Sampling GIS data
Since it is virtually impossible to obtain detailed information regarding every location on the Earth's surface (specifically for GIS), it is best to take samples from a smaller representative subset. GIS deals with explicitly spatial data. The two ways in which spatial data can be sampled is:
-Directed sampling and
-Non-directed sampling
Directed sampling involves making decisions about the objects to be viewed and cataloged. This population occupies an area and both the area and the sample population form the sampling frame.
Non-directed, probability-based sample is preferable since it eliminates bias.
Probabilistic sampling is of four types:
-Random sampling
-Systematic sampling
-Stratified sampling and
-Homogeneous sampling
The above sampling types can be combined to create a hybrid sampling scheme.
Random spatial sampling is the most basic sampling design. It allows each point, line, area or surface feature to be selected with the same probability as the next.
Systematic sampling use a repeatable pattern as a basis for selection.
Stratified spatial sampling selects small areas within which individual spots, objects or features are sampled. Stratifying simplifies the process by dividing the task into small regions.
Data quality refers to the fitness for use of data for intended applications. The components of data quality are listed below:
-The data must be reliable and accurate to be considered as usable. It should be in agreement with the real world being represented.
-The data must be current and up-to-date for the intended application.
-The data must be relevant in terms of content and level of detail
-The data must be timely for the application
-The data must be complete and precise or the degree of uncertainty must be indicated
-The data must be concise and intelligible (comprehensible to the user)
-The data must be stored in a format that can be conveniently handled (data handling involves one or more of the following operations: maintenance, transmission, distribution, classification, re-sampling, retrieval and updation)
-The data must be documented regarding its source to enable users to determine their suitability for a certain application
-The data must be stored in map projection that best meets the requirements of the application with regards to the preservation of area, shape, distance and direction
-The data should be captured at a scale using a classification scheme suitable to the application for which it will be used
-Cartographic properties (physical condition of mapping media, quality of line work, use of color and symbology, classification of features, map maintenance and revision cycle) play an important role in determining the quality of data.
Since it is virtually impossible to obtain detailed information regarding every location on the Earth's surface (specifically for GIS), it is best to take samples from a smaller representative subset. GIS deals with explicitly spatial data. The two ways in which spatial data can be sampled is:
-Directed sampling and
-Non-directed sampling
Directed sampling involves making decisions about the objects to be viewed and cataloged. This population occupies an area and both the area and the sample population form the sampling frame.
Non-directed, probability-based sample is preferable since it eliminates bias.
Probabilistic sampling is of four types:
-Random sampling
-Systematic sampling
-Stratified sampling and
-Homogeneous sampling
The above sampling types can be combined to create a hybrid sampling scheme.
Random spatial sampling is the most basic sampling design. It allows each point, line, area or surface feature to be selected with the same probability as the next.
Systematic sampling use a repeatable pattern as a basis for selection.
Stratified spatial sampling selects small areas within which individual spots, objects or features are sampled. Stratifying simplifies the process by dividing the task into small regions.
Data quality refers to the fitness for use of data for intended applications. The components of data quality are listed below:
-The data must be reliable and accurate to be considered as usable. It should be in agreement with the real world being represented.
-The data must be current and up-to-date for the intended application.
-The data must be relevant in terms of content and level of detail
-The data must be timely for the application
-The data must be complete and precise or the degree of uncertainty must be indicated
-The data must be concise and intelligible (comprehensible to the user)
-The data must be stored in a format that can be conveniently handled (data handling involves one or more of the following operations: maintenance, transmission, distribution, classification, re-sampling, retrieval and updation)
-The data must be documented regarding its source to enable users to determine their suitability for a certain application
-The data must be stored in map projection that best meets the requirements of the application with regards to the preservation of area, shape, distance and direction
-The data should be captured at a scale using a classification scheme suitable to the application for which it will be used
-Cartographic properties (physical condition of mapping media, quality of line work, use of color and symbology, classification of features, map maintenance and revision cycle) play an important role in determining the quality of data.
No comments:
Post a Comment