Geoscience Reference
In-Depth Information
maintenance of the data, the modeling and analysis tasks and the presentation of the
data. Two data models can be distinguished:
Vector data model that represents geometrical objects defined as points, lines or
polygons, including 3-dimensional (3D) point data.
Raster data model , which is used to represent continuous or categorical data on
gridded surfaces.
Each data model has advantages and disadvantages with respect to its storage
efficiency as well as the types of processing that can be performed. To some extent the
choice of data model is often pre-determined by the user's experience and preference
or by the nature of what it represents in the real world. The two data models will be
discussed in detail throughout the topic. While the emphasis will be on the vector and
raster data model, Chap. 15 will deal exclusively with point data. Vectorial data are
composed of two components: the geometry and the attribute data associated with
each geometric feature. Depending on the underlying database associated with the
vector dataset, the links between the features and attributes can be direct one-to-one
joins or consist of one-to-many joins. The diagram in Fig. 1.1 illustrates how features
can be represented using the raster and vector data models.
This topic deals exclusively with digital spatial data, that are predominantly used
in remote sensing and digital image processing. Each chapter deals with a specific
spatial data type that includes vector, raster and 3Dpoint/raster datasets. Earth observ-
ing imaging sensors are acquiring imagery of the planet at increasingly higher spatial
and temporal resolutions, resulting in large data archives.
Some of these datasets are freely available, for instance the Landsat and MODIS
archives operated by the USGS and NASA and PROBA-V archive operated by the
ESA. These archives are increasingly being relied upon by the Earth observing com-
munity and Earth system scientists.
These disparate sources of spatial data are providing tremendous opportunities to
answer questions related to environmental science, climate change, socio-economic
analysis and disaster management, as well as being integrated into Web mapping
systems. Improved technology, in terms of the image quality, means that the size
and quantity of data is also increasing, requiring more efficient tools to process and
analyze the datasets using easy-to-use, easy-to-modify software solutions.
1.4 Earth Observation Data
Remote Sensing can be very loosely defined as a process of collecting information
without coming in contact with the object. With respect to Earth observation, it can
be considered to relate to the acquisition of imagery of the Earth's surface (Richards
and Jia 1999).
Efforts to acquire aerial imagery began at the start of the twentieth century
using cameras mounted to balloons, kites and aeroplanes. These technologies were
adapted and refined largely for military reconnaissance during the First and Second
 
Search WWH ::




Custom Search