The spatial configuration of local climate zones reveals effects on wayfinding in human walking

improved local connectivity or city accessibility, vice versa. Instead, it simply indicates the extent of spatial coherence between local and city levels. Future spatial planning should emphasize the synchronous pattern of improving local and urban spatial configuration for improving social wayfinding through physically spatial co-development at the local, urban, and regional levels. 4.1.2 Empirical associations between spatial configuration and wayfinding. Empirical evidence in this study revealed significant correlations between landscape metrics of LCZ types and intelligibility values (Table 2), achieving the research objective of defining statistical relationships between intelligibility-based wayfinding and LCZ-based urban forms. Thus, we claim that the wayfinding in human movement can be improved by adjusting configurations of LCZ types. The issue of wayfinding in human-related social science is then transformed into the issue of configurational changes in space-related urban science. Our claim is also supported by several previous studies that have detected the influences of spatial information cognized during the trip on wayfinding or spatial orientation [51]. Spatial layouts at both local and city levels, such as spatial symmetry or complexity, were considered important determinants of human walking behavior, spatial knowledge acquisition, and then the level of wayfinding [12, 52]. [10] proposed a model to demonstrate how the spatial geome- try of street corners affects wayfinding based on the 532 sets of surveys from 20 participants.

[53] found that regular geometry in an urban setting can increase intelligibility and thus way- finding based on stimulation. Identifying statistically significant relationships between spatial configuration and wayfind- ing, this study is unique in that it investigated the spatial configuration impacts of different LCZ-based urban forms. In other words, rather than focusing on general spatial layouts with- out taking into account urban function and land use, we provide further details about the impacts of various LCZ types, allowing for more precise spatial adjustments in practice. We not only complement the literature by pointing out the direction and magnitude of correla- tions between spatial configurations and wayfinding but also show how spatial planning strate- gies can be implemented. 4.2 Scale affects landscape metrics quantifications Landscape metrics quantification is scale-dependent [54], which means that the numeric val- ues as representations of spatial configurations within a city or a region may differ according to the spatial scale of the land data. Scale, in the context of landscape metrics, generally includes the ‘extent’ and ‘grain’ [55]. Extent is the geographical scope, which refers to the boundary of the GBA in this study; Grain refers to the spatial resolution of the observation units [56]. Scale effects then occur when quantifying landscape metrics-based spatial configu- rations. Firstly, landscape metrics of LCZs may change when choosing each GBA city as the

extent instead of the entire regional extent. Moreover, the grain of patch-level landscape met- rics depends on the spatial resolution of the basic land data. In this study, the grain of spatial configurations of individual LCZ patches is the same as the spatial resolution of LCZ classifica- tion, which is 100*100 m2. However, it may vary after changing the data resolution. For exam- ple, [57] used vegetation raster data with a grain of 20*20 cm2 for measuring spatial configuration. Meantime, focusing on the overall LCZ configurations, the observation unit shifts from LCZ patch to the LCZ type, and then the coarser grain of class-level landscape met- rics may suggest results that differ from patch-level landscape metrics. The opposite effects of LCZ D (low plants) on wayfinding between patch- and class-level landscape metrics (Table 2) support this statement. To improve wayfinding, a more clustered spatial layout of the overall LCZ D is recommended based on the class-level landscape metrics, whereas patch-level results PLOS ONE The spatial configuration for wayfinding PLOS ONE | https://doi.org/10.1371/journal.pone.0289780 September 8, 2023 13 / 18