Environmental protection of urban areas requires a wide range of data about air pollution, water pollution, soil pollution, waste management, noise, transport, landscape, and energy supply. In order to perform impact assessment and model predictions, various techniques of short-term, medium-term and long-term monitoring are required for spatial processing and time series analysis. The Geographic Information System (GIS), with data from remote sensing, monitoring networks and field sampling, is used to create datasets. In order to raise GIS spatial processing and dynamic modelling to a more efficient level, new data management tools are developed for basic data management, more complex exploratory spatio-temporal data analysis (ESTDA), dynamic modelling by partial differential equations and ordinary differential equations, and cell-based modelling focused on multicriteria analysis. In order to accelerate numerical tasks, the data structures are based on information depicted in 3-dimensional space, in a data cube, where two dimensions correspond to the surface coordinates, and the third corresponds to time. One of the strongest aspects of environmental information arranged into a data cube is its capability to support analytical methods derived from spatial modelling and dynamic modelling. The data cube is made up of cells. Each cell occupies a specific portion of 3-dimensional space in a regular grid. It can represent spatio-temporal data predicting the values of air pollution concentration, or data predicting another type of environmental pollution such as surface water pollution, soil contamination, noise, or other disruptions of our environment. Multicriteria analysis with cell-based modelling helps to find the worst sites characterized by high levels of environmental pollution. This can be derived from interpolations based on sample points of terrain measurements. The methods must reclassify the datasets to a common scale, and weight those that are more important for the analysis. Finally, the dataset represented by the weighted average of individual factors is used to identify the most exposed sites. In fact, Inverse Distance Weighted (IDW) is implemented as a deterministic method, and ordinary kriging is used for geostatistical interpolation. While the chosen interpolations in 3-dimensional space are based on relatively simple numerical tasks and statistical principles, information in cells should be supported by more powerful modelling tools that can encompass transport (consisting of advection, diffusion and dispersion), sorption, decay or degradation, reaction (either kinetic or thermodynamic) and other phenomena. In case of air pollution spreading in the atmosphere, the effect of diffusion, advection in a constant unidirectional flow field, and decay are described by an analytical solution derived from partial differential equations. Mathematical models based on partial differential equations are often transferred to the simplified form of a set of ordinary differential equations. Surface water pollution is modeled with this simplified form that deals with mass accumulation and delayed transport with compartment models. The attached case study describes the environment in the City of Prague and characterizes the basic monitored compartments of environmental pollution, such as air pollution, surface water pollution and land-use changes. A few case studies are described to show data processing and spatio-temporal analysis of short-term, middle-term and long-term observations.
A short term spatio-temporal analysis of air pollution in the central part of the City of Prague is focused on producing interpolated continuous surface maps from point samples and from air pollution models during a single day. A medium-term evaluation of surface water pollution illustrates changes in N-NO3 concentrations in selected profiles over the period of a decade based on compartment models. A long-term spatio-temporal analysis describes land-use changes derived from classified aerial images in the area of interest in the 1938-2006 period. The presented spatio-temporal analyses indicate changes in living conditions in the city of Prague, which can be useful in decision-making processes and outline common guidelines for environmental predictions. It also demonstrates spatio-temporal tasks in dependence on various time scales and spatial scales.