Search

This paper presents a result of landslide inventory mapping at the Bednja Municipality and Lepoglava City study area in Hrvatsko Zagorje region, NW Croatia. The landslides were interpreted from the high resolution (30 cm) digital elevation model (DEM) and its derivatives (slope and contour map, hillshade). The DEM was interpolated from the point cloud obtained by airborne laser scanning undertaken in spring 2020. In the study area of 20.22 km2, the total number of interpreted landslides is 912, making the average density of 45.1 ls/km2. The average size of the recorded landslides is 448 m2. According to the spatial plans, most of the studied area is covered by forests, agricultural areas, pastures, and artificial areas. The highest density of landslides is also in the forest areas, while the lowest is in the artificial areas. Furthermore, almost 64% of the mapped landslides are located within 50 m of the roads, and more than 39% of the mapped landslides are located within 100 m of the buildings and residential houses. Due to the level of detail provided and its completeness, the presented landside inventory map is an important tool for risk management at the local level because it gives detailed information necessary for risk evaluation as well as to decide about feasible options for risk mitigation, e.g., stabilisation measures vs relocation of the development to a more favourable location.

A landslide inventory presents a detailed register of the distribution and characteristics of past landslides in a specific area. Landslide inventory maps and other maps such as landslide susceptibility, hazard and risk maps present an essential tool in landslide risk management, supporting authorities, practitioners and decision-makers in the more appropriate and sustainable land planning and risk mitigation strategy development. In recent years, Light Detection and Ranging (LiDAR) data have been commonly used to map landslide morphology and estimate landslide activity. LiDAR is a consolidated remote sensing technique used to obtain digital representations of the topographic surface for areas ranging from a few hectares to thousands of square kilometers. From elevation point clouds obtained by laser scanning, a detailed digital elevation model (DEM) and different DEM derivatives, such as slope, hillshade or contour maps, can be produced. In this study, a historical landslide inventory map of the Hrvatsko Zagorje area (NW Croatia), interpreted from LiDAR high-resolution DEM (HRDEM) derivates, is presented and analyzed regarding geomorphological characteristics. The study area comprises 20.22 km2 of the hilly terrain (88% of the area has slope angles >5°), mostly covered by forests (52%). The area is composed of Triassic carbonate rocks, Miocene clastic sedimentary rocks and soils and Quaternary alluvial soils. LiDAR data for the study area was acquired in the framework of the project „Methodology development for landslide susceptibility assessment for land- use planning based on LiDAR technology (LandSlidePlan)“ financed by the Croatian Science Foundation. The topographic derivative datasets used to interpret the landslide morphology were hillshade maps, slope maps and contour lines. Landslide identification on the LiDAR HRDEM derivatives (0.3 m resolution) was manual and GIS- assisted, based on recognizing landslide features (e.g., concave main scarps, hummocky landslide bodies, and convex landslide toes). The mapping was performed at a large scale (1:100–1:500) to ensure the correct delineation of the landslide boundaries. Totally 912 landslides were mapped. The total area of mapped landslides is 0.408 km2 or 2.02% of the study area, and the mean landslide density is 45.1 slope failures per square kilometer. The average landslide area is 448 m2 (median = 173 m2). The small size of the landslides is probably the result of geological conditions (mainly Miocene marls covered with residual soils) and geomorphological conditions, where the differences between the valley bottoms and the top of the hills are rarely higher than 100 meters. Geomorphological characteristics of mapped landslides were compared with characteristics of the stable terrain using different DEM derivatives, such as roughness and curvature. According to the analyses, roughness and curvature values are distributed differently in landslides and stable terrain. Knowledge of the difference between the geomorphological characteristics of landladies and stable terrain provides valuable information for the automatic mapping of landslides and potently unstable slopes.

The objective of this study is a presentation of the landslide susceptibility assessment on a national scale for the Republic of Croatia using a heuristic approach. The effects of controlling factors on landslide susceptibility were estimated using the Fuzzy logic approach based on a multiclass overlay of landslide predictor maps. The predictor set relates to topographic variables, geomorphological factors, geology factors, and land cover. The selection of relevant landslide factors and the final landslide susceptibility assessment depends on subjective factors, such as researcher knowledge of the study area, respectively knowledge of different landslides types and processes in the study area, etc. For these reasons, it was necessary to verify the final landslide susceptibility map with data on known landslides. During the last few years, the scientists from the Faculty of Mining, Geology and Petroleum Engineering in Zagreb were systematically collected data on landslides, and the database of 2, 186 landslides with the exact location of the occurrence was conducted. The Area Under the Receiver Operating Characteristic Curve (AUROC) was used to validate derived landslide susceptibility maps and select the final one for further classification into three susceptibility zones. With over 90% of mapped landslides falling in high and very high susceptibility zones, the results are considered satisfactory for national scale landslide modeling. The landslide susceptibility map of Croatia was created to give a general overview of problem areas for an entire country, and it can be used to inform national policymakers and the general public. The analysis showed that approximately 20% of the area of Croatia is potentially prone to sliding. Particularly landslide-prone areas in Croatia are lowlands and hills in the Pannonian Basin, the hills of the Istrian Peninsula, and isolated narrow valleys in the Dinarides, such as Rječina River Valley and Vinodol Valley in Primorje.

This paper presents the application of Light Detection and Ranging (LiDAR) data for landslide identification and mapping in the pilot area at the Istria Penninsula (Croatia) and the analyses on the influence of expert knowledge on the quality of landslide inventory. Visual interpretation of landslides was carried out on high-resolution airborne laser scanning (ALS) LiDAR dataset. Scanning was taken in March 2020 for the pilot area in the City of Buzet. Based on the characteristics of the acquired LiDAR Point Cloud, a bare-earth digital elevation model (DEM) with 30 cm resolution was created. Different topographic derivative datasets such as slope, hillshade, contour lines, and roughness maps were created to interpret the LiDAR data. Eight experts with different levels of expert knowledge on LiDAR interpretation were given one week to carry out visual identification and mapping of potential landslides in the pilot area (0.3 km2) at a large scale (1:200) to provide detailed landslide mapping. Statistical analyses were performed based on the collected data to determine differences in the mapping accuracy and the number of recognized landslides by the experts. Results show that the experts familiar with the geology of the study area and potential landslide mechanics obtained better results than the experts who mapped landslides based on only ordinary topographic and geomorphological features specific for landslides.

The scientific research project Methodology development for landslide susceptibility assessment for land-use planning based on LiDAR technology (LandSlidePlan, HRZZ IP-2019-04-9900 ), funded by the Croatian Science Foundation, deals with new and under-investigated subjects in respect of inventory mapping of small and shallow landslides and presents innovative approaches to scientific research of landslide susceptibility assessment using cutting-edge LiDAR technology for collection of input data. The project has three main scientific goals. The first goal is to create the best optimal digital model of the bare-earth terrain that shows realistic landslide footprints and differences between disturbed and undisturbed land that may influence land use. The second goal is to create the most reliable large-scale landslide susceptibility map with the best differentiation of landslide-prone and non- susceptible areas using scientific methods customised to specific engineering geological conditions of Croatian environments with sliding threats. And the third goal is to create maps depicting information about landslides tailored according to the needs of the system of physical planning in Croatia (particularly land-use planning), encompassing local and regional levels under, harmonised at the national level. Due to different natural conditions and land uses in different parts of Croatia, the methodology will be developed for pilot areas in the City of Zagreb, Hrvatsko Zagorje and Istria, selected based on characteristic geological settings and degree of urbanisation.

Landslide maps produced by LiDAR (Light Detection and Ranging) are very clear and detailed representation of the phenomena and in many cases reveal evidence of past landslides that are virtually invisible by other detection techniques due to vegetation cover. Over the last decade, airborne laser scanning (ALS) has been made available and has been used to identify and map landslide. LiDAR elevation data prove particularly effective for mapping of small and shallow landslides in areas that are partially or completely covered by dense vegetation that are difficult or impossible to identify using conventional identification techniques. The first topic covered in the talk is landslide detection and mapping using very high-resolution LiDAR DTM to obtain complete historical inventories of shallow soil slides. Few examples of landslide inventory maps from different geomorphological environment of the Republic of Croatia will be presented to show typical landslide distribution. For landslide spatial prediction, the talk focuses on the results of landslide susceptibility modelling and zonation performed for a range of coverages, starting with the largest encompassing whole territory of Croatia. Landslide susceptibility zonation was also performed for areas of two counties (Primorsko-Goranska County and Karlovac County) and few smaller pilot areas of cities and municipalities in the Pannonian Basin and in External Dinarides. The main objective of presented susceptibility zoning in different scales (national, regional and local scale) is to enable analysis of usefulness and reliability of map information (i.e., spatial distributions and rating of the terrain units according to their landslide propensity) for application in physical and urban planning. The needs of the decision-makers, planners and other stakeholders involved in landslide risk prevention are analyzed through the series of round-table discussions organized in Croatia in the framework of the project of applied research PRI- MJER (KK.05.1.1.02.0020). The ultimate goal is to create maps depicting information about landslides tailored according to needs of the system of physical planning in Croatia (particularly land use planning), encompassing local and regional level, harmonized at the national level. The research of the mapping methodologies is part of the project “Methodology development for landslide susceptibility assessment for land-use planning based on LiDAR technology, LandSlidePlan” (HRZZ IP-2019-04-9900) fully supported by Croatian Science Foundation.

