Spatial analysis and data mining of urban trees

Abstract

Tree coverage in urban spaces is a theme of great importance for current societies, given all the benefits that green spaces provide to the population, especially in large cities. Trees fulfill a very important role to ensure quality of urban living and urban environmental quality, and as a result trees are considered to be an element of urban infrastructure. In spite of the recognition of the importance of tree coverage, events in which a street tree falls or needs to be preventively cut down are quite frequent, damaging property and causing disturbances in the routine of the population. From a rich dataset on urban trees for the city of Belo Horizonte (MG, Brazil), this paper proposes contributions towards the identification and solution of problems related to tree coverage, with special emphasis on felled trees. Data mining techniques are employed in search of consistent patterns, expressed as association rules or temporal sequences, that are related to felling events. We also show a VGI tool to updating and expanding the original dataset.