Spatial Planning Decision Based on Geomorphic Natural Hazards Distribution Analysis in Cluj County, Romania

Abstract

Spatial planning decisions form the basis of territorial progress by enhancing the resilience and opportunities for local and regional development. Therefore, decisions made as a result of multidisciplinary studies based on GIS assessment of all involved factors can provide a real and up-to-date image of the analyzed territory. In this context, geomorphic processes are among the factors restricting development, affecting built-up areas, transport infrastructure, and economic activities. This paper assesses geomorphic processes at the level of Cluj County, Romania, which mainly consist of active landslides that directly impact the accessibility of communities and their degree of safety, while accelerated soil erosion severely affects the output of land used for agriculture. GIS technology and a semi-quantitative model for determining the landslide hazard were used to classify landslides across occurrence probability classes. This methodology was implemented in Romania through Government Decision no. 447/2003. The USLE model was used to determine the soil erosion. As a result, the territory of Cluj County, which is the study area of this paper, was entirely included in one of the classes of risk concerning active geomorphic processes. Another important aspect consisted of categorizing transport infrastructure according to risk classes. A population risk assessment was also performed, taking into account the degree of accessibility of the territorial emergency department in the event that such hazards and processes occur. These results form the basis of proposals to efficiently plan the county territory, adapting decisions to the present trends in the evolution.

1. Introduction

The dynamic geomorphic processes that develop across a territory depend on the morphometric, morphological, and bedrock features, as well as on the degree of human intervention. They may be heterogeneously distributed and have a minimal size, providing diversity to the territory, but most of the time they have a negative impact on a territory, because of the area covered and the rate of development [1,2,3,4,5,6]. One of these processes is soil erosion, which affects large areas at a European level, especially in the context of climate change [7,8,9]. One can also remark on landslides, which have also grown larger due to the increase in extreme rainfall events, small-sized seismic movements repeated over short periods of time, and the decrease in forested areas in some places [9,10].

The revised universal soil equation erosion model was used to classify the analyzed territory according to classes of soil erosion, taking into account the conditions of their occurrence and the works intended to reduce soil erosion [11].

A key component of natural risk management is spatial planning, which can predict an area’s growth and set boundaries based on the possibility of a risk appearing [12]. In the context of the increasing trend of expansion of built space, studies are needed to avoid building on land vulnerable to risk processes [13,14]. Monitoring landslides is a necessary activity, given that Romania is one of the European countries with the most extensive areas with high and very high susceptibility [13]. The creation of a landslide inventory that is constantly updated could help to avoid the issuance of building permits in high risk areas [12,15]. However, it is not only areas currently affected by slope processes that need to be highlighted but also susceptible areas, as there is a risk of building permits being issued in areas where there is the potential for hazards to manifest themselves. Therefore, maps showing the degree of risk are necessary for good spatial planning [12,16,17,18,19,20]. Moreover, these risk studies need to be constantly updated and validated with field observations, to allow risk mitigation measures to be adjusted according to the development of hazards [20,21].

The purpose of the studies is not only to avoid authorizing construction or infrastructure projects, but also to argue for the allocation of funds for land stabilization measures [12]. In addition, the estimation of potential financial losses due to landslides could help to increase awareness of the negative effects of these risk processes in the population and local governments [12,14].

The risks from natural hazards are often underestimated by the authorities managing these problems, who lack expertise in the field. The consequence is poor preparedness for natural hazards [22]. Therefore, it is desirable that natural hazard mitigation strategies are implemented by local governments in collaboration with specialists from various fields [13,23,24].

Through territorial planning, contingency plans can be drawn up tailored to the specific area and the likelihood of risk occurrence. Spatial planning strategies should be determined in cooperation with emergency specialists, who can provide ideas and feedback [18].

Carrying out studies that analyze the likelihood of natural hazards helps to create short- and long-term management strategies [12,18,25]. Short-term strategies include measures that are adopted in emergency situations to minimize the impact of the occurrence of natural hazards, while long-term strategies focus on technical prevention measures, as well as spatial planning measures [18].

Law No 575 of 22 October 2001 on the approval of the National Territorial Planning Plan-Section V-Natural Risk Areas regulates geographically delimited areas within which there is the potential for the occurrence of destructive natural phenomena that may affect the population, human activities, and the natural and built environment, and cause damage and human casualties [26]. The natural risks referred to in the law include earthquakes, floods, and landslides. The importance of risk studies at county level derives from this law, which states that the declaration of an area as a natural risk area is made by decision of the county council on the basis of natural risk maps approved according to para. (Art. 2). County councils may also apply for funding from the state budget for the execution of natural risk maps and natural risk prevention and mitigation works. Annex 6 also provides two maps at national level showing the potential for landslides and the type of landslides [26]. Unfortunately, this law is outdated at present, as it has not been updated; the general recommendation is that natural risks should be analyzed in every local planning document. Hence the importance of risk studies, which are designed to provide a current picture of landslides.

Although Law 575 of 2001 was the first mapping of the UATs affected by landslides, subsequent studies have detailed and represented at national level the areas subject to landslides [19,27,28]. The importance of studies at the national level is to provide a basis for public authorities in making decisions on the routing of major roads or utility infrastructure [28].

At the national level, there are policies that regulate how to manage risk processes. Romania’s Territorial Development Strategy is the long-term programmatic document that outlines the development directions of the national territory for horizon 2035. This document also underpins the entire national spatial planning system and is the basis for strategic documents at regional, county, and local level (territorial development strategies, spatial planning plans, regional development plans, urban development plans), as well as other development strategies at national level with territorial relevance and impact [29]. The territorial diagnosis mentions landslides as processes occurring in hilly and lowland areas, one of the causes being massive deforestation.

Strengthening measures to protect the natural environment against risks is one of the specific objectives under general objective 4 (protecting natural and built heritage and enhancing the elements of territorial identity), and it proposes protecting heritage and promoting measures to regenerate natural capital. Specific objective 4.3 (reducing the vulnerability of areas subject to natural risks) also proposes the priority rehabilitation of areas with high vulnerability to natural risks (floods, landslides).

The next territorial level at which natural hazard policies exist is the development region. At the level of the North-West Region, a Development Plan for the period 2021–2027 has been prepared, which includes information related to landslides, and 192 administrative territorial units have been identified where landslides could occur. A high potential for landslides characterizes a share of 45% of the total 192 TAUs, and together with the medium-risk potential, this share increases to 60%. Compared to the situation highlighted in Law no. 575/October 2001 [26], there has been a lack of updates to the landslide database, resulting in the need to map risk areas [30].

Landslides have also been analyzed for various purposes concerning spatial planning issues: assessment of land vulnerability to risk processes in urban development [31], impact of land use change on landslide susceptibility [25], assessment of geo-hazards for increasing resilience at the local level [14], landslide management in the urban environment [32], and analysis of how landslides are integrated into spatial planning policies at the European level [13].

One of the objectives of this paper was to use GIS technology to identify the probability of landslide occurrence using a semi-quantitative model, which allows for the classification of different territories across Romania according to classes of landslide occurrence probability. This methodology has been enforced in Romania through Government Decision no. 447/2003 [33]. The inclusion of certain areas in the high-probability class leads to the identification of hotspot zones, allowing for the prioritization of actions to reduce the negative effects caused by these natural slope processes. In addition, decisions have to be made regarding the planning of the studied territory and regarding the design of new infrastructure elements outside areas that have the potential for the emergence or reactivation of landslides.

Another objective of this study was to identify spatial planning measures to help mitigate risks to infrastructure and human settlements and to prevent negative consequences in the event of potential landslide activation.

Study Area

This research focuses on Cluj County, situated in the North-West Region of Romania, covering 6674 km² (2.8% of Romania’s territory) and with a population of 679,141 as per the 2021 census. It is a strategically positioned and economically influential county, ranking second after Bucharest in Romania’s economic landscape [34]. The diverse landforms include the Apuseni Mountains in the west, surrounded by basins, plateaus, and hills, forming a complex topography [35] (Figure 1). Because of its geographical location and degree of economic development, Cluj County has a strategic position that allows for the polarization of large areas within the Transylvanian Basin and even beyond. Economic activities as well as its demographic, cultural, and tourist attractiveness provide a high potential for economic development and therefore an increased impact and higher degree of influence over the neighboring areas.

The land use in Cluj County comprises 63.8% agricultural land, 25.1% forest, 2.9% construction, and 1.8% transport infrastructure, with 5% identified as degraded or unproductive. Slope processes, including landslides and soil erosion, pose natural threats, impacting agriculture and infrastructure. A comprehensive study mapped 3836 active and stabilized landslides, emphasizing the need for spatial planning to address natural risk assessments and mitigate potential disruptions to human activity.

The housing stock in Cluj County has seen rapid growth, especially after 2006, with 2011–2018 witnessing the highest construction activity. Considering this accelerated urbanization, it is crucial to assess potential risks associated with territorial expansion.

In order to study the spatial distribution and determine the conditions for the emergence of such processes, 3836 active and stabilized landslides were mapped after thorough field investigations. Stable landslides may potentially reactivate and all landslides were assessed to obtain cause–effect-type information and to create a complete inventory of all the geomorphic processes that endanger human activity and the slope balance.

The spatial planning process therefore takes into account aspects related to natural risk assessment, the identification of cause–effect-type relations, and the impact assessment on human activities.

The housing stock in Cluj County has shown a greater and faster expansion of its built-up stock compared to the whole country. While between 1990 and 2006 the growth was constant, after this year the built stock registered a rather vertiginous increase, with 2011–2018 being the construction period with the most dwellings by total number, namely 33,000 [34]. It is therefore important that the territory is assessed in terms of potential risks that may manifest themselves in the context of accelerated urbanization.

In Annex 7 of Law 575 of 22 October 2001 on the approval of the National Territorial Planning Plan-Section V-Natural risk areas [26], the administrative-territorial units affected by landslides are listed by county, and the landslide potential and type (primary or reactivated) are mentioned for each locality. In the case of Cluj county, 27 TAUs are included, including two municipalities (Cluj-Napoca and Dej), one town (Huedin) and 24 communes. Regarding the landslide production potential, the high potential predominates, present in 56% of the TAUs, followed by medium (26%), medium-low (15%), and low (4%). In 55% of the TAUs, primary landslides predominate, and in 45%, there are reactivated landslides. In Cluj-Napoca, Cojocna, and Sânpaul, both types of landslide (primary and reactivated) are encountered. According to the North-West Region Development Plan, in 2017, an area of 95,500 ha was affected by landslides in the region; in Cluj county, the total area affected by natural processes in 2018 was 47,508 ha, which represents 50% of the entire region [30]. These documentations and laws show that Cluj County faces problems related to landslides and that updated studies are needed in the context of high urbanization rates, to avoid authorizing construction in areas susceptible to landslides.

2. Materials and Methods

There are many methodologies used to classify a territory across classes of probability of landslide occurrence, as well as to determine soil erosion, such as cause–effect methodologies, deterministic methodologies, expert knowledge methodologies, and methodologies based on statistical relations according to the degree of correlation of the studied phenomenon or process, which may represent quantitative or semi-quantitative analysis models.

Recent developments in technology and scientific software have allowed for the assessment of increasingly large databases that better describe the studied reality. They also allow for the emergence of better methods for assessing and performing pre- and post-event predictions that have a higher validation rate.

This study aimed to identify the best planning proposals for the county territory. Therefore, semiquantitative methods were used to analyze geomorphic slope processes. These methods have been recognized nationally and were accepted unanimously when used in feasibility studies performed for smaller administrative units, as well as counties and development regions. They are detailed below.

2.1. The Assessment of Soil Erosion in Cluj County

Studies performed at continental level for the identification of soil loss have revealed the existence of large territories in the Transylvanian Basin liable to strong soil erosion [36,37,38]. This fact is mainly caused by pluvial erosivity, improperly applied crop technology, soil features, and morphometric characteristics. These studies provided a generalized image of the areas impacted by high erosion. However, this study increased the level of detail and a new model was constructed to determine the amount of eroded soil and identify the affected areas, which allows making the best choice in terms of practical measures for intervention and the reduction of negative effects.

To identify the soil loss in the study area, GIS technology was used, as it allowed the modeling of the factors that needed to be studied to determine the amount of soil eroded. For this purpose, we used the equation drawn up by Moţoc, M. et al. [39,40] for the Romanian territory, based on the universal relation used by the U.S. Soil Conservation Service (the Revised Universal Soil Loss Equation) and taking into account the climate conditions in Romania.

$$E=K\ast S\ast C\ast C_s\ast L^m\ast L^n$$

where E—average annual erosion (t/ha/year), K—the erosivity coefficient based on climatic aggressiveness, S—the correction coefficient for soil erodibility, C—the correction coefficient for the crop effects, C_s—the correction coefficient for the effect of anti-erosion works, L^m—slope length (m), and Lⁿ—slope (%).

Considering that this equation has a generalized form, one needs to objectively quantify the values for each factor taken into account according to the specificity of the assessed territory [41,42,43,44].

2.2. Determination of the Probability of Landslide Occurrence in Cluj County

The probability of landslide occurrence in Cluj County was calculated using a semi-quantitative model, to determine the spatial probability of landslide occurrence according to methodological norms regarding the drawing and the contents of natural risk maps concerning landslides, a model which has also been used in other similar studies in Romania. In the first stage, this model involved the creation of a database of all the factors that may cause and/or trigger landslides, taking into consideration the previous studies performed in Romania. Such factors include lithology (K_a), geomorphology (K_b), structural features (K_c), hydrological and hydrogeological features (K_d), hydrogeological features (K_e), seismic coefficient (K_f), the degree of forest coverage (K_g), and the level of human intervention in the territory (K_h).

In order to compute the probability of landslide occurrence in Cluj County, GIS technology was used, as it allows for manipulation of the database and extraction of information needed for the assessment of these types of natural processes [45].

The geological structure and lithology condition regarding the emergence of landslides were considered [46,47,48], as they are among the most important factors causing landslides [49,50]. In the studied territory, the analysis of the landslide distribution according to geological classes obtained through digitization of the Geological Map of Romania, 1960, highlighted that most landslides develop in geological classes dominated by marls and tuffs (28.5% of landslides mapped in the county), marly clays, sands and tuffs (22.5% of the landslides located in the hilly sector of the county), clays, sandstones with coal, marls, and marly schists and tuffs (14% of the landslides), etc. (Figure 2).

Taking into account morphometric features such as slope and altitude, influence coefficients were provided according to the intervals defined by Government Decision no. 447/2003 [34]. The elevation was reclassified according to the landslide probability classes defined by the Government Decision no. 447/2003 using a digital terrain model (DEM) with a resolution of 10 m, downloaded from https://www.ancpi.ro, accessed on 1 March 2023. The slope was derived from the DEM using ArcMap 10.7.1 software and reclassified according to the same methodological specifications.

Therefore, the landforms characterized by low slope values (0–20) are affected by insignificant erosional processes. These are drained by valleys in an advanced stage of maturity and have a low probability of landslide occurrence, thus receiving a 0.1 coefficient. Currently, about 1% of active landslides are located in such areas (Figure 3).

An average or high probability of landslide occurrence, characterized by coefficients between 0.3 and 0.5, is found in the hilly areas with moderate or steep slopes, fragmented by valleys which have reached an advanced stage of maturity, with 80% of the active landslides in the county being located in these areas. The other 19% of the landslides affect the hilly landforms characterized by slopes steeper than 150, where there is a very high probability of landslide occurrence (Figure 3).

Rainfall is one of the most important factors in triggering landslides [51,52]. Most studies researching the probability of landslide occurrence aimed to identify the threshold values needed for triggering new events but also for the reactivation of older and unstable landslides [43,53].

The annual average amount of precipitation modelled for Cluj County represented the basis of analysis for the identification of landslide occurrence probability according to the hydrological and climatic coefficients [6]. In this respect, territories characterized by a medium amount of precipitation, as well as catchments in the mountain and hilly areas generally controlled by the precipitation at the level of their area, have an average probability of landslide occurrence, receiving a value of probability coefficient between 0.1 and 0.3. Territories characterized by moderate amounts of precipitation and mature valleys that have tributaries affected by floods, presenting sectors with active vertical erosion, have a medium-high probability of landslide occurrence and a value of probability coefficient between 0.3 and 0.5. Territories receiving an amount of precipitation higher than 850 mm/year have a high probability of landslide occurrence and a value of probability coefficient between 0.5 and 0.8 (Figure 4).

The impact of hydrogeology on the probability of landslide occurrence was determined using a hydrogeological map of Europe. Areas characterized by a depth of over 5 m for the phreatic water level correspond to a low probability. An average-high probability of landslide occurrence was considered for areas where phreatic waters flow on steeper slopes or at the feet of slopes where springs occur. These areas received a value of probability coefficient between 0.3 and 0.5. The areas with a high permeability of the upper deposits correspond to a high probability of landslide occurrence and received values of probability coefficient between 0.5 and 0.8 (Figure 5).

Cluj County is mostly located in a VI seismic area on the MSK seismic scale for earthquakes that have an intensity higher than 6 degrees on the Richter scale. As a consequence, in terms of this triggering factor, there is an average probability of landslide occurrence for about 80% of Cluj County, which received a value of probability coefficient between 0.3 and 0.5 (Figure 6). The probability of landslide occurrence induced by the seismic factor is average-high only in the South-East of the county, receiving a value of probability coefficient between 0.5 and 0.8, because this area is located in a VII seismic area on the MSK seismic scale according to STAS 11100/1993 and the parameters for the zoning of seismicity according to act P100/1992 for earthquakes that have an intensity higher than 7 degrees on the Richter scale.

Regarding the stabilizing role played by the vegetation in case of landslides, previous studies have shown that there is a lower density of landslides on forested slopes [53,54]. Data from the Copernicus Land Monitoring Service EEA were used to model this coefficient. Therefore, territories characterized by forest cover on more than 80% of their area correspond to a low probability of landslide occurrence, receiving a coefficient of 0–0.1. These are mainly mountain areas, where about 55% of the territories correspond to this description and the density of landslides is low. Territories covered by between 55% and 80% forest present an average probability of landslide occurrence and a value of probability coefficient between 0.1 and 0.3. Agricultural lands, as well as transition (deforested) lands, are characterized by a high and very high probability of landslide occurrence, and therefore received a value of probability coefficient between 0.3 and 0.5, as the stabilizing effect of the vegetation is extremely low (Figure 7).

In order to stress the influence of construction on the overloading of slopes and the high density of infrastructure elements, the territories without constructions were characterized with a low probability of landslide occurrence and a minimal value of probability coefficient, 0.1. In contrast, overloaded slopes have a high probability of landslide occurrence and a coefficient of 0.9 (Figure 8).

Therefore, using the methodology regulated by Government Decision 447/2003 [33], factors for determining the probability of landslide occurrence were provided on a scale from 0–0.1, for a very low landslide probability, characteristic of geomorphologically stable sectors, to 1, which indicates a very high probability of the occurrence of landslide processes.

The entire analysis was performed in an ArcGIS project in which specific layers for each coefficient were manipulated, resulting in the associated rasters, as well as a spatial database concerning the probability of landslide occurrence in Cluj County.

In order to obtain the average hazard coefficient (Km), the following formula was used, using the raster calculator mode in ArcMap 10.7.1 software:

$$\mathrm{Km}=\mathrm{square}\mathrm{root}(\frac{K_a\ast K_b}{6}{K}_c+K_d+K_e+K_f+K_g+K_h),$$

where Km—average hazard coefficient, K_a—lithological coefficient, K_b—geomorphological coefficient, K_c—structural coefficient, K_d—hydrological and climatic coefficient, K_e—hydrogeological coefficient, K_f—seismic coefficient, K_g—forest cover coefficient, and K_h—anthropogenic coefficient.

2.3. Choosing the Best Development Strategies for Cluj County Taking into Consideration the Natural Risks

Once there is an exact image of the natural risks generated by geomorphic processes (mainly soil erosion and landslides), the best decisions can be made in terms of spatial planning, in order to enhance protections and mitigate risks.

This database of geomorphic processes generating negative effects should be used in the planning of built-up areas and to ensure the proper management of agricultural lands. Decisions made in this respect should be included in general urban masterplans at the level of administrative units, as well as in the county general masterplan, taking into account the vulnerability of populations and infrastructure, the present trends in the evolution, and the size and frequency of events.

At county level, when updating spatial planning documentation, a series of studies are carried out to analyze the existing situation, identify problems in the territory, and propose solutions to remedy them. One such study is dedicated to the natural environment and risks in the territory. The proposals are then included in the development strategy and project portfolio. According to Law 575 of 2001, County Councils can apply for funds from the national budget to carry out landslide mitigation works. Furthermore, each administrative–territorial unit must take into account the provisions of the county-level documentation and include studies on natural hazards as well as mitigation measures in their urban planning documentation (general urban plans, detailed urban plans) [26]. Local planning regulations can also establish building bans in affected areas, protection zones, and other rules.

Blaga, Josan, and Ilieș (2014) [55] stated that the assessment of the probability of natural hazards is one of the general objectives of sustainable development plans and can be achieved through several steps: identification of areas with potential for natural hazards and the creation of maps; determination of the vulnerability of the territory to natural phenomena and processes; short and long-term planning of measures to prevent the effects of hazards; and the establishment of monitoring, prediction, and forecasting systems.

Mateos et al. (2020) [13] argued that landslide management should be regulated at European level through a series of specific measures: preliminary analysis of existing landslides; risk maps of the susceptibility of the territory to landslides; implementation of risk area management plans focused on prevention, protection, and the preparedness of the population in the event of landslide occurrence, as well as the adoption of post-disaster strategies; and information for the population and local administrations, as well as the organization of training sessions for spatial planning specialists and local decision makers.

3. Results

The above-presented methodology was applied and a database was created as a result. This database is required for the classification of Cluj County territory into classes of soil erosion and classes of probability of landslide occurrence, allowing choosing the best development strategies for the analyzed territory.

3.1. Soil Erosion in Cluj County

According to the data provided by the National Agency for Land Improvement, the analysis of works made to mitigate soil erosion covers only 30 administrative units out of the total of 81: Aghireșu, Aiton, Apahida, Așchileu, Baciu, Bobâlna, Borșa, Căianu, Călățele, Căpușu Mare, Câmpia Turzii, Chinteni, Ciurila, Cluj-Napoca, and so on (Figure 8). It is therefore necessary to inform the population about the influence of agricultural technologies applied to maintain the soil quality within their territory.

A general soil loss equation was applied. As a result, we identified the amount of soil loss at county level, as well as the most affected administrative units. At the level of these administrative units, the class of high erosion covers extended areas, so there is a need for special agricultural works depending on the soil type and the terrain features (slope, degree of fragmentation).

There are 3175 hectares affected by soil erosion that were included in the medium class, representing 7.8% of the total area of Cluj County. There are 113 hectares included in the medium-high class, and 10.5 hectares included in the high and very high class in terms of soil loss, where the amount of soil eroded exceeds 20 t/ha/year (

Table 1. Distribution of soil loss classes in Cluj County.

Class of Soil Erosion
Very Low		Low		Medium		Medium-High		High		Very High
ha	%	ha	%	ha	%	ha	%	ha	%	ha	%
54,865.4	82.6	8220	12.4	3175	4.8	113.5	0.2	10.3	0	0.2	0

).

Most of the county territory (82.6%) was however included in the class of very low soil loss, under 3 t/ha/year. These areas are mainly located in the mountains and in administrative units with a higher percent of forest cover.

3.2. The Probability of Landslide Occurrence in Cluj County

The analysis of the landslide spatial distribution at the level of administrative units highlighted the existence of large areas affected by stabilized landslides (with a certain potential for reactivation) in the following administrative units: Cluj-Napoca (2.97% of the total affected area in the county/519 hectares), Aiton (2.62%/111 hectares), Ploscoș (2.48%/103 hectares), Feleacu (2.32%/153 hectares), Ceanu Mare (2.29%/217 hectares), and Apahida (2.03%/215 hectares). In Cluj County, an area of 4948 ha is affected by landslides (Figure 9).

Analyzing the spatial distribution of landslides in Cluj County in relation to land use at a detailed level, it was found that 45.37% of the landslides affect arable lands. In these areas, agricultural technology is affected and the geomorphic risk might be increased because of the overlapping of landslides and local soil erosion. Permanent grasslands represent another land use class highly affected by landslides, as 37.7% of the landslides in Cluj County occur in these areas. In this case, there is a cumulative negative effect induced by the landslides and the decrease in the grasslands’ production capacity, not to mention the negative effect produced by the density of animal tracks in triggering landslides on slopes. In total, 7923 parcels of land belonging to different land use classes were identified as affected by active landslides (

Table 2. Distribution of landslides in land use classes.

Land Use Parcels	Number of Parcels	Area (Hectares)	%
Arable land	483	2251.7	45.37
Permanent grasslands	66	1839.9	37.07
Unproductive lands covered by reed or cattail, swamp vegetation	880	304.8	6.14
Waste heaps or dumps	15	165.2	3.33
Yards and constructions	908	122.4	2.47
Forest vegetation	258	96.0	1.94
Roads and railways	2	74.6	1.50
Permanent crops other than vineyards: orchards	247	55.8	1.12
Streams	306	48.1	0.97
Vineyards	2038	2.8	0.06
Areas near water bodies	2720	1.8	0.04
Total	7923	4963	100

). This means that there are 4963 hectares that require extensive consideration in the future in order to mitigate the negative effects and to lower the probability of the occurrence of new landslide events or the reactivation of the already existing landslides.

A probability map at county level was drawn up as a result of applying the equation concerning the probability of landslide occurrence (Figure 10).

The class of high probability of landslide occurrence, characterized by a value of hazard coefficient (Km) between 0.5 and 0.8, covers 92.3 km², representing 1.4% of the analyzed territory. A large part of the study area, 45.1% and corresponding to 3007.7 km², presents a medium probability of landslide occurrence (

Table 3. The distribution of classes of landslide occurrence probability in Cluj County.

The Probability of Landslide Occurrence
Low (Km < 0.10)		Medium (0.10 < Km > 0.30)		Medium-High (0.30 < Km > 0.50)		High (0.50 < Km > 0.80)
ha	%	ha	%	ha	%	ha	%
630.8	9.5	3007.7	45.1	2938.5	44.1	92.3	1.4

). These areas are mainly located in the hilly sector dominated by geological deposits consisting of clays, marls, and intercalation of tuffs.

A medium-high probability of landslide occurrence characterizes 1.4% of the study area, including most of the administrative units located in the mountains, such as Beliș, Valea Ierii, Margău, Mărișel, and Băișoara, where more than 5% of the administrative territory was included in the class of medium-high probability. In this case, the medium-high probability was determined by morphometric features, especially the slope, but also the higher amount of precipitation and the higher primary drainage network density, factors which cause and trigger landslides.

Based on examples of measures from the literature, each landslide affected area was adapted to the specificity and severity of the hazard occurrence.

Receiver operating characteristic (ROC) and area under the curve (AUC = 0.783) validation were used to assess and compare the obtained information map and spatial analysis techniques (Figure 11), where AUC values above 0.7 are generally considered satisfactory in landslide research [55,56].

One of the most common causes of landslides is erosion at the base of the slope, which can be caused by excavations for the construction of buildings or to access infrastructure. The risk maps produced were intended to identify slopes prone to landslides and to contribute to the establishment of building restrictions in those areas, manifested in general urban plans or zonal urban plans at the level of administrative–territorial units.

Other anthropogenic activities that can cause the manifestation of geomorphological processes include agricultural works, infrastructure development through embankments, overloading of the land with bulky constructions, modification of surface and groundwater drainage, vibrations produced by heavy traffic, deforestation, etc. [57].

Land use and urban planning can be used to establish measures to prevent and combat landslides, such as earthworks, drainage, fixing with posts or pillars, support and anchoring works (using gabions, sheet piling, cellular walls), and afforestation [57].

Other examples of long-term measures that can be implemented through spatial planning include establishing building bans in risk areas, restricting land use patterns to prevent risks from occurring, relocating existing activities away from risk areas, refusing to issue building permits in risk areas, and imposing building standards (height, building materials, foundation structure, etc.). Short-term measures include evacuation plans, establishing emergency shelters and ensuring accessibility [15,18]. Surd et al. (2005) [58] primarily proposed the afforestation of slopes and avoidance of overloading with construction and technical infrastructure. Other measures (that are more expensive however) include the construction of drainage ditches and stone or concrete walls.

4. Discussion

Taking into account the results of research on the anti-erosion role of vegetation [39], as well as the identification of areas corresponding to a high class of susceptibility to soil loss in Cluj County, in the future certain recommendations should be made in planning and strategy documents such as the County Land Use Plan, the general urban plans of each administrative–territorial unit, and zonal urban plans. There are various invasive methods that could mitigate the negative effects of soil loss. These can include crop rotation practices, where permanent crops might be replaced with perennial plants and alfalfa, which provide a better soil protection because of their higher density of roots, which have an anti-erosional role, as well as their higher water consumption, which hinders the occurrence of landslides. Other measures to mitigate soil loss should include works oriented along hills and valleys, works along the contour lines, as well as the creation of terraces to mitigate the occurrence of landslides.

Landslides are geomorphic processes that produce significant material damage, because they destroy buildings and block water, gas, and other supply systems, as well as roads. They also modify the landscape and affect large plots of agricultural land. The modeling performed to analyze the distribution of landslides and the associated probability of the occurrence of new events or the reactivation of older landslides took into consideration the current state of art, mainly based on bibliographical research, the use of expert knowledge within the research team, as well as a database including the totality of factors causing and triggering landslides, together with a very detailed inventory of all events prior to analysis, as provided by the national public administration authorities [59].

The frequency of landslides that needed the intervention of emergency (ISU) workers in the territory of Cluj County was assessed for a long period, between 1970 and 2018. One can see that there were some years with an extremely high number of events and periods of relative stability, when the slope dynamics were reduced. In the period between 2005 and 2010, the number interventions exceeded 30 events annually in 2006 (32 events) and 2010 (34 events).

The distribution by decade of landslide events highlights the 2001–2010 decade, with 144 events that needed the intervention of emergency (ISU) workers in the field. The number of events during other decades was approximately equal (45 in 1970–1980, 50 in 1981–1990, 42 in 1991–2000, and 31 in 2011–2018) (Figure 12).

In order to mitigate the negative effects of these events, several important measures need to be taken into consideration. Agricultural terraces should be created on slopes where there is the potential for the occurrence of landslides. Slopes should not be overloaded with constructions or a dense road network, which add weight to the deposits located above clays. These deposits may become wetter after a period of consistent rains and/or snow melting and this may lead to the reactivation of landslides, even superficial ones. To reduce the water infiltration into the sliding surface, drainage tubes may be used to remove the excess of water and therefore reduce the degree of ground erosion.

In terms of spatial planning strategies, it is important to ensure that the population exposed to geomorphic risk phenomena are informed properly. People should be educated and trained to understand the impact of human activities in the territory, the effects of anthropogenic pressure on slopes, and at a higher scale, the effects of deforestation and improper land use.

In the decision-making process, geotechnical studies are needed to determine the probability of landslide occurrence according to the resistance of rock deposits before new residential areas and transport infrastructure are built. People living in areas prone to risk should insure their assets. Recommendations can also be made to afforest slopes with high landslide potential using hydrophilic arboreal vegetation with a rapid growth and adaptation rate (such as acacia (Robinia pseudoacacia), pine (Pinus sylvestris), etc.); slopes can be terraced and vines or fruit trees can be planted on them, which are well suited to the soil and climatic conditions of the area.

It is necessary for institutions to use monitoring systems to reduce the time needed for the field intervention of specialized emergency services in the case of landslides. The universal soil loss equation (USLE) method and landslide semi-quantitative models are valuable tools for assessing erosion and landslide susceptibility by taking into account various factors such as rainfall, soil type, slope, land cover, and anthropic influences. However, there are limitations associated with these methods, which can result in oversimplification of complex interactions and potential inaccuracies, particularly in heterogeneous landscapes. We thus draw attention to using the most detailed models and current databases, and to validate the modelled results through direct field mapping of natural processes modelled using GIS technology. Following the background studies carried out for the updating of the County Land Use Plan, the County Council has the possibility to apply for funds to mitigate the effects of slope processes, in order to stabilize areas with a high potential for landslide activation. A first category of measures that can be adopted through spatial planning concerns the legislative field. Each administrative–territorial unit is obliged to take over the provisions of higher-level land-use plans and include them in their urban development plans. The correct approach to risk phenomena starts with the drawing up of documents that define the areas exposed to risk and establishing measures to regulate land use. More specifically, local planning regulations can implement building bans in risk areas, create protection zones, and set regulations on height, land occupation percentages, land use coefficients, foundation depth, and the obligation to carry out geotechnical studies. Of course, these documentations must be carried out in collaboration with experts from multidisciplinary fields, as well as representatives of emergency services [13,22,23,24], in order to ensure the correct approach to risks. Starting from the legislative level, the representatives of local public administrations are obliged to inform the public when issuing planning permissions, whether or not the areas in which they wish to build are at risk of landslides, and it is forbidden to issue building permits in areas at risk. Any building activity must also be supervised to avoid illegal (unauthorized) siting of buildings in risk areas.

The second category of measures includes concrete action on active landslides or areas susceptible to landslide processes. These include the stabilization of slopes using retaining walls, drainage, and afforestation. It is also advisable to avoid overloading the land with buildings in landslide-prone areas and to avoid carrying out works that could worsen the current land situation.

Another category of measures that could be taken by local public administrations concerns informing the population about natural risks and the damage they can cause. It is also necessary to collaborate with the Emergency Situations Inspectorate in carrying out public information campaigns.

5. Conclusions

The GIS modeling of soil loss and of the probability of landslide occurrence is a very important activity in territorial analysis at national, regional, or local levels. In this case, the analysis was performed at the level of Cluj County, the most developed county in the North-West Region of Romania. The territory of this county is affected by soil erosion and a significant number of active and potentially reactivated landslides.

The spatial planning decisions for Cluj County, Romania, presented in this article emphasized the critical role of multidisciplinary studies, particularly those utilizing geographic information system (GIS) assessments. The comprehensive analysis of geomorphic processes, specifically landslides and soil erosion, revealed the intricate dynamics shaping the territory and influencing its development potential. The integration of GIS technology and semi-quantitative models, such as the landslide hazard assessment and the universal soil loss equation (USLE) model, provided a nuanced understanding of the challenges faced by the region. The findings underscore the impact of active landslides on community accessibility, safety, and the compromised agricultural outputs resulting from accelerated soil erosion. By categorizing risk classes for landslides and soil erosion, decision makers can gain valuable insights into vulnerable areas, enabling informed choices for efficient territorial planning. The identified hectares affected by soil erosion, landslide occurrence probabilities, and risk classes for transport infrastructure can all contribute to a comprehensive risk assessment for Cluj County. It is noteworthy that a substantial portion of the study area faces medium to high probabilities of landslide occurrence, particularly in hilly and mountainous regions. Quantitatively, it is noteworthy that 3175 hectares within Cluj County exhibit soil erosion, categorizing them into the medium class and constituting 7.8% of the total county area. Furthermore, 113 hectares fall within the medium-high class, and an additional 10.5 hectares are classified as high and very high, with an annual soil loss surpassing 20 t/ha. The high probability class for landslide occurrence, characterized by a hazard coefficient (Km) ranging between 0.5 and 0.8, encompasses 92.3 km², equivalent to 1.4% of the analyzed territory.

A substantial portion of the study area, precisely 45.1% or 3007.7 km², presents a medium probability of landslide occurrence, primarily concentrated in the hilly sector dominated by geological deposits such as clays, marls, and intercalation of tuffs. A medium-high probability of landslide occurrence characterizes 1.4% of the study area, notably affecting administrative units situated in mountainous regions like Beliș, Valea Ierii, Margău, Mărișel, and Băișoara. In these cases, this medium-high probability is influenced by morphometric features, particularly the slope, as well as higher precipitation levels and the primary drainage network density, serving as pivotal factors triggering landslides.

These findings necessitate geotechnical studies before initiating new construction projects, ensuring the resilience of residential areas and transport infrastructure. Additionally, risk mitigation strategies, such as insurance for residents in high-risk areas and the implementation of monitoring systems, become imperative elements in the decision-making process.

County and local governments are responsible for regulating risk areas and for constantly monitoring and updating the landslide database. Appropriate spatial planning can ensure the avoidance of the negative consequences of risk processes for settlements and infrastructure and the sustainable management of affected areas.