Approaching Flood Risk Management by Creating a Three-Dimensional Model at the Level of a Watershed

Abstract

Globally, the number of major floods has been consistently significant in recent years. By using several methods of acquiring and processing geospatial data, this study aimed to develop a digital terrain model that supports the modeling of sudden increases in water levels in a river to provide a true picture of the areas at risk. The main contribution of this research is provided by the method of performing coupled geospatial, hydrological, and hydraulic calculations within the area of interest. This approach includes an analysis of all the hydrotechnical works executed in the riverbed. The research highlights the characteristics of the water flow corresponding to the maximum flows with exceedance probabilities of 10%, 1%, 0.5%, and 0.1%, as well as those associated with maximum discharges resulting from scenarios involving the failure of the storage dam in the area. The research results indicate that the creation of a 3D model at the river basin is probably the most important step in flood risk management, as the results obtained at this stage can also influence other measures that can be applied.

1. Introduction

Currently, phenomena characterized by random natural hazards are particularly complex in terms of their occurrence, intensity, manifestation, and material damage. Understanding the complex interactions that take place when hazard phenomena occur requires the implementation of integrated measurements and the use of complex monitoring systems, many of which are based on digital terrain models and digital surface models generated by LiDAR scanning technology. Integrated measurements involve the collection, representation, interpretation, and analysis of accurate and up-to-date geographic data to study, monitor, and manage the interactions that occur in addressing the problems generated by climate change, the organizing of the territory, and the monitoring of various natural phenomena, assessing the risks and the effects they could have on human life, material goods, and the environment. The technological progress recorded in the field of sensors intended for collecting geospatial data finds immediate applicability in the study of the factors that determine the occurrence of natural hazards. The rapid collection and processing of accurate geospatial data that can be easily integrated into dedicated software platforms are essential elements for supporting the risk management process and are an essential component for developing prevention and protection measures for affected areas.

This paper aims to present how different methods of acquiring and processing geospatial data are used to create a complete, accurate, and uniform digital terrain model (DTM) for a hydrographic basin; based on this, a series of maps can be generated—hazard maps and flood risk maps—with various exceedance probabilities. To perform the numerical modeling of the propagation of the flood waves that cause floods, a precise mapping of the risk areas is necessary; the very-high-resolution digital terrain model (DTM) obtained from LiDAR measurements can be used to do this.

The question from which this research starts is related to how to create a three-dimensional model at the level of a river basin for a flood risk management approach. Correlated with this question, the novelty element offered by this work involves the method of performing coupled geospatial, hydrological, and hydraulic calculations, which are carried out in the area of interest and which will highlight the characteristics of the water flow corresponding to the maximum flows with the exceeding probabilities of 10%, 1%, 0.5%, and 0.1%, as well as those corresponding to the maximum flows resulting from the scenarios of breaking the storage dams in the area. As previously mentioned, the simulation related to the propagation of the flood waves through the riverbed is based on a methodology that involves the reproduction of synthetic flood waves in different sections along the river, which are mainly hydrometric stations supplemented by upstream and/or downstream sections of the confluence with the main tributaries. The important novelty and differentiating element of this work is provided by the inclusion of the analysis of all the hydrotechnical works executed in the riverbed and the establishment of the influence of each hydrotechnical work on the hydraulic regime of the river. In this way, the coupled hydrological and hydraulic model includes all the constructive elements of the existing hydrotechnical works on the water course (longitudinal and transversal), which influences the hydraulic regime. This significant contribution can be supported by the analysis conducted by Hailemariam et al. [1], who examined current methods for flood risk assessment. The authors’ analysis was further complemented by an evaluation of how these methods can be implemented in practice. It can be noted that the authors identified methods with good efficiency; however, no concerns were raised regarding the importance of hydraulic structures.

The development of flood hazard maps aims to implement preventive measures in vulnerable areas to reduce the risk of natural disasters. The Inter-Agency Standing Committee (IASC) [2] states that an emergency can turn into a risk when there are failures in disaster response decisions. Therefore, the risk, if not controlled, can easily escalate into a disaster due to the limited possibilities to propose countermeasures for such a situation. If, in one way or another, the risk can be managed, then the disaster is averted in most cases. By using specialized risk management tools, the course of a disaster can be modified, and the response action during an emergency and afterwards limits the negative effects on human and economic activity, as well as on the environment.

Risk management [3] is a process of systematically identifying, analyzing, and reporting risk factors. It involves maximizing the probabilities and consequences of positive events while minimizing the probabilities and consequences of negative events. In this sense, the present research approach is strengthened by including the analysis of existing hydrotechnical works conducted by specialists in the field [4,5], seeking efficient methods to control the effects of a disaster and advancing the research to a higher level.

The risks generated by floods have always been a source of concern for all interested parties, including the civilian population, local and central authorities, and researchers. Each of the aforementioned parties has had the protection of life and of the civilian and economic objectives as its main objectives in situations where floods occur. In their research, Lim et al. [4] conclude by suggesting the development of measures for preventing flood damage. Similarly, Parsian et al. [5] invoke the results of their research for protecting life and civilian objectives.

An exceptional perspective is provided by Bayraktar et al. [6], who, in their research, focused on the detection of plants in exposed areas, as well as the classification of these zones. This approach effectively complements the perspective centered on safeguarding human lives and the integrity of civil infrastructure. Through the application of the proposed model, the number of plants in wild areas can be determined. This methodology, which emphasizes the inclusion of nature among the priorities that civilization must protect, holds significant importance.

Hu et al. [7] sought to evaluate how the interpolation of multi-source spatial data can lead to efficient results for the quantitative estimation of precipitation. This approach starts from the premise that floods have their origin, to a large extent, in the fact that the increase in flows is produced by large volumes of precipitation. The conclusion of Hu et al. [7] was that there has been a shift from the interpolation to the fusion of multi-source data for the spatial estimation of precipitation. This conclusion is in line with that of the present work, which uses the principle of summarizing multi-source data to determine the floodability of the analyzed areas. Das [8] has the same approach, concluding upon the need to prepare maps of susceptibility, vulnerability, and flood risk. This conclusion generates directions of action regarding the construction of structures that can and should prevent floods [8]. In reality, these structures limit and control the effects of floods, more so than preventing them. The same philosophy regarding the analysis of flood effects is also reflected in the research of Renschler et al. [9], who use data provided by the floods caused by Hurricane Irene on August 28, 2011. Their research makes a comparison between the estimated situation, which should have been within controllable limits, and the reality caused by the release of a large amount of water to avoid the dam breaking in the affected area, which led to obtaining results that meant completing the information provided by government reports or scientific studies. These results concretely support the usefulness of multi-source data fusion. The same usefulness is also proven by Dash et al. [10], who identify and validate flood maps, or by Albano et al. [11], who use the method proposed for the analysis and mapping of large-scale damage. The research of Gangani et al. [12], which was performed using HEC-RAS (River Analysis System) [13] and evaluates several open datasets for the DEM (Digital Elevation Model), tried to find ways to simulate peak floods. Their conclusions showed cyclical events of this type, offering a superior perspective in terms of forecasting. The same type of open data were used by Xafoulis et al. [14], who corroborated a dataset obtained via a topographic measurement mission. Through their research, the authors showed the importance of the precision of the topographical measurements in terms of reducing errors when assessing the extent of floods.

For areas or even countries that are frequently crossed by meteorological phenomena that can generate floods, information with this specificity is very useful—a fact proven by Saint-Martin et al. [15] by proposing the development of a DamaGIS database, which collects and identifies data on flood damage from several sources, including media sources or social networks. The efficiency of the existence of information sets is proven, especially since they are accessible through GIS software. The analysis conducted by Rosser et al. [16] happily completes the idea of Saint-Martin et al. by efficiently using the technological level reached by mobile phones. The analysis shows how smartphone users document the events they witness and upload them to social media networks, in fact sending geotagged information that can be effectively used to complete the information necessary for assessments and subsequent analyses or simulations.

Shifting the focus from flood control to flood mitigation, Liu et al. [17] attempted to estimate the impact of floods on affected areas by determining the impact on the lives of the population after a flood. Such an analysis can provide extremely useful results for increasing the efficiency of the recovery process of affected areas, applying a principle found in Business Continuity Management, ISO 22301 [18]. However, to be able to do such an analysis, it is necessary to have the other one, which provides information on flood maps. A similar concern is addressed by Naktaniek et al. [19], who propose the DEMATEL method for evaluating factors associated with the risk of pluvial flooding, providing a valuable tool for planning and rescue service activities. Likewise, the study by Fu et al. [20], which analyzes pluvial flood risk over the past 30 years, indicates that the evaluation method based on the integration of GIS and AHP (Analytic Hierarchy Process) is effective in such situations.

2. Data and Methodology

2.1. Location of the Analyzed Area

The Prut River (Figure 1) is the second longest tributary of the Danube [21], which is the most important river in Romania. The Prut River, with a length of 952.9 km, forms the border between Romania and Ukraine for 31 km, from Oroftiana to Lunca Ivancăuți, and between Romania and the Republic of Moldova for 711 km. From a hydrological perspective, the Prut River is divided into two areas of interest—the upstream sector of the Stânca Costești reservoir (Figure 2), which is approximately 139 km long, and the downstream sector, which is 591 km long. The Stânca Costești reservoir has an approximate surface area of 38 km² and represents the most important hydraulic construction on the Prut River.

According to the National Institute of Hydrology and Water Management [22], in the Prut–Birlad area, where the Prut River has the determining weight, there are 46 areas with a potentially significant risk of fluvial flooding, of which 14 are at a low risk. For these, 32 strategies have been developed so far, with the intention of covering all areas with a potentially significant risk of medium- and high-level floods. One of the important components for the management of this hydrographic area was the development of hydrological risk maps.

2.2. Methodology

In accordance with Directive 2007/60/EC [23], which establishes that EU Member States shall produce hazard maps and risk maps adapted to the zonal context, four categories of works are established, as necessary, to build the dataset required to produce hydrological risk maps for the Prut River, as follows:

• Carrying out topo–bathymetric measurements on the Prut River and the Stânca Costești reservoir;
• The creation of an integrated digital terrain model for the Prut floodplain on both banks by integrating the topographic–bathymetric data produced within the project into the available digital terrain models;
• The acquisition of satellite images/production of orthophotos and other types of geospatial data necessary for the production of flood hazard and risk maps;
• Hydrological and hydraulic modeling in order to generate flood hazard and risk maps along the Prut River.
• Figure 3 presents the execution diagram of risk maps.

To create these maps, in accordance with the provisions of Directive 2007/60/EC, three distinct flood scenarios have been developed.

These are the low-probability scenario (for maximum flows with a 0,1% probability of exceedance, corresponding to floods that have a chance of occurring once in 1000 years); the medium probability scenario (for maximum flows with a 1% probability of exceedance, corresponding to floods that have a chance of occurring once in 100 years); and the high-probability scenario (for peak flows with a 10% probability of exceedance, corresponding to floods that have a chance of occurring once every 10 years).

It is important to note that the flood hazard map is a synthesis tool that includes a number of essential details. These details are relevant for each exceedance probability considered. One such detail is the flood boundary, which represents the extent of water in each scenario analyzed. Another key detail is the water depth or level, which is divided into three classes: a water depth below 0.5 m, a water depth between 0.5 and 1.5 m, and a water depth greater than 1.5 m.

On the other hand, the flood risk map is a document that provides information on potential material and human damage in administrative–territorial units, depending on the different probabilities of exceeding the maximum flow in flood-prone areas.

2.2.1. The Digital Terrestrial Model

To streamline the process of creating risk maps, existing data in government collections are used, consisting of images with digital models, satellites, and orthophotos, which are verified, validated, and then converted into formats compatible with the GIS system chosen for implementation.

The existing digital models for the two banks of the Prut River have different characteristics, due to the fact that they are specific to the three riparian countries. For the Republic of Moldova and Ukraine, structures in the UTM 35 projection system and the Baltic Sea 77 altimetric system were used, and there was no possibility of assessing the accuracy of the digital model with the help of control points. For the right bank, located in the territory of Romania, the data structures used were in the Stereographic 1970 projection system and the Black Sea 1975 altimetric system. In this case, an assessment of the accuracy could be made using 21 control points. The result of the assessment can be seen in

Table 1. Data from the assessment of the accuracy of data from the existing digital model.

Nr. Pct.	X Stereographic 1970	Y Stereographic 1970	Z from Measurements	Z from the Existing Digital Model	Z Differences
1	446,166.399	749,856.371	5.177	5.191	0.014
2	446,166.972	749,122.036	5.834	5.895	0.061
3	446,167.195	749,118.364	5.694	5.669	−0.025
4	446,462.235	748,807.676	7.657	7.647	−0.010
5	446,464.728	748,812.093	7.549	7.517	−0.032
6	448,296.490	747,855.194	7.637	7.667	0.030
7	448,297.113	747,859.075	7.605	7.457	−0.148
8	449,729.415	746,928.357	8.196	7.563	−0.633
9	450,739.187	746,382.863	7.915	7.832	−0.083
10	450,741.408	746,379.560	7.979	7.841	−0.138
11	461,635.305	741,140.586	8.498	7.752	−0.746
12	462,311.121	741,727.466	7.973	7.573	−0.400
13	462,311.396	741,723.993	8.036	7.732	−0.304
14	463,635.554	745,686.487	9.225	8.686	−0.539
15	463,637.353	745,686.895	8.957	8.706	−0.251
16	473,769.710	745,910.468	10.989	10.523	−0.466
17	473,769.897	745,915.437	11.197	10.863	−0.334
18	487,243.490	740,493.701	12.998	12.689	−0.309
19	487,247.435	740,493.219	13.049	12.652	−0.397
20	680,717.057	678,520.475	49.973	50.905	0.932
21	708,131.935	667,821.267	104.640	104.685	0.045

.

Similarly, the orthophotomap and satellite image components show differences between the two banks of the Prut River for the same reasons. For the right bank of the Prut River, color orthophotomaps (scale 1:5000; georeferenced in the 1970 stereographic projection system, edition 2011) were taken. For the left bank of the Prut River, satellite images (in the 2008 edition (pdf format), non-georeferenced, and in 2011 (.jpg format), georeferenced in the UTM 35 projection system) were taken.

The information gaps on the left side of the Prut River, but also in the Stanca–Costesti accumulation area, required the completion of the model with data obtained from aerial laser scanning for an area of approximately 750 km², with a density of 2 points/m².

An extract from the flight plan for completing and updating the data from the Stanca Costesti area can be seen in Figure 4.

Details of the three flights performed to update data from the Stanca Costesti area are presented in

Table 2. General flight details.

Name	Value	Unit
CCNS4 project name	PRUT
CCNS4 area name	PRUT
Coordinate system	UTM North—WGS84 SPH—EGM96:35
Sensor name	LM7800 60O 400 kHz [Mode 7]
Magnitude Variation	0	Deg
Minimum side gap	154	m
Data Annotation 1	WGS84
Data Annotation 2	Alt [ft]

,

Table 3. CCNS4—computer-controlled navigation system summary.

Name	Value	Unit
Number segments	3
Segment length	62.090	Km
Operation time	0.386	h
Time per turn	120.0	S
Mean speed	54	m/s
Raw storage	11.498	GB

and

Table 4. CCNS4—computer-controlled navigation system line parameter listing.

Flight Line Name	Segment Name	Mean Height Above Mean Sea Level [m]	Mean Average Ground Level [m]	Mean Ground Level [m]	Azimuth [deg]	Segment Length [km]	Mean Dots Per Area [*dots/m2]	Eaw Storage [GB]
1	1	2242	2147	95	162	20.782	2.0	3.848
2	1	2242	2157	84	162	20.698	2.0	3.833
3	1	2242	2136	105	162	20.610	2.0	3.817

.

The digital terrain model for the hydrographic basin was developed according to the following three levels of detail:

• Level A—very detailed, with a resolution of 1–2 m, covering restricted areas of particular complexity and importance, which require detailed analysis.
• Level B—detailed, with a resolution of 4–5 m, covering the river courses and their main tributaries of sufficient length so that the results of hydraulic modeling are relevant.
• Level C—low detail, with a resolution of 10–15 m for the rest of the river basin.
• Data sources for the three levels of detail are as follows:
• The digital terrain model of Level A was obtained exclusively through aerial laser scanning, performed in such a way as to obtain a density of 2–4 points/m².
• The digital terrain model of Level B can be obtained either by aerial laser scanning, performed so as to obtain a density of 1–2 points/m², or by extracting the digital terrain model from images.
• The Level C digital terrain model was generally obtained from the integration and calibration of existing data, namely the acquisition of the digital terrain model with a resolution of 5–10 m from the National Agency for Cadastre and Real Estate Advertising or the vectorization of contour lines on 1:5000-scale topographic plans.

The time required to acquire the data necessary to generate the digital terrain model of Levels A and B through aerial laser scanning was approximately 20 h, including the movement of the aircraft in the project area.

From the LiDAR point cloud, the digital terrain model (DTM) and the digital surface model (DSM) were obtained by classifying them. Following the classification, the following point classes were generated: ground, low vegetation, medium vegetation, high vegetation, constructions, and unclassified points. The ground class also included points located on the water surface. It is worth mentioning that scanning systems based on LiDAR technologies allow the collection of a very large number of three-dimensional points that define a surface or an object, with the help of laser waves. The advantages of using this method are given by the high precision and the relatively high density of points, a fact that allows us to obtain three-dimensional models of surfaces or objects.

Figure 5 illustrates the diagram of common aerial data collection operations and their integration with GPS (Global Positioning System) and IMU (Inertial Measurement Systems) data for processing and obtaining the digital model. Airborne sensors separately collect 2D and altitude data. The laser scanner and the IMU/GPS system are synchronized with the GPS reference signal and allow for a way of data acquisition that facilitates their further processing. In this way, three-dimensional datasets are obtained, forming what is known as the “point cloud”.

The collected image, Figure 6, set required further processing, which was performed by using the ULTRAMAP V 3.9 software [24] on a computer with an Intel Core i7-8565U, four cores, 1800 MHz, and 32 GB of RAM. This software, which stands out for its manual editing tools for the digital surface model, its 3D visual analysis for a better perspective, and its good facilities for data export, requires the existence of some initial data, as can be seen in Figure 7.

Thus, to create a project in ULTRAMAP, the following initial data are required:

• Raw images, specifying that the image format is specific to the ULTRA-CAM Lp (dragonfly) camera [25];
• Internal orientations, which describe the internal geometry of the camera at the time the images were captured with it.

2.2.2. Topographic and Bathymetric Measurements

Topo–bathymetric measurements were conducted to restore and complete the existing support network, acquire topographic data necessary to generate transverse and longitudinal profiles, acquire data regarding water depth, and conduct surveys of engineering structures.

The following equipment was used to conduct the topo–bathymetric works:

• GPS model SP80 [26], used for determining the elevation at the water level and connected to the sonar in kinematic measurement mode;
• Sonar model HUMMINBIRD 798ci SI Combo, dual beam [27];
• Boat model StarCraft SL 350;
• Appropriate protective equipment.

The terrestrial reconnaissance procedure of the analyzed area is based on existing information, which is registered with the institutions in the government area that manages hydrological information. In order to conduct a terrestrial reconnaissance of the area, the points of the existing geodetic support network at the institutions that manage hydrological information were identified. The inventory of the conservation status of the support network points allowed for the identification of areas with insufficient information, where it is necessary to complete the network with new points. This step, which is very important in carrying out the measurements, allowed for the preparation of the network project and the establishment of the stationing/measurement sessions. The establishment of the optimal periods for executing static GNSS measurements was achieved by using a free planning software—GNSS Planning Tool 2.90.1.

After collecting bathymetric data, they were processed to apply corrections and calculate the exact position of bathymetric points using the SonarTRX program [28].

2.2.3. Conducting Cross-Sectional Profiling Within the Prut Riverbed and the Stânca Costești Accumulation Area

Following the analysis, 434 transverse profiles were determined and established. The distance between the transverse profiles was relative and varied depending on the morphological characteristics of the minor riverbed. For the bathymetric profiles, intermediate measurements were also performed, particularly on shorter profiles in the areas of the inflection of the watercourse. Transverse profiles were also performed in areas of sudden slope changes, respectively, and in areas with reverse slope, to the extent that these areas can be identified in the field.

For each cross-sectional profile, common points were determined to connect the elevations resulting from the digital terrain model with the topo–bathymetric measurements. The coordinates were determined using GPS measurements and the Real Time Kinematic (RTK) measurement method, using the closest ROMPOS reference stations to the project area [29], respectively, i.e., the RO VRS 3.1 GG virtual station system. The measurement accuracy of the points on the cross-sectional profile was +/− 5 cm in the plan (Stereographic System 1970) and +/− 10 cm for normal elevations (Black Sea reference 1975). In the terrestrial measurements on the cross-sectional profiles, points were measured at a maximum of 20 m, as well as at changes in the slope of the terrain.

The transverse profiles overlapped the trust values from the existing digital terrain model found in government archives. The overlap area was approximately 50 to 100 m, without exceeding the levee line. Figure 8 illustrates this overlapping action for the sector downstream of the locality of Prisacani.

2.2.4. Hydraulic Modeling

The creation of flood maps is based on the hydraulic calculation 2D model implemented in the Hydrologic Engineering Center’s (CEIWR-HEC) River Analysis System (HEC-RAS) [13].

The following datasets are applied to the hydraulic calculation model:

• The main route of the Prut River and its main tributaries;
• The calculation of cross-sections;
• Roughness coefficients;
• Hydraulic structures in the riverbed;
• Maximum flow values.

The HEC-RAS [13] mathematical model offers the possibility to determine the characteristics of the moving water flow, whether permanent or non-permanent, under uniform or gradually variable hydraulic conditions, both for rivers with single-line beds as well as for dendritic or annular ones.

This mathematical model is based on the integration of water motion equations using the finite difference method [30], addressing both the case of water in temporary motion and the case of water in permanent motion. Free water surface elevations are calculated between consecutive profiles using an iterative routine known as the standard step method.

2.2.5. Creating Flood Maps

The aim is to develop hazard and risk maps corresponding to the exceedance probabilities of 10%, 1%, 0.5%, and 0.1% for the Prut riverbed through coupled hydrological and hydraulic modeling. This approach allows for the determination of the contribution of the Prut River tributaries corresponding to the flood waves on the Prut River.

The synthetic data obtained after transposing the results of hydraulic modeling on the map for each of the hydrological scenarios with the exceedance probabilities of 10%, 1%, 0.5%, and 0.1%, as modeled in the existing flow regime and which are attached as attributes of the layer of the GIS of the inundation limits and of the transversal profiles, are as follows:

• The surface of the flooded area in the existing regime;
• The width of the flood lane in the existing regime;
• The average speed of water flow in the transverse profile;
• The level of the free water surface in the existing regime;
• The share of various water depth classes;
• The areas/localities affected under the existing regime.
• The flood risk map primarily encompasses the delineation of the following categories of flood risk zones:
• Major risk zones—areas where the construction of permanent structures is prohibited because of the frequency of flooding, water depth, water velocity, and the duration of flooding events, which render these areas as pathways for significant water drainage.
• Medium risk zones—areas that require protection through structural and non-structural measures in accordance with current legislation and regulations.
• Low risk zones—areas characterized by a low level of damage, where only local protective measures are necessary.
• The following colors are used to represent the three categories of areas in the flood risk maps:
• Red for major flood risk;
• Orange for medium flood risk;
• Yellow for areas with minor flood risk.

3. Results

Hydrological modeling was conducted on river sectors between two successive hydrometric stations on the Prut River, using six hydrometric stations.

• The number of calculation cross-sections was 434, of which:
• 360 sections were downstream of the Stânca Costești Accumulation;
• 41 sections were in the Stânca Costești Accumulation area;
• 33 sections were upstream of the Stânca Costești Accumulation.

In total, 20 flood maps were generated, containing the orthophotomap, information about localities and land use, and the boundaries of flood zones generated for different exceedance probabilities.

Figure 9 and Figure 10 provide comparative images of the results obtained; Figure 9 presents the area of the Prisacani hydrometric station with a 0% exceedance, and Figure 10 presents the same area of the Prisacani hydrometric station with a 0.5% exceedance. By comparing the two figures, the impact of a 0.5% overshot, indicated in red in Figure 10, becomes apparent.

Figure 11 presents a map of the Mastacani–Galati sector. This is one of the 20 maps generated from the hydrological model. The map shows information for floods corresponding to maximum flows with exceedance probabilities of 0.1%, 0.5%, 1%, and 10% in this sector of the Prut River. The significant importance present in the retrievable results, including this map, is provided by the differentiator of the method, by including the analysis of hydrotechnical works. The exceedance probabilities are selected in accordance with the provisions of Directive 2007/60/EC.

4. Conclusions

Activities related to the organization and development of the territory, as stipulated by Law no. 350/2001 regarding territorial development and urban planning [31], as well as to urban planning and administration, require the correct and up-to-date use of hazard and risk maps, as developed at appropriate scales and levels of detail. The procedure for creating and using these documents in the context of natural risk factors is regulated at the European level by directives and at the national level by administrative acts that establish responsibilities at the central, regional, county, and local levels, as well as the management of associated databases. However, regarding the risks generated by human activities, regulation is less developed. Although the expansion of instability phenomena induced by anthropogenic activities is more limited compared with that of natural phenomena, their impact can be significant at the local or regional level. For instance, subsidence caused by mining operations affects a substantial proportion of territories in certain areas. Current legislation [31] refers to mining areas only in the sense that they can be considered as disadvantaged areas. This categorization is used in the legislation [31] from a demographic point of view and is not related to the actual impact caused by instability phenomena. Therefore, it is crucial to develop and adopt rules and standards in this field, as well as to introduce the obligation to include specific information in the process of developing plans for land use and urban planning.

In the context of sustainable development and mitigating the negative impacts of extraction activities, mitigation measures and risk management strategies are essential. What can be considered a set of proposals generated by this research for strategies that can be implemented to reduce the risk of flooding includes the following: the creation of retention basins; the ecological rehabilitation of exploited areas; the stabilization of slopes; and the efficient management of stormwater. The creation of retention basins, the ecological rehabilitation of exploited areas, the stabilization of slopes, and the effective management of stormwater are some of the strategies that can be implemented to reduce flood risk. Furthermore, the involvement of local communities and stakeholders in the development and implementation of risk management plans can contribute to the creation of solutions that are well adapted to the specific context of each site.

The ecological health and safety of human communities are closely linked to the efficient management of natural resources and landscapes. Quarries and ballast pits, in the context of the changes they induce in the environment, require special attention in the modeling and interpretation of flood risk maps, thus guaranteeing effective planning and responses to extreme events. The rapid expansion of the urban environment is one of the most important global issues today. In order to successfully control situations resulting from this rapid urbanization, geospatial information needs to be as accurate and up-to-date as possible.

Correct geospatial information leads to accurate calculations, which is reflected in the increased effectiveness of hydrological forecasts and warnings in case of floods. Probably the most important finding of this research is that creating a 3D model at the level of the river basin represents the first and most important stage in flood risk management, because flood maps are developed based on geospatial data, and a plan of measures is generated to combat and mitigate the effects of floods. This statement is because the other stages of flood risk management, subsequent to the creation of a three-dimensional model at the river basin level, can be influenced by the results of this stage. In this regard, the main contribution of this research is provided by the method of performing coupled geospatial, hydrological, and hydraulic calculations, which are conducted in the area of interest and include an analysis of all hydraulic works executed in the riverbed. These highlight the characteristics of water flow corresponding to maximum flows with exceedance probabilities of 10%, 1%, 0.5%, and 0.1%, as well as those corresponding to maximum flows resulting from scenarios of the failure of the storage dam in the area.

To create a 3D model of a hydrographic basin, current technologies enable the simultaneous acquisition of LiDAR aerial data and digital images. This represents one of the most used methods for developing a comprehensive cartographic database over relatively large areas. Although LiDAR technology and the software for processing these data are quite expensive, and specialists in the field are few, this technology has garnered significant interest across many domains. This is due to the elimination of the need for direct physical contact with the ground surface, as well as the vast number of data points—on the order of millions—that can be collected in a single measurement campaign.