Wildland-Urban Interface (WUI) maps of Poland

Description

The dataset contains six Wildland-Urban Interface (WUI) maps of Poland, based on different radii used to calculate building density and forest cover share. The maps are based on the WUI definition of US Federal Register (USDA and USDI, 2001) as operationalized by (Radeloff et al., 2005), which distinguishes two kinds of WUI: intermix, where housing intermingle with wildland vegetation, and interface, where settlement abuts the wildland areas. Either WUI type requires a housing density higher than 6.17 houses/km2 (1 house/40 acres in the US context). In intermix WUI, there has to be also > 50% wildland vegetation, while the interface WUI, has < 50% wildland vegetation but is within 2.4 km of a wildland vegetation patch larger > 5 km2. As there are visible differences in building density and forest cover share calculations for different radii, we applied a circular moving window algorithm testing six radii: 100, 250, 500, 750, 1000 and 1500 meters, which resulted in six respective maps. Smaller radii have the advantage that even small groups of buildings are mapped as WUI and focus on ecological processes that operate on short distances (e.g., noise pollution), whereas larger radii have the advantage that they focus on larger WUI areas, and reflect ecological processes operating at longer distances better (e.g., habitat fragmentation for wildlife species with large territories. The recommended WUI map, for those, who are not interested in a specific application is the one based on 500 m. More details on the consequences of using a particular map are shown in the research paper linked to this dataset. Acknowledgments The study was supported by the National Science Centre, Poland, contract no. UMO-2019/35/D/HS4/00117 and by the NASA Land Use and Land Cover Change Program.

Steps to reproduce

We mapped the WUI in Poland based on exact building locations and detailed wildland vegetation maps. Specifically, we analysed countrywide building locations from the official Polish Topographic Objects Database (BDOT10k), where each building in Poland is mapped as a polygon (14,895,581 polygons in 2019). In terms of wildland vegetation, we focused on forest cover only, because forest is the natural, wildland climax vegetation in Poland. We analysed the 10-m resolution Sentinel-2-based S2GLCraster product, covering the entire European Union, and freely available for scientific purposes (Malinowski et al., 2020). According to the accuracy assessment, forests were very accurately mapped (Broadleaf tree cover had user and producer accuracies of 94.6%, and 95.3% respectively; Coniferous tree cover 95.8% and 97%) (Malinowski et al., 2020). To calculate building densities and shares of forest cover, we applied a circular moving window algorithm testing six radii: 100, 250, 500, 750, 1000 and 1500 m. While the forest cover data were stored in LAEA coordinate system, and buildings were stored in PL-1992 coordinate system, the final maps are also presented in PL-1992 coordinate system. The vector maps attributes include: WUI_type - 1=Intermix WUI, 2=Interface WUI num_build - number of buildings in each WUI patch

Institutions

Categories

Funding

Licence