Izvorni znanstveni članak
https://doi.org/10.37023/ee.12.1-2.9
REMOTE SENSING AND QGIS FOR REGIONAL LANDSLIDE RISK MAPPING: A CASE STUDY OF VARAŽDIN COUNTY, CROATIA
Anja Bek
; Geotehnički fakultet Sveučilišta u Zagrebu, Varaždin, Hrvatska
*
Filip Dodigović
; Geotehnički fakultet Sveučilišta u Zagrebu, Varaždin, Hrvatska
Krešo Ivandić
; Geotehnički fakultet Sveučilišta u Zagrebu, Varaždin, Hrvatska
Jasmin Jug
; Geotehnički fakