Food-Water-Energy ADE
The primary objective of FWE ADE is to extend the existing CityGML data model, to store FWE and its nexus related properties for developing a digital twin of the study area. By doing so, not only the status quo can be mapped, analysed and visualised, but future analytical scenarios can also be built, simulated and visualised. For stakeholders, such an integrated data model and visualisation tools can further support them to sketch sustainable urban development policies which is a need of the time. Keeping in mind the challenges of data availability and its granularity, FWE ADE introduces four thematic modules: FWEBuilding, FWELandUse, FWEArea and FWESystem.
While FWEBuilding and FWE LandUse extends the CityGML core thematic classes of building and landuse with FWE specific attributes, FWEArea and FWESystem are two new classes having multisurface geometry and related FWE attributes introduced in the CityGML core. Newly defined FWEArea object is defined as a city object having multisurface geometry covering the entire study area. It can be neighbourhood boundaries, zonal boundaries, municipality boundaries, or political boundaries which are typically static do not change with time. Based on the underlying FWEBuilding and FWELandUse, FWE and its nexus related properties along with temporal changes of population and LandUse change are summarised and introduced at this spatial level. Additionally, properties which naturally belongs only at FWEArea level or datasets of certain FWE and its nexus properties which are only available at FWEArea level are introduced here. Similar to FWEArea, FWESystem is also defined as a city object having multisurface covering the entire study area. Extending FWEArea up by a spatial level, FWESystem covers entire system boundary of the study area which can be borough, city or even an entire district. These boundaries are also static and typically do not change with time. Since FWE ADE is an integrated data model of different spatial levels, for any given time and use case, data related to properties of population, FWE and its nexus can easily be scaled up or down from a taken spatial level using data inter or extrapolation. FWE ADE as a package is flexible in a way that based on use cases each of its thematic modules can be used as a standalone data model or can be scaled up from the smallest spatial level i.e FWEBuildings to the highest spatial level i.e FWESystem or vice-versa by using aggregation functions in SQL and data inter or extrapolation. A high level UML package diagram can be represented as below. More information including the FWE ADE UML diagram and its XML schema definition (XSD) can be found here.
An example implementation workflow using FWE ADE enriched CityGML LandUse dataset with SimStadt biomass workflow using Feature Manipulation Engine (FME) can be represented as below.
As shown in the above diagram, as a first step, using FME software, CityGML LandUse dataset will be enriched with the required input parameters of surfaceArea, surfaceAreaUnit and vegetationCoverCategory available in the FWE ADE XSD (as defined in the parameters list under chapter B). surfaceAreas can be calculated using FME while vegetationCoverCateogry is custom codelist which is based on land cover satellite images. For the German case study of Landkreis Ludwigsburg, vegetationCoverCategory is derived from the 2016 national scale crop and land cover map of Germany based on imagery acquired by Sentinel-2A and Landsat-8 (Griffiths et al., 2018). These input parameters are mandatory for calculating biomass potential from the CityGML LandUse datasets using SimStadt. The enriched CityGML LandUse is then inputted to the SimStadt’s biomass workflow. The output CityGML LandUse file with output parameters from the FWE ADE XSD will be directly added to the 3DCityDB which is already extended to store the FWE ADE parameters. From 3DCityDB, based on the user needs either the enriched CityGML LandUse file with biomass potential can be extracted or the enriched CityGML FWEArea/FWESystem file with biomass potential values can also be extracted. Below image shows an exemplary visualisation of the above process.
![FWE ADE concept](img/FWEADE3.jpg)