A Tool for Identifying Suitable Places for the Placement of Blue-Green Infrastructure Elements, a Case Study on the Cities of the Moravian-Silesian Region, Czech Republic

By inergency On Jan 28, 2024

1. Introduction

Digitization in the construction industry has been in the forefront of the professional public’s interest for several decades. Currently, BIM (Building Information Modeling) and CIM (City Information Modeling) methods, GIS environments, and other modeling tools are widely used and have already become the standard, especially for the design of public buildings in many countries. Even though these methods provide a tool with many functionalities, they are usually used only in the process of building design. As stated by [1,2,3], there is great potential in the use of these technologies for the entire life cycle of buildings, mainly because the costs of operating buildings are among the largest costs, exceeding the purchase price of the property.

Apart from modeling buildings, it is advisable to focus research interest on public spaces. The importance of planning public spaces is highly influenced by the degree of urbanization, which is constantly increasing. The urbanized environment of settlements is usually made up of an enormous amount of impervious surfaces, which, due to the often completely uncontrolled influence of urbanization, increase and thus cause the overflow of the sewage system and sewage treatment plants, the flooding of smaller watercourses, and the emergence of subsequent and increasingly frequent torrential floods [3,4,5].

Many authors deal with the planning of aboveground public spaces and the implementation of blue-green infrastructure [6,7,8,9,10,11]; others mention the influence of urbanization that leads to the development of underground public spaces (Japan, Finland, Canada) [3,12,13]. In both cases, the authors [2,3,14,15] agree that the key to a properly created public space is planning, both spatial planning and strategic planning, but also detailed planning that concerns a specific problem.

As a result of extreme situations caused by climate change, buildings and their immediate surroundings are affected by various climate effects, which also have a significant impact on the quality of life of the population, functionality of the infrastructure, natural ecosystems, etc. [14,16,17]. In order to effectively respond to these phenomena, it is necessary to set the right strategies and plan for the development of buildings and areas. With the help of digital technology, it is possible to carry out various simulations in the territory and predict possible scenarios, on the basis of which administrators can more easily react to these situations. Refs. [6,15] state that, as a result of climate change, society is compelled to plan, manage, and develop technologies that will allow residents and administrators to be warned. As a disadvantage of these procedures, the author finds that information technologies are dependent on electricity, and therefore, in extreme situations that negatively affect the energy system, these technologies may not work. Therefore, Ref. [14] suggests the further development of offline functions.

The available publications deal with a wide range of climatic phenomena, but the following text will be mainly devoted to the problem of rainwater and the possibility of its use and retention in an urbanized area, because it is important to ensure the management of this commodity in order to ensure quality of living conditions and the habitability of human settlements [4,18].

The concept of rainwater is often analyzed due to its extreme accumulation and the occurrence of floods, as these consequences are visible and cause great damage to property and lives. Refs. [19,20,21,22] appeal for the creation of a tool for the prevention and prediction of extreme situations, stating that if the capacity of the sewage network is exceeded, the risk of flooding increases. Ref. [6] proposes two types of possible solutions: technological (blue-green infrastructure, development of the sewage network) and soft tools (technology for predicting situations, informing residents). Depending on terrain geometry and hydromechanical properties, some areas may be affected by landslides due to unstable subsoil, long exposure to moisture, and high slope gradients. According to [6], the prevention of landslides involves the application of blue-green infrastructure, slope reduction, soil treatment, retaining walls, etc. Another extreme situation is the increasingly frequent droughts and lack of groundwater. In order to ensure their functionality, these territories should apply certain measures to secure at least some water supplies and water retention. Refs. [8,17,19,21] analyze areas facing this problem and summarize that it is necessary to ensure the effective management of rainwater, e.g., by building dams, underground reservoirs, or by applying elements of blue-green infrastructure.

Blue-green infrastructure elements are those that include vegetation and water features, including buildings and facilities. These elements include, for example, polders, infiltration basins, bodies of water (lakes, ponds, fountains, streams), green walls, green roofs, rain gardens, infiltration paving, and any combinations thereof [6,9,10,11]. The integration of blue-green infrastructure into the infrastructure of a territory is an extension of the current access to rainwater, i.e., it allows rainwater to be soaked on the same land where the building is located (it is built), or it involves the addition of soaking with a retention device, or this water is drained into the watercourses using storm sewers [23].

