Within the framework of BANG, i3mainz worked on a novel information system in cooperation with the General Biology and Atmospheric Physics departments of Johannes Gutenberg University (JGU) Mainz, as well as the Max Planck Institute for Chemistry. BANG aimed to contextualise bioaerosol data measured at sensors with relevant spatial data. The developed system enables an analysis of the correlations between measured data, geodata and calculated atmospheric data.
In the research field of biogeochemistry, bioaerosol data are currently collected, classified and managed in local databases by individual institutes. In the process, scientists usually lack tools for the visual analysis of their measurement data, especially on their geographical context. Moreover, researchers can initially only access their own data, as interfaces between the different databases do not exist. Only after a lengthy process do the individual institutions publish their data in a central database of the “National Center for Biotechnology Information” (NCBI), where they become accessible to other research groups.
The system, which was further developed in 2016, links bioaerosol data from JGU Mainz and the Max Planck Institute for Chemistry (MPI-C) with relevant geodata (such as land use or the occurrence of tree species) and information on air movements in different atmospheric layers. The focus of the work was on a validity check of the applied method. Calculated air flows, provided by the Institute of Atmospheric Physics Mainz, are spatially represented and interpreted. The flow model results are stored multidimensionally and allow for a multitude of possible two-dimensional grid combinations, which makes a manual visual interpretation difficult. Therefore, methods for the automated generation of rasters and the derivation of potential maps were developed within the scope of a master thesis.
In several steps, raster information is generated from the multidimensional format, summed up, interpolated and averaged weekly. Areas with rainfall events above certain threshold values are excluded, as these significantly influence the dispersion of particles in the air. The calculated weekly grids are then offset against the percentage cover of a selected tree species (e.g. alder or birch). The resulting potential maps show those areas with tree cover that potentially emit pollen at a given time period.
The integration of the data sets used for the analysis into the WebGIS is based on OGC-compliant web services. Access is realised interoperably by means of WMS, WFS and WCS.
The standard-compliant interfaces enable both other research groups and the system’s web client to access the data interoperably. The focus of the web client is to enable the user to visually analyse the various data. Initially, both static and dynamic spatial data are visualised in a classic web GIS, which is currently being supplemented by atmospheric data. Beyond the visual interpretation possibilities, automated processing for analysis is also planned in the application via a WPS.