Cartographic Generalization in Multi-scale Environment: Case study of Lamu County, Kenya

At survey of Kenya (SoK), map generalization is still carried out manually. This results in datasets that are not harmonized, process is slow, large data storage demands, loss of detail in the downscaling and duplication of effort when multiple scales are involved. For a faster generalization of a topographic map at large cartographic scales a generalization workflow was developed in this work that ensures a harmonized and linked multi-scale database using a base data at 15, 000 containing feature classes to different scales. The main objective of this work is to generalize geospatial data using ArcGIS and QGISs operators. It discusses the process of vector based cartographic generalization using a case study of Lamu, Kenya. The vector dataset used was at basic scale of 15, 000. The datasets contained the following feature categories transportation, water features, vegetation boundaries, swamps and other special and unclassified features. General and Survey of Kenya specifications and constraints for each scale of generalization were used to symbolize the layers after generalization. Contour and spot height features were separately regenerated by varying the contour interval and spot height spacing, for each scale, using Global Mapper analysis tools using lamu DEM to create contours and spot heights. Results obtained were compared with old maps and new features for quality assessment. Findings indicated that, cartographic generalization using GIS softwares provides a good opportunity to generalize large scale data and this can be possible to generalize up to two steps for optimal results as per the current GIS generalization operators. However, there is a requirement of editing and symbolization to preserve important details and to add other map elements. Results obtained indicate optimal solutions for two steps in the generalization hence further research is recommended to enable more steps of generalization and also explore possibilities and outcomes of change the order of generalization to check for significant variations if any on results.