The fact that issues with rainwater should be solved at the level of municipalities and cities is indicated by the number of already published and constantly updated strategic documents, which, among other criteria, deal with the increased amount of rainwater falling on territories and, on the other hand, with droughts. One of the default documents is the EU Adaptation Strategy [24], which was implemented in the Czech Republic through a document entitled “Strategy for adapting to climate change in the conditions of the Czech Republic” [25], and, subsequently, many more detailed documents were prepared by individual municipalities [26,27,28,29,30], e.g., for solving problems with drought, floods, and other natural phenomena [31,32]. Unfortunately, these documents are created with the tendency to acknowledge that these phenomena are a problem and that it is necessary to solve them, but they no longer point out how, where, why, etc. By creating a strategy, municipalities consider it as a finished solution, but it is only a starting point for other tasks. There are many examples of good practice, but each territory is specific, and often these practices cannot be completely and fully adopted into everyday life [33].

As mentioned earlier, many municipalities have a developed strategy, but it is rarely reflected in practical life. A big obstacle is the lack of good practice examples in the territory of the Czech Republic and the lack of companies dealing with this. On the other hand, many interesting and revolutionary projects have been seen, but their execution was carried out inexpertly and, therefore, did not fulfill their function. The municipalities therefore have a clear idea that it is necessary to pay attention to the management of rainwater, but it is necessary to create a tool that would help them to identify critical places that require more attention [34].

2. Study Area

The territory of the Czech Republic is divided into 14 regions, and for our study, we have chosen the Moravian-Silesian Region, which is located in the northeastern part of the republic, characterized by its geographical diversity and is the third most populous. Historically, this is an area significantly affected by intensive coal mining and the related heavy industry. The Beskydy and Jeseniky Mountains reach high rainfall totals, which often threaten the area with floods. The average annual rainfall is 600–1200 mm and it is classified as an area with variable weather due to the presence of high mountain areas and the existence of the so-called Moravian Gate, which affects the weather.

The following cities of the Moravian-Silesian Region were selected for the purposes of the research: Karvina, Havirov, Frydek-Mistek, Ostrava, and Opava (see Figure 1). The reason for the selection of larger cities was primarily the need for a sufficient database about the territory, which is the most extensive in these cities. Small municipalities ordinarily do not have the essential information base, and therefore were not suitable for an analysis where data is paramount. The selected municipalities are those with more than 50,000 inhabitants; moreover, the city of Ostrava is the third most populated city in the Czech Republic with more than 280,000 inhabitants. These municipalities therefore have a clear idea that it is necessary to pay attention to the management of rainwater, but it is necessary to create a tool that would help them to identify critical places that require more attention [34,35].

Almost 13% of the area of the city of Karvina is not suitable for the absorption of rainwater, due to the geological composition of the soil and fragmentation of the terrain. Existing and new buildings are negatively affected by the excessive accumulation of rainwater, which causes waterlogging and flooding, not as a result of extreme rainfall events, but as a result of the subsidy of this water being distributed almost evenly throughout the year. Several hydrogeological assessments were prepared, which confirmed the unsuitability of this area for infiltration. There is almost no rainwater drainage there, so it is necessary to figure out where to place suitable solutions using blue-green infrastructure, while minimizing costs. The extreme load of rainwater causes the formation of an active landslide area in the steepest parts, which already affects existing buildings and public infrastructure [36]. The conditions for the use of land unsuitable for logging are even implemented in the master plan, and are therefore required for potential builders.

Unlike Karvina, the other cities do not have a detailed analysis of the territory in relation to the accumulation of rainwater. In general, rainwater is usually trapped or drained into watercourses through discontinuous storm drains. There is practically no use of rainwater in the cities’ infrastructure; the use of water is rarely considered, especially in the case of development and business plans.

3. Materials and Methods

Refs. [2,3,19,20,21,37,38] find potential in the prediction of crisis situations and planning of areas affected by stormwater in BIM, CIM, GIS, and other tools derived from them. Many publications and analyses [3,19,20,21,38] are only processed in a database environment in GIS (ArcGIS). According to [19,39,40], effective tools for obtaining data about the territory are a 3D laser scan or unmanned aerial vehicles, the output of which is a cloud of points that can be used to create a digital twin [40] of the object or territory. It is an effective method for displaying building outlines using a semantic segmentation approach. Three-dimensional laser scanning is also used for the exteriors of buildings, but a prerequisite for processing the received data is the need for high-quality computer equipment.

Ref. [41] used the Vensin model tool, which is publicly available, to model unpredictable building collapses due to improper maintenance and age. This model is based on the creation of an algorithm that graphically depicts mutual relationships, enabling the creation of flow charts. It is a suitable tool for detecting potential threats (risk assessment). To extend its function to prediction, warning, and other modeling and planning, it is necessary to link it with GIS. Similar to the previous author, Ref. [21] uses ERDAS Imagine software 2 to model the area of interest in GIS and subsequently create statistical analyses, heatmaps, etc.

Another possible approach for assessing the risks of extreme situations (e.g., floods), especially for more complex cases, is, according to [19], a new method that integrates Grey-DEMATEL into AHP (analytical hierarchical process). It is a module for multi-criteria risk assessment with the help of artificial intelligence. Ref. [19] confirms the dependence of factors through matrices that can be drawn on a map and identify the risk locations in the territory.

3.1. Research Questions

The basic research questions that were determined before the actual solution of the project were the following:

Can the provided and detected data be used for implementation into the proposed application?
Will it be possible to run the map application on a regular desktop computer?
Is the application really set up correctly?
Will the usability of the application be ensured by municipalities in practice?
Were any disadvantages of the model identified?

In the following text, answers to these research questions will be provided. However, the authors of the contribution are already aware at this moment that the application can be perceived as innovative only for the territory of the Czech Republic, while, in other countries, this issue may already have been solved.

These research questions were established based on previously conducted research and developers’ experience, which are shown in Figure 2.

3.2. Data Collection

3.2.1. Precipitation

The occurrence of rainwater in the selected cities is very variable and always depends on the geographical location of the locality. From the available statistical data of the Czech Hydrometeorological Institute [35], it was found that the largest proportion of precipitation in the Moravian-Silesian Region occurs in the period May–October (see Table 1). The phenomenon of climate change is also related to changes in the occurrence of another type of precipitation—snow—the share of which is constantly decreasing in the winter.

3.2.2. Data Collection Background

With the gradual digitization of construction and public administration, it is possible to use quite detailed data, but only for public spaces and real estate in the administration of the state or cities. Other data on real estate owned by private entities are not published (except for data from the real estate cadaster, which is required by law), and, therefore, were not processed as part of the submitted contribution [42,43,44]. Considering the amount of real estate and the rights to it, it was not even possible to capture private land.

Many cities have approached the passporting of their property responsibly and have a passport for almost everything that is located on their territory. They realize that it is important to know the state of their assets in order to manage their cities thoroughly, and that if they have the necessary data, it is possible to subject that data to analysis, research, and simulations that will help their development. The digital development of settlements is unstoppable, and if municipalities want to be competitive, they have to move with the times. In terms of data acquisition, information was divided into two types: publicly available without restrictions and that received upon request. All data or passports given by the municipalities are shown in Table 2.

The data obtained and subsequently analyzed were the following:

Basic map (default background)—Czech Office for Surveying, Mapping and Cadaster;
Satellite map—Czech Office for Surveying, Mapping and Cadaster;
Cadastral map—Czech Office for Surveying, Mapping and Cadaster;
Flood map—5-year water level—Ministry of the Environment;
Flood map—20-year water level—Ministry of the Environment;
Flood map—100-year water level—Ministry of the Environment;
Passport of communications—municipality;
Passport of parking areas—municipality;
Passport of paved surfaces—municipality;
Passport of unpaved surfaces—municipality;
Passport of greenery—municipality;
Sewer network passport (only points of street drains, heights of street drain covers, or replacement of missing data for contour lines)—municipality; network administrator;
ψ values—the rainwater runoff coefficient for the given drained area—technical standard.

In addition to the previously mentioned data, publicly available map databases were used, which created the main basis for the subsequent work. Other map documents had to be purchased from the relevant providers. Maps were further expanded with 3D models of buildings and supplemented with a map display (base map, aerial photo) with a link to available registers of the real estate cadaster and real estate territorial identification register [45,46].

Considering the fact that all the data needed to perform the input analysis are fragmented into several sources and not all documents are publicly accessible, it was therefore necessary to purchase some data and obtain some from the municipalities in a demanding manner. After the collection of all necessary sources, it was found that many were transmitted in a different form, so the data was synchronized into a usable format that can be used to work with the tools used in this study.

3.3. Methods and Tools

Identification of weak points can be done both by manual calculation and by certain software. This method is very time-consuming and requires attention during data processing. In addition, some facts regarding the type of surface and the area vary over time. Regarding the fact that contemporary societies create simplifying methods to avoid unnecessary mechanical calculations, this method is especially suitable for the school environment, where it is necessary to ensure an understanding of the given calculation principles. Another disadvantage is the limited options for updating input data [47,48].

3.4. Calculation Methods for Drainage of Rainwater from the Territory

To manually calculate the amount of rainwater, the relationship described in the Czech technical standard ČSN 75 9010 rainwater absorption equipment can be used [49]. The calculation includes the size of the regulated floor plan area from which rainwater is drained, which is determined from the following relationship:

$A_{red} = \sum_{i = 1}^{n} A_{i} * ψ_{i} [m^{2}]$

(1)

where:

A_red—total reduced drained area from 1 to n,

A_i—real drained area in m²,

ψ_i—rainwater runoff coefficient for a given drained area,

n—the number of drained areas of a certain type.

For the calculation of the reduced area A_red, the runoff coefficient ψ_i is important; it depends on the permeability of the given surface or its material and is determined in the range from 0 to 1, where a value of 1 represents an absolutely impermeable (usually non-absorbent) surface, a value of 0, on the contrary, a surface with a maximum absorption. The amount of rainwater flowing from the reduced area is determined from the following relationship:

Q_tot = A_red × q_s [l/s]

(2)

where:

Q_tot—the total amount of discharged rainwater m²,

A_red—total reduced drained area from 1 to n, in m²,

q_s—standard rain intensity of the considered periodicity in the given area [ls⁻¹ ha⁻¹] (within the Czech Republic, it varies approximately between values of 110 and 140 ls⁻¹ ha⁻¹).

In order to be able to calculate the amount of drained rainwater, it was necessary to classify individual plots according to their type. For this step, publicly available data of the Czech Cadastral and Land Surveying Office were used and in the case of ambiguously defined areas, aerial photographs had to be used. This classification was carried out for all cadastral territories of the selected cities (as shown in Table 3).

Each land type was then classified in order to assign the stormwater runoff coefficient. This was followed by the mechanical addition of values to Formulas (1) and (2), as can be seen in Table 4. The received result is only static data that can be edited, but the resulting value does not indicate the behavior of rainwater in the field. For the purposes of further use of the determined values, it was necessary to use available and new software, especially for the purposes of creating simulations.

3.5. Tools Used for Computer Representation of Locations

The R-WIM tool was created in the smart urbido s.r.o. software 2.0 environment. This software was specifically developed for wide use in the fields of facility management, City Information Modeling, and BIM. The company cooperates mainly with the municipalities of the Moravian-Silesian region. Those cities place orders for models, visualizations, property databases, and related analyses. R-WIM (Rainwater Information Management) is an innovative web-based mapping application designed to work with city data to display key geographic information. This application was created to provide a useful overview of the urban environment and improve the ability to identify critical locations within the city where stormwater retention measures need to be taken. For the purposes of creating the R-WIM interactive tool, it was necessary to use the following tools when working in urbido.

Jupyter Notebook 7 is a tool that was used to work in the Python programming language. This free web tool (software) was created to facilitate work with this programming language, to enable editing commands, checking outputs, creating data file analyses, visualizations, etc. This tool can also be used with other languages and is user-friendly, even for beginners who are learning how to program in Python, but it is most useful for scientists, and anyone who needs to graphically represent the analyzed data. Using this software, not only graphs, but also map outputs can be created. Users can write code in several programming languages, including R and Julia in addition to Python, and instantly display results, graphs, and tables right in the document. In addition, it is possible to insert text cells for comments and documentation. Jupyter Notebook also allows to export documents in different formats and easily share them with others, making it easier to present and share results and projects. It is a useful tool for data analysis, machine learning, and research. Also, it is important to emphasize that Jupyter Notebook provides interactivity within code cells. Users can run the code step by step, which is useful for debugging and experimenting with different parts of the code. This interactive processing capability allows researchers and data analysts to more easily explore data and models [50,51].

Mapbox is a leading platform for creating and managing interactive maps and geolocation services. With a high level of flexibility and customization options, Mapbox offers solutions for various fields, including web and mobile application development, navigation, data visualization, and analytics. One of the key features of Mapbox is its ability to allow the creation of custom map styles and layers, allowing users to create maps that perfectly match their needs and esthetic preferences. Mapbox is also known for its ability to provide detailed geolocation data and quicky display maps on different devices. This platform is popular among developers, businesses, and organizations that need a powerful tool for working with geolocation data and maps. Mapbox also supports the manipulation of spatial data and enables the creation of interactive map applications with dynamic layers and visualizations. Combining powerful data visualization, navigation, and geolocation analysis tools with excellent map styles and a simple development environment, Mapbox is a very useful tool for those who need to work with geographic data and provide professional-level map services [52,53].

QGIS (Quantum GIS) is a powerful open-source geographic information system (GIS) and desktop application that offers users the ability to visualize, analyze, and edit geographic data. However, the functionality of this tool can be extended using plug-ins written in C++ or Python. One of the key features of QGIS is its ability to display a wide variety of geographic data, including vector and raster data, and to allow users to create their own map styles and layers. This gives users the flexibility to customize maps and visualizations according to their needs and aesthetic preferences. The tool also offers many features for analyzing geographic data, including geocoding, distance analysis, joining, and data filtering. This is useful for performing complex geographic analysis and extracting useful information from the data. Another important aspect of this tool is its ability to edit geographic data, which allows users to edit vector data directly in the application. QGIS is also capable of working with raster data, which includes the classification, analysis, and processing of raster images. Overall, it can be said that QGIS is an efficient and pleasant tool for working with geographic data and maps, whether it is for professional GIS analysts, urban planners, or other professionals who need to work efficiently with geographic information [51,54,55].

Mapbox and QGIS tools were applied using the Jupyter Notebook programming language, not only for the visualization and data manipulation of individual cities, but also for the visualization of rainfall runoff in the territory, terrain modeling, location of street drains, etc. Linear interpolation and heatmaps were used for further data presentation.

Linear interpolation of data was used for the presented analysis, as the materials for the visualization of various entities were usually not sufficient and it was necessary to use linear interpolation of the available data. It is a method used in numerical data analysis and computer graphics. Basically, we have two points in space with coordinates A[x₀;y₀] and B[x₁;y₁], and we are looking for the coordinates of the point C[x;y], located on the line that was created by connecting these two points [56,57,58,59]. This is expressed using the relation (3):

$y = y_{0} + (x - x_{0}) \frac{y_{1} - y_{0}}{x_{1} - x_{0}}$

(3)

A heatmap was used to visualize the density of inlets in the resulting map. This kind of map serves as a universal graphical representation of the scale of a certain value using color markings superimposed on the map background to express different values that can be modeled, analyzed, and scaled. The most frequent application of heatmaps is maps identifying temperature islands in an urbanized area [60,61]. In this case, a close-to-human display was used, i.e., the most critical areas are marked in red, neutral areas in orange, and areas that are minimally or not at all affected by the given phenomenon in green.

