The nature of human habitats : revealing outdoor recreation preferences through landscape utilization

Abstract

A reduction of accessible green space has deteriorated peoples’ opportunities for recreation. To be able to assess landscapes’ potential for recreation, indices and frameworks have been developed. These have mostly relied on expert knowledge rather than analyses of empirical data. Empirical data could be achieved by studying actual landscape usage by recreationists via Public Participatory GIS (PPGIS), where surveys are employed to gain spatial data of peoples’ recreational habits. Analysing such data is challenging, as other aspects than preference, for instance accessibility, also affect where recreation occurs. This thesis investigates what landscape characteristics are important for recreationists, how an index of recreational potential can be created, and how PPGIS methodology can be improved to better understand recreation. It also evaluates the Perceived Sensory Dimensions framework, a proposed design tool based on how humans perceive environments. These aims are achieved through a literature review of which forest characteristics are preferable, combined with two PPGIS studies employing novel methodology to analyse the choice of location for recreation in Sweden. The literature review resulted in a proposal for a recreation potential index for forests in Sweden, where large trees, proximity to water, and the absence of traces of forestry were identified as the most important elements. The PPGIS studies showed that the improved methodology, including the use of machine learning models and viewshed analysis, yielded accurate models. The models indicated several characteristics of particular importance for recreationists, such as proximity to water, recreational infrastructure and lack of urban noise. Finally, the evaluation of the PSD framework revealed it to have good internal validity, aligning with theoretical expectations. However, it also concluded that it is unsuitable as a tool for mapping landscapes based on their characteristics.