URBAN GEO BIG DATA project
URBAN GEOmatics for Bulk Information Generation, Data Assessment and Technology Awareness (URBAN GEO BIG DATA) is a project of national interest (PRIN), funded by the Italian Ministry of Education, University and Research (MIUR)―id 20159CNLW8. The scientific partners of the project are Politecnico di Milano, Politecnico di Torino, National Research Council (CNR) Institute for Electromagnetic Sensing of the Environment (IREA), Italian Institute for Environmental Protection and Research (ISPRA), University of Padua and Sapienza University of Rome. The scientific partners produced various datasets in different domains within the URBAN GEO BIG DATA project, focusing on five cities: Milan, Turin, Padua, Rome and Naples in Italy.
Communicating research outputs to scientists, public and private sector administrators, and citizens regarding various real-world phenomena using dynamic, interactive, and understandable visual interfaces is essential for effective decision-making. As a result, visualization of the datasets generated within the URBAN GEO BIG DATA project has been an integral part of the project. The datasets can also be queried and processed. The geoportal hosting these functionalities is developed at Politecnico di Milano using free and open source software. The geospatial visual interface of the geoportal is a virtual globe, created using CesiumJS API. The geoportal is available on the Web at http://urbangeobigdata.como.polimi.it/ and hosts various use cases.
The first use case explores the 3D visualization of buildings in Milan, Turin, Padua and Naples using the OGC standard CityGML. The CityGML data is generated from data in Esri shapefile format by one of the scientific partners of the project, University of Padua, and converted to KML/COLLADA/glTF format, so that they can be visualized on the virtual globe. Thematic information of the buildings can be queried. This use case also overlays the flood risk map published by the Lombardy region of Italy as open data using the OGC standard Web Map Tile Service (WMTS) and simulates floods to aid in the informed decision-making process regarding adaptation measures and mitigation of flooding effects (Figure 1).
Figure 1: Flood simulation in Milan with a flood risk map and 3D buildings visualization and query
The second use case visualizes maps of deformation of Earth’s surface, produced by CNR IREA, on the virtual globe. The technique used by CNR IREA is widely acknowledged for the investigation of ground deformation and allows the generation of mean deformation velocity maps, and for each point on the ground, deformation time series. The mean deformation velocity maps for Milan, Turin, Padua, Rome and Naples are overlaid on the virtual globe using the OGC standard Web Map Service (WMS). Each mean deformation velocity point on the map can be clicked on to plot the cumulative deformation time series from 1992 to 2010 (Figure 2). The data used to create the plot is retrieved using the OGC standard Web Feature Service (WFS). Moreover, raster maps that are generated from the cumulative deformation time series are animated on the virtual globe. Each frame of the animation, i.e., raster map, is retrieved using WMTS (Figure 3).
Figure 2: Visualization of mean deformation velocity maps of Rome and query of a deformation point on the virtual globe
Figure 3: Visualization of a raster file of 2005 from raster deformation time series of Turin on the virtual globe
The third use case visualizes and processes EO-derived Land Use and Land Cover (LULC) datasets (land consumption maps from ISPRA of 2012, 2015, 2016 and 2017; land cover map from ISPRA of 2012; GlobeLand30 of 2000 and 2010; and Global Human Settlement (GHS) built-up of 1975, 1990, 2000 and 2014). All the datasets are overlaid on the virtual globe using WMTS, and the ones of multiple years are animated using the technique used for animating the raster deformation time series. Moreover, datacubes are created from the datasets, and they are processed using the OGC standard Web Coverage Processing Service (WCPS). When a user clicks on a dataset of a single year, WCPS returns the value of the clicked pixel, which is the land cover class. On the other hand, when a user clicks on a dataset of multiple years, WCPS returns the change of land use or land cover classes for the clicked coordinates (Figure 4). Besides executing processing for a set of coordinates, WCPS also calculates the amount of change of selected land use or land cover class for an area drawn by the user for two selected years (Figure 5).
Figure 4: Change of classes for the clicked coordinates of GHS built-up maps from 1975 to 2014
Figure 5: Amount of change of artificial surfaces, around 3%, for the drawn area around Campania in Italy from 2000 to 2010 according to GlobeLand30
The source code of the geoportal with version 3 of the GNU General Public License (GPL) is available at https://github.com/kilsedar/urban-geo-big-data-3d.
Candan Eylül Kilsedar, who developed the geoportal of the URBAN GEO BIG DATA project, is a PhD student and research fellow at Politecnico di Milano and a charter member at OSGeo. She can be contacted on LinkedIn, Twitter and via email at candaneylul.kilsedar [at] polimi.it.