Effects of raster terrain representation on GIS shortest path analysis

Area of Science:

• Geographic Information Science
• Spatial Analysis
• Geospatial Data Modeling

Background:

• Spatial analysis relies heavily on the quality and preparation of input data.
• Data representation choices can significantly distort analytical outcomes.
• Understanding these impacts is vital for reliable geospatial insights.

Purpose of the Study:

• To investigate the consequences of two common data preparation techniques on shortest path analysis using raster terrain data.
• To evaluate the impact of network connectivity and attribute scale on analysis results.
• To provide recommendations for accurate and unbiased geographic information system (GIS) data representations.

Main Methods:

• Shortest path analysis on raster terrain data.
• Examining network connectivity by linking raster cells to neighbors.
• Assessing the influence of attribute scale ranges on cost assignment.
• Conducting biobjective shortest path experiments.

Main Results:

• Shortest path analysis outcomes are highly sensitive to data representation parameters.
• Varying attribute reclassification and network generation methods yield significantly different results.
• The choice of data preparation significantly impacts the location of linear features.

Conclusions:

• Data representation choices in spatial analysis, particularly for shortest path calculations, can lead to substantial result variability.
• Recommendations are provided to ensure GIS data representations yield accurate and unbiased analytical results.
• Awareness of data preparation impacts is essential for reliable spatial modeling.