URBiNAT Scientific Partners Marcel Cardinali and Uta Pottgiesser from the Institute for Design Strategies at the OWL University of Applied Sciences and Arts (Germany), are the co-authors of a new paper on the automization of indicator development in green space health research. The paper which draws on data collection and analysis conducted as part of the URBiNAT project, is published in the journal Software Impacts (Elsevier, ScienceDirect, 2023). 

The relationship between green spaces and health is attracting more and more societal and research interest. The research field is however still suffering from its differing monodisciplinary origins. Now in a multidisciplinary environment on its way to a truly interdisciplinary field, there is a need for a common understanding, precision in green space indicators, and coherent assessment of the complexity of daily living environments. In several reviews, common protocols and open-source scripts are considered a high priority to advance the field. Realizing these issues, we developed PRIGSHARE (Preferred Reporting Items in Greenspace Health Research). It is accompanied by an open-source script that supports non-spatial disciplines in assessing greenness and green space on different scales and types. The PRIGSHARE checklist contains 21 items that have been identified as a risk of bias and are necessary for understanding and comparison of studies. The checklist is divided into the following topics: objectives (3 items), scope (3 items), spatial assessment (7 items), vegetation assessment (4 items), and context assessment (4 items). For each item, we include a pathway-specific (if relevant) rationale and explanation. The PRIGSHARE guiding principles should be helpful to support a high-quality assessment and synchronize the studies in the field while acknowledging the diversity of study designs.

PRIGSHARE was developed within QGIS v3.22 with the graphical modeler feature to automate specific repetitive processes that are necessary to compare different indicators and/or different buffers. Next to Euclidean buffers, 25 m buffered service areas (BSA) to calculate network distances are available. BSA are a more precise version of network distances, especially for smaller distances. The algorithm will calculate the chosen indicators for distances between 100 m and 1500 m with 100 m increments. AID-PRIGSHARE allows the user to choose which of the following green space indicators should be generated by the algorithm:

AID-PRIGSHARE was developed within QGIS v3.22 with the graphical modeler feature to automate specific repetitive processes that are necessary to compare different indicators and/or different buffers. Next to Euclidean buffers, 25 m buffered service areas (BSA) to calculate network distances are available. BSA are a more precise version of network distances, especially for smaller distances. The algorithm will calculate the chosen indicators for distances between 100 m and 1500 m with 100 m increments. AID-PRIGSHARE allows the user to choose which of the following green space indicators should be generated by the algorithm:

• (a) Mean vegetation index (e.g. NDVI) in Euclidean distance (−1 to 1)
• (b) Mean vegetation index (e.g. NDVI) in BSA (−1 to 1)
• (c) The amount of public green space within BSA (m2)
• (d) Public green space ratio (public green space/total buffer area) within BSA (0-1)
• (e) Access to green infrastructure within BSA (m2)
• (f) Distance to the nearest public green space (rounded in steps of 100 m)
• (g) The amount of green space uses (playgrounds, sports fields, gardening,..) within BSA (number)
• (h) Diversity of green space uses within network distance (number)
• (I) The amount of private green space of an individual (m2))
• (j) The amount of semi-public green space of an individual (m2)

The architecture of the script combines the necessary steps to generate a specific indicator in a task chain that is repeated for every distance. Next to the mandatory input of the geolocation of points (e.g. the home address of an individual surveyed) and corresponding ID field, the input is optional and depends on the requested tasks. The script uses Boolean operators to indicate which tasks should be performed. To reduce computation time, the script makes use of intermediate results and reuses them to create additional indicators.