4. Rainwater Information Management

In the following lines, the process of creating the resulting map model is presented.

The initial step for the creation of the resulting model was the processing of the terrain model. The terrain model was created from the provided data and refined based on information about the locations of the sewer network and inlets (i.e., the real heights of street drains in the given area, which accurately reflect real conditions, or real slope ratios).

The next step was to create a geojson file, i.e., information about geospatial data, from an available CSV (i.e., a format enabling tabular data exchange). Subsequently, it was necessary to load all the points into three separate fields (latitude, longitude, altitude).

A grid of new points was generated from the created fields, for which a value of 5/100,000 of latitude and longitude was chosen, which determined the value of the distance between individual points. This value was determined as a suitable compromise between the indicative value and the time or memory requirement.

In the case of information about the location of inlets, data of differing quality were provided by the municipalities. Some municipalities had a passport processed; others did not have any data. Linear interpolation was therefore applied to the original list of all inlets, their height values and the new grid set, thanks to which it was possible to thicken the network of inlets in the area. This step allowed the height value determination of the generated network with sufficient accuracy in the following procedures.

This process needed to be firstly applied to a data sample, in this case, data from the center of the area was chosen, in order to verify the functionality of set procedure (see Figure 3). This step was important for the time-consuming nature of the whole process of modeling the territory. In a small area, it was found that the data interpolation was set correctly, and, therefore, it was possible to use it without any problems, even in a larger area.

The same algorithm was applied to the entire area (urbanized territory of the municipality) and, at the same time, a conversion to DataFrame was used (data converted to a two-dimensional table), to facilitate the easy saving of results to a CSV file.

The resulting CSVs were reduced by the QGIS program before importing them into the database, so that all surfaces with a lower permeability index were used, leaving out grassy surfaces, fields, etc. Through this process, the reduction of the grid from 3.6 million points to less than 90 thousand was achieved, which resulted in extreme memory saving and high performance of the used computer devices [45,54,62]. The resulting interpolation of the entire area is shown in Figure 4. Figure 4 is the graphically represented terrain, i.e., the elevations of the terrain, in which the missing elements were filled through correctly set interpolation.

From the data interpolation and statistical values of the precipitation, it was possible to develop a 3D model enabling the display of the slope of each individual parcel and, thus, it was possible to determine the places where the accumulation of rainwater will occur. Since it was based on the calculated relief based on the height data of the inlets, this model is only approximate and does not capture, for example, potholes and the unevenness of roads, etc., which, in practice, can cause an issue due to water accumulation. To solve these problems, it would be necessary to use terrain mapping with another technology, which would induce a huge time, technical, and financial demand. However, these problems could occur only in certain spots, without significant influence on the wider area. In addition, within the scope of the territory being processed, this fact, or the level of detail, was not the subject of the solution.

However, with the help of created 3D model, it is very easy to distinguish where water accumulates in a certain place and where it drains from the urbanized area. This was the main purpose of the presented project, i.e., to create a 3D model that would be user-friendly for the broadest range of users.

The 3D model was created on the basis of terrain data, which visually represents the amount of rainwater falling on individual parcels in the form of a simplified column display. Each plot was assigned a runoff coefficient and a slope. This model does not only show static absolute data. The model accounts for the reduction of the amount of rainwater caused by the slope of the terrain, the possibility of infiltration, and the surface runoff to the storm drain. The resulting model captures the area where excessive accumulation of rainwater occurs. In these areas, it is appropriate and desirable to implement precautions such as infiltration beds, flower beds, infiltration paving, etc.

The next step was the visualization of the drains, which captures the area that can be served by each drain. Standard values were used, where it is assumed that each inlet would serve an area of approx. 400 m². A heatmap was used for the representation of inlets where dense or sparse coverage of the sewerage inlets was immediately visible. In the Mapbox library, it is possible to create this layer automatically and further specify various parameters, such as area radius, transparency, intensity of points, color, and many others.