An innovative technique for detailed landslide inventory mapping is airborne laser scanning and LiDAR-derived DTMs in high resolution. LiDAR data used in this study was obtained in the framework of the “Methodology development for landslide susceptibility assessment for land use planning based on LiDAR technology (LandSlidePlan IP-2019- 04-9900)” project fully supported by the Croatian Science Foundation. To select the optimal digital terrain model (DTM) for landslide delineation, quantitative and qualitative assessments were done individually for three landslides. The quantitative assessment included a comparison of minimum, maximum, mean, and standard deviation values of DTMs derived by using four interpolation methods (Kriging, IDW, Natural Neighbor, and ANUDEM) in six raster resolutions (0.15, 0.3, 0.5, 1, 2, and 5 m). Furthermore, by comparing point cloud LiDAR data and interpolated DTMs elevation values, the mean-absolute-error difference (MAE) and root-mean-square-error (RMSE) were calculated. Hillshade, roughness, and curvature morphometric maps were derived for 24 DTMs per landslide, resulting in the qualitative assessment of 216 different morphometric maps. The quantitative assessment showed minimum and negligible differences between DTMs for landslide areas ; therefore, the qualitative assessment prioritised determining the optimal DTM for deriving morphometric maps needed for landslide delineation. Based on visual interpretability of landslide parts (i.e. crown, ridges, and toe) and the terrain quality (i.e. expressed details, irregularities, and blurriness) on the derived morphometric maps, the LiDAR DTM derived using the Kriging method in 0.3 m resolution was selected for landslide inventory mapping in further studies.

Cilj ovoga rada je bilo izraditi preliminarnu kartu podložnosti na klizanje u regionalnom mjerilu na temelju postojećeg inventara klizišta Grada Karlovca i dostupnih prostornih podataka primjenom bivarijantne statističke metode. Ukupno je analizirano 17 preduvjeta klizanja izvedenih iz digitalnog modela terena rezolucije 25 m, pokrova i namjene korištenja zemljišta Corine Land Cover, prometnica dostupnih na Open Street Map, te digitalizacijom Osnovne geološke karte M 1:100.000 i topografske karte M 1:25.0000. Za analizu utjecaja pojedinih klasa faktorskih karata klizanja i definiranje težinskih faktora primijenjene su dvije metode, metoda Weight of Evidence i metoda informacijske vrijednosti. Karte podložnosti na klizanje izvedene su za osam različitih scenarija, a verificirane su analizom ROC krivulje. Analiza je pokazala da najviši stupanj točnosti (96%) i predikcije (94%) ima karta podložnosti izvedena na temelju 14 faktora klizanja, uključujući nagib terena, orijentaciju padine, hrapavost i zakrivljenost terena, litologiju, udaljenost od geoloških granica i rasjeda, vlažnost terena, udaljenost od stalnih i povremenih vodotoka te izvora, udaljenost od prometnica i namjenu zemljišta. Konačna karta podložnosti na klizanje klasificirana je s obzirom na četiri klase podložnosti na klizanje, te je oko 60% područja klasificirano kao nisko podložno na klizanje, oko 16% je klasificirano kao srednje podložno na klizanje,oko 14% kao područja visoke podložnosti i 11% kao područja vrlo visoke podložnosti na klizanje. Iz daljnjih analiza raspodjele klasa podložnosti na klizanje s obzirom na namjenu klizišta i gustoću stanovništva može se zaključiti da postoji opravdani rizik od klizanja za stanovništvo, odnosno materijalna dobra na području Grada Karlovca. Iz navedenog se zaključuje da su nužna daljnja istraživanja koja će rezultirati detaljnijim i pouzdanijim kartama klizišta., The purpose of this thesis is to create a preliminary landslisde susceptiblity map on a regional scale based upon an existing landslide inventory from Karlovac City and available spatial data using bivariate statistics. In total, 17 sliding preconditions were created from a digital elevation model with 25 m spatial resolution, Corine Land Cover land use, traffic road data available on Open Street Map and by digitalization of General Geological Map M 1:100.000 and a topographic map M 1:25.000. For the analysis of the influence of certain classes from factor landslide maps and defining weight factors two methods were used, Weight of Evidence and Information Value. Susceptiblity maps were created by defining eight different scenarios and verified by ROC curve analysis. Analyses showed that the susceptibility map created from 14 landslide factors: slope gradient, aspect, roughness and curvature of the terrain, lithology, proximity from geological boundary and faults, terrain humidity, proximity from permanent and intermittent watercourses and springs, proximity from traffic roads and land use has maximum success rate (96%) and maximum prediction rate (94%). Final susceptibility map was classified in terms of four classes of landslide susceptibility; 60% of the territory was classified as low susceptibility, around 16% was classified as medium susceptibility, around 14% as areas of high susceptibility and 11% as areas of very high susceptibility. Further analyses of the division of susceptibility classes regarding land use and population density showed there is a justified risk of sliding for the population, and material goods in the Karlovac City area. From the above it can be concluded that further research is necessary, which would result in more detailed and reliable landslide maps.

Svrha ovog završnog rada je odrediti podrijetlo vode s izvora na području Ledine, u općini Preseka u Zagrebačkoj županiji. Laboratorijskom analizom uzoraka vode dobivene su vrijednosti stabilnih izotopa vodika 2H i kisika 18O koji su zatim uspoređivani s lokalnom linijom oborinske vode - LMWL Zagreb i Ljubljana te sa regionalnom linijom oborinske vode - RMWL Hrvatska i Slovenija. Podatci o LMWL i RMWL su dobiveni iz globalne mreže izotopa u precipitaciji - GNIP čiji se podatci koriste globalno u mnogim hidrološkim, hidrogeološkim te ekološkim istraživanjima i radovima. Na temelju interpretacije rezultata određeno je da je voda podrijetlom iz oborina, točnije s područja Ljubljane budući da dolazi do najboljeg podudaranja s LMWL Ljubljana.

Cilj ovog diplomskog rada bila je izrada modela podložnosti na klizanje za dio podsljemenske zone Grada Zagreba korištenjem bivarijantnih statističkih metoda analize podložnosti na klizanje, točnije metode Weight of Evidence i metode informacijske vrijednosti. Izrađeno je ukupno 18 modela podložnosti na klizanje prema predviđenim scenarijima. Scenariji se razlikuju s obzirom na prostornu rezoluciju korištenih faktorskih karata te prostornu rezoluciju i način geometrijskog prikaza klizišta iz inventara. Faktorske karte korištene za izradu modela podložnosti na klizanje bile su karte nagiba terena, karte raščlanjenosti terena, karte stratigrafskih jedinica, karte udaljenosti od geoloških granica, karte udaljenosti od drenažne mreže i karte namjene zemljišta prema tipu pokrova, prostornih rezolucija 1 m, 2 m i 5 m. U izradi modela podložnosti na klizanje korišteni su inventari klizišta prostornih rezolucija 1 m, 2 m i 5 m, pri čemu su klizišta bila prikazana poligonima, centroidima ili točkama na glavnim pukotinama. Nakon što su izrađeni svi modeli podložnosti na klizanje pristupilo se određivanju stupnja točnosti i predikcije pojedinih modela pomoću analize ROC krivuljom. Usporedbom rezultata analize ROC krivuljom utvrđeno je da svi modeli podložnosti na klizanje daju vrlo dobre rezultate, ali model podložnosti na klizanje prema scenariju S12 daje kombinaciju najviših stupnjeva točnosti (92%) i predikcije (92%) područja istraživanja. Navedeni model dobiven je primjenom metode informacijske vrijednosti i korištenjem faktorskih karta prostorne rezolucije 1 m te inventara klizišta iste prostorne rezolucije u kojem su klizišta prikazana točkama na glavnim pukotinama., The objective of this thesis was to create landslide susceptibility models for the Podsljeme area of the City of Zagreb using bivariate statistical analysis methods, specifically the Weight of Evidence method and Information Value method. A total of 18 landslide susceptibility models were created according to the predetermined scenarios. Each scenario is defined by the spatial resolution of the used factor maps and the spatial resolution and type of geometric representation of the landslides from the inventory map. The factor maps used to develop the landslide susceptibility models were in three different resolutions (1 m, 2 m and 5 m) as follows: slope maps, terrain breakdown, stratigraphic units, distance from geological contacts, distance from the drainage network and land use according to cover type. Landslide inventories with spatial resolutions of 1 m, 2 m and 5 m were used in the creation of the landslide susceptibility models, in which the landslides were represented either by polygons, centroids or points on the main cracks. After all the landslide susceptibility models were created, the levels of accuracy and prediction of individual models were determined using the ROC curve analysis. By comparing the results of the ROC curve analysis, it was determined that all models of landslide susceptibility give very good results, but the model of landslide susceptibility according to scenario S12 gives a combination of the highest degrees of accuracy (92%) and prediction (92%) of the research area. The mentioned model was obtained by applying the Information Value method and using factor maps with a spatial resolution of 1 m and landslide inventory of the same spatial resolution, in which landslides are represented by points on the main cracks.

U ovom radu prikazuje se postupak izrade karte podložnosti na klizanje na području Grada Slatine primjenom bivarijatne statističke metode. Inventar klizišta koji je korišten za procjenu podložnosti na klizanje preuzet je iz Urbanističkog plana uređenja Grada Slatine iz 2003. godine. Analizirano je ukupno devet preduvjeta klizanja izvedenih iz dostupnih izvora podataka: Osnovna geološke karta Hrvatske M 1:100 000, Hrvatske osnovne karte (HOK) M 1:5 000, ortofoto snimaka područja i iz Prostornog plana uređenja Grada Slatine. Digitalni model terena rezolucije 5 m izrađen je u okviru ovog rada digitalizacijom podataka s HOK-a, a iz njega su izvedene najvažnije faktorske karte morfoloških i hidroloških parametara. Za analizu utjecaja pojedinih faktorskih karata i klasa primijenjen je χ2 test, te dvije metode, metoda informacijske vrijednosti i metoda Weight of Evidence. Karte podložnosti na klizanje izvedene su za četiri scenarija s odabirom različitih kombinacija faktorskih karata. Na sve karte su primijenjene četiri vrste klasifikacije. Za konačnu kartu odabrana je karta koja sadrži morfološke, hidrološke i antropogene faktore, a izrađena je primjenom metode informacijske vrijednosti klasificirane natural breaks metodom. Konačna karta podložnosti na klizanje klasificirana je na tri klase podložnosti na klizanje (niska, srednja i visoka), a verificirana je sa zonama podložnosti iz 2003. definiranim na temelju izravnog kartiranja. Usprkos opisanim ograničenjima, pouzdanost konačne karte podložnosti na klizanje je visoka. Ova karta se može primijeniti kao tematska podloga za prostorno planiranje jer prikazuje zone u kojima je potrebno definirati posebne uvjete za gradnju u odnosu na klizišta., In the thesis the procedure of development of a landslide susceptibility map of the Slatina City using the bivariate statistical method was presented. The landslide inventory map used for landslide susceptiblity assessment was taken from the Urban Development Plan of the Slatina City from 2003. A total of nine landslide precondition maps were analysed. All are created from the following available data sources: Basic Geological Map in the scale 1:100 000, Croatian Basic Topographic Map in the scale 1:5 000, orthophoto images of the area and Spatial plan of the Slatina City. Digital elevation model of 5-m spatial resolution was created in the framework of the thesis by digitalizating all elevation points and isohypses. It was used as one of the most important input data set for morphological and hydrological parameters.. For the analysis of the influence of individual factor maps and classes, the χ2 test was applied, as well as two methods, the information value method and the Weight of Evidence method. Landslide susceptibility maps were created for four scenarios with a different combinations of factor maps. Four types of landslide susceptiblity classification were applied to all maps. For the final map, it was selected a map derived based on analysis of morphological, hydrological and anthropogenic factorsusing the information value method, classified with the natural breaks method. The final landslide susceptibility map was classified into three landslide susceptibility classes (low, medium, and high landslide susceptibility) and it was verified by comparison with susceptiblity zones defined in 2003 by direct mapping. Despite described limitations, realiability of the final landslide suceptibliy map is high. The created map can be used as a thematic base map for spatial planing because it depict zones that posses special conditions for construction in relation to landslides.