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

city was also the hot spots of PD values in LCZ 3 (compact low-rise), LCZ 5 (open mid-rise), and LCZ 6 (open mid-rise) as well as the hot spots of IJI values in LCZ 9 (sparsely built) and LCZ D (low plants) (Fig 6(b)–6(f)), which means that LCZ 3, LCZ 5, and LCZ 6 types in Zhongshan city were significantly subdivided than other cities, and LCZ 9 and LCZ D types were highly interspersed into other LCZ types. Moreover, LCZ 3 and LCZ D types were both particularly disaggregated in Guangzhou city shown as the hotspots of both PD and IJI (Fig 6 (b), 6(f)). Dongguan city also had separately distributed LCZ 6, similar to Zhongshan city (Fig 6(d)). LCZ 9 and LCZ D types were highly juxtaposed with other LCZ types in Hong Kong (Fig 6(f)). For patch-level metrics of ENN and SIMI, Zhongshan city was still the cold spot of all patch-level landscape metrics, in which patches in LCZ1 (compact high-rise), LCZ 2 (compact mid-rise), and LCZ D (low plants) were more spatially connected due to their significantly shorter interpatch distances, compared to other cities (Fig 7(a), 7(b) and 7(d)). Patches of LCZ 9 type exhibited the least similarity to their adjacencies in Zhongshan and Hong Kong (Fig 7 (c)). Comparing the spatial clusters of LCZs between class and patch levels, LCZ 9 showed the same patterns among cities (Figs 6 and 7(c)), whereas a contrast configuration of LCZ D type between class and patch levels was observed (Figs 6(f) and 7(d)). In detail, LCZ D type was overall intermixed with other LCZ types on average, and the patches of LCZ D type were con- trarily spatially clustered.

Overall, hot spot analysis for landscape metrics contributed to determining the priority cit- ies for spatial reconfigurations. The locations showing hot or cold spots had significantly dif- ferent landscape metrics of LCZ types, in which spatial reconfigurations had the potential to improve wayfinding greatly. In our case, spatial clusters of landscape metrics were shown in Zhongshan, Guangzhou, Dongguan, and Hong Kong. Here, we defined Zhongshan as the pri- ority location for spatial reconfigurations, because Zhongshan is the only city that shows hot or cold spots of each landscape metric (Figs 6 and 7). The details about the spatial distributions of LCZ types can refer to S1 Fig in S1 File. 4. Discussion Explaining intelligibility-based wayfinding using spatial analysis of urban forms, this study provides not only an insight into understanding wayfinding as a social issue from the Fig 6. Hotspot results of class-level metrics in LCZ types. The spatial clusters of the values of SPLIT in LCZ 2 (a), the values of PD in LCZ 3 (b), LCZ 5 (c), and LCZ 6 (d), as well as the values of IJI in LCZ 9 (e) and LCZ D (f) among cities. https://doi.org/10.1371/journal.pone.0289780.g006 PLOS ONE The spatial configuration for wayfinding PLOS ONE | https://doi.org/10.1371/journal.pone.0289780 September 8, 2023 11 / 18