The result is a set of specialized maps consisting of the following:

Runoff coefficient—a map showing the basic range of public areas classified according to the water permeability of the given surface, or according to the runoff coefficient of the given area. Individual sub-areas are interactive, and it is thus possible to view individual attributes of a given map element (area, area type, surface material, total surface runoff, etc.).
Drains—a map extracting data about the city sewer system, with inflow objects being graphically represented here—street drains, or objects used to divert surface runoff rainwater into the sewer network.
Heat map—an interpolation map showing the density of sewer inlets. Based on a simple graphic representation, the map shows the accumulation of street drains (point objects) on the municipal sewer network in a given municipality.
Visualization of precipitation in 3D—an interactive map showing in 3D part of the total precipitation that creates surface runoff from a given area. The individual totals are thus represented by the height (of the water column), which is influenced by the coefficient of runoff of rainwater typical for the given surface.
Terrain—an extrapolated 3D terrain model of the given area. A map showing a 3D model of the urbanized surface of the given municipality, which was created based on the processing of available datasets. Individual heights are extrapolated ×10 for clarity (i.e., lowest point on the map ×0, mean value ×5, maximum value ×10).
Water column—mathematical model of critical points. Map showing in 3D the direction of the surface runoff of rainwater to the points of accumulation of this water (maxima of columns). The calculation was performed using an inverse function applied to the terrain map in an algorithmic combination with a 3D precipitation visualization map. The result is a graphic representation of the places where rainwater accumulates from the surface runoff.

Users of the application can get information in one place, not only about precipitation statistical data and the behavior of precipitation in the area, but also information about, for example, the type of surface, ownership, absorbency, terrain, etc. The terrain model, which has been refined using different data files and passports, has been refined and it can be said that the accuracy of the presented model is many times higher than those submitted by individual municipalities.

The novelty of the WIM tool is primarily due to the processing of basic (input) models, which can be processed in the form of interconnected layers, including connections to various data environment (e.g., data from the real estate cadaster, data from the Czech Hydrometeorological Institute, etc.) algorithms providing calculations and analyses, e.g., in the form of calculations of surface runoff depending on the type of surface and runoff coefficient.

One of the key outputs of the WIM tool is interactive maps of the urbanized area with a representation of runoff conditions. These maps form a comprehensive basis for processing the analysis and management of rainwater in an urbanized area, while the individual maps are processed in the form of interconnected layers. By combining partial map sources (base map of areas, maps of terrain morphology, drainage networks, permeability of areas, or maps of total precipitation), the resulting interactive map of the surface background will be created, which will determine problem areas, i.e., places where rainwater accumulates, etc.

5. Results

The resulting R-WIM application can be expressed by individual consecutive steps from problem identification to implementation of a computational interactive model and its verification in a real area (Figure 5).

The R-WIM application enables its users to graphically express real problems with rainwater in an urbanized area by means of an interactive computational simulation of spatial phenomena in the area. The R-WIM application is freely available online via the link: “wim.urbido.cz/uvod (accessed on 25 January 2024)”. From this link, specialized maps of individual statutory cities in the Moravian-Silesian region can be accessed online (Frydek-Mistek, Havirov, Karvina, Opava, and Ostrava). Within the web application, individual map layers can be viewed cumulatively or separately through the interactive display of public areas of individual statutory cities in 2D and 3D views, including the display of selected attributes and the possibility of modeling over time. The export of the model is shown in Figure 6 and Figure 7, other outputs are shown in Appendix A.

Resulting Replies to Set Research Questions

In this section, answers to research questions are given.

An elementary problem occurring at the very beginning of the project solution was the insufficient database of data suitable for their subsequent application. This was mainly due to the incompleteness of the data, the high error rate, but especially the lack of a basic passport of the municipalities’ property. Another problem was the promise of the necessary data from their administrator, but in subsequent communication, data were no longer provided to the authors. Fortunately, most of the selected municipalities are already ready for the digitization of the construction industry according to the new Building Act [42], and therefore maintain a very extensive database of all the necessary data. Many data had to be purchased, searched in public sources, or subjected to detailed analysis and interpolation.

Considering the high detail of the model and the detailed background of the embedded data, the display speed is dependent on the user’s hardware equipment. The processing of each map itself took several days, or weeks, using the services of the VŠB-TUO supercomputer. Displaying the maps itself is no longer so demanding, as the data calculation has already been carried out and the user can see the result. The map application was also created as a public web interface accessible to everyone without restrictions, and, therefore, some of the map’s functionalities had to be limited for possible viewing on a regular desktop computer. However, this limitation does not affect the quality of the display and is not noticed by the average user.

During the annual heavy rains in spring, which regularly occur in the area of interest, field mapping was carried out in places where excessive amounts of rainwater accumulate and thus cause problems in the sewer network and also in the daily life of residents (inappropriate access to the land, aquaplaning of vehicles). At the same time, this water can be effectively accumulated in the elements of blue-green infrastructure and ensure the supply of underground water, the creation of water features, or its use for watering public greenery, etc.

One possible example was the situation on 17 Listopadu street in Karvina, where rainwater regularly accumulates on the local road (Figure 8). At the same time, the evaluation of the detected condition was carried out in the same place as the results of the proposed application, and it was found that the problem was identified in the place that was also identified in the presented model (Figure 9). Figure 9 shows the place where all surface water from the surrounding area flows and is therefore the worst place in the given area.

The advantage of the application processed in the software environment smart urbido s.r.o. is that it is possible to respond to current weather trends, as the database can be continuously downloaded and the model thus changes according to the current situation in the area (the model update frequency has not yet been determined).

Considering the fact that the application was created specifically for municipalities, which will incorporate its results into their strategic documents and planning, the real effect of the model will only be clear after some time. However, the advantage is that, thanks to the fact that we use our own software, we are not dependent on other paid software. Another advantage of the software smart urbido s.r.o. is that it allows the insertion of additional modules that are needed for further territorial planning, based on the needs of municipalities.

The map model, including other project outputs, will be applied in the planning and management of individual municipalities. At the beginning of the project solution, there was communication with the representatives of the municipalities, who reacted very positively to the creation of the presented application, because all strategies regarding rainwater only inform that it is necessary to solve this problem, that elements of blue-green infrastructure have already been applied somewhere, and that, in the era of climate change, it is important to manage rainwater efficiently. However, these general definitions and insights do not indicate how to solve these problems. Therefore, the authors came up with a proposal to create a free web application that will allow municipalities to search for places where it is necessary to apply elements of blue-green infrastructure and thus ensure better habitability of the territory. Memoranda on the applicability of the output were drawn up with individual municipalities and the Moraviansilesian Region, who promised and welcomed its inclusion in their strategic plans.

As mentioned above, the model can be changed depending on the needs and current situations in the territory, and, therefore, the authors will provide maximum cooperation for the final setting of each map tailored to all selected municipalities.

Apart from the advantages of the map model, the authors had to deal with many obstacles. The model processing schedule was tight because it was necessary to generate certain basic data about the territory from sources other than municipal. The difficulties during the creation of this model were mainly connected to its demands on the performance of the necessary equipment, when, at a resolution of 0.00005 latitude and longitude, almost 4 million points were created in the territory of Karvina. The double resolution (fineness) increases the points number to 16 million, which would increase requirements of memory, time, and performance demands. Regarding the size of, for example, the statutory city of Ostrava, it was unimaginable to use such a high resolution.

It was also necessary to assign the generated grid points to individual parcels, which would take too much time using purely mathematical methods, so the QGIS tool was used instead and the intersection of the parcel layer with this grid was performed. To speed things up, the points were limited only to the city center and only those parcels that had a lower permeability (paved surfaces, paving, etc.) were selected. Even so, the calculation of points took less than 6 h. In the case of a denser points network and larger areas (e.g., the entire city), the calculation process would take a few days. The possibility to parallelize the model and divide it into segments is its advantage.

However, when looking at the visualization in the application environment, it is possible to find that some surfaces are displayed jaggedly. This effect is caused by the creation of a model that is based on the cadastral delimitation of the territory, and data is then applied to each parcel. This situation was created only in some places in the territory, where linear structures, such as sidewalks, streets, etc., are included in multiple parcel numbers.