IconMap-Based Visualisation Technique and its Application in Soil Fertility Analysis

Abstract

With the development of geographic information science and the progress in the research of scientific visualisation, the visualisation of multivariate geospatial data has attracted much attention. This paper discusses an IconMap-based visualisation technique that enables multiple geospatial variables be represented in a single raster layer. This is achieved by extending the conventional pixel-based image structure to the three-tiered iconic design. Thereby the spatial pattern generated by the interaction between multivariate geographic variables can be disclosed, and the goal of geospatial data mining is achieved. A visualisation case study on soil organic matter and nutrients for the Shuangliu County of Chengdu City, China, was performed using the proposed approach. The result shows that the static IconMap can better display the distribution tendency of the high-end and low-end soil organic matter and nutrients, and that the dynamic IconMap can both reflect the interaction between the organic matter and the nutrients variables, and display the soil fertility levels in a comprehensive way. Thus, the IconMap-based visualisation approach is a non-fused, exploratory analytical approach for multivariate data and valuable for visually analyse the soil fertility condition.