro-logically similar units (HSU) by DEM data, soils maps, and satellite-derived land use. They also used satellite data to determine a vegetation index (NDVI) and a leaf water content index (WCI), which are combined to delineate areas where a subsurface supply of water is available to vegetation. The distribution of microwave remotely sensed near-surface (0-5-cm deep) soil moisture was analyzed to identify areas of high soil moisture gradients (Mattikalli et al., 1998). This analysis showed a direct correlation between soil moisture dynamics and soil texture. Soil moisture data were employed in a hydrological model linked to a GIS, to predict subsurface hydraulic conductivity.

Remote-sensing systems use raster format for collection and acquisition of data. Many of the commonly employed GIS systems (e.g., Arc/Info) mainly use the vector format to store data layers. In this format, data are collected as points, lines, and polygons, where each structure holds information for a specific region (Fig. 3). Both the vector and raster structures have advantages and difficulties that are well described in the literature (e.g., Peuquet, 1984; Burrough, 1990), yet their fundamental differences make the integration a complicated task (Piwowar et al., 1990). In the recent past, many commercial GIS have been adapted to offer raster image display and handling capabilities (e.g., Arc/Info Version 6.0 or later), and several others offer both raster and vector capabilities (e.g., GRASS). The integration of remotely sensed data with GIS data occurs naturally in a raster GIS because data structures are approximately the same for both sources. In a vector system, the integration requires more effort, and several technical problems need to be overcome for the true integration. Important problems in the integration are the raster/vector dichotomy, generalization, and accuracy of digital information (Piwowar et al., 1990; Lunetta et al., 1991). Although the raster/vector dichotomy is a major impediment for a true integration, a significant advancement has been made to resolve the issue (e.g., McKeown, 1987; Conese et al., 1992; van der Laan, 1992; Westmoreland and Stow, 1992). These studies have employed a variety of approaches including use of quadtrees, object-oriented methods, knowledge-based systems, expert systems, artificial intelligence, etc. to achieve the task of true integration (Fritsch, 1992; Molenaar and Janssen, 1992). Examples of some commercially available systems

Origin

35

35

35,

36

36

22_

-if

33

-if.

i

36

36

37

36

36

J6_

h

35

35

37

37

38

¿5

35

?7

37

37

A

38

38

38

34-

37

37

tf-

Si

JV

T\

41

39

39

39

lo

40

K

41

39

39

39,

'40

40

40

40S

--.33

35

36 )

35

37

JL 38

41

39

f 40

Figure 3 Representation of spatial data in a GIS: (a) raster-formatted data consists of a sequence of orderly placed pixels (or picture elements) and (b) vector-formatted data consists of polygon entities to represent features.

that have some integration capabilities include GRASS, Arc/Info Version 6.0 onwards, ERDAS IMAGINE, and PCI.

Integration of raster and vector data types requires an efficient raster-to-vector (and/or vice versa) conversion routine. Mattikalli et al. (1995) developed a methodology for the separate but equal type of integration, in which the key process is a raster-to-vector (and vice versa) conversion. The procedure makes use of some built-in routines commonly available in most vector GIS, and some intermediate data formats, viz. lattice and SVF (single variable file). This approach has been employed by Mattikalli (1995) to integrate remotely sensed satellite data derived from both fine- and coarse-resolution sensors with digitized map data.

0 0

Post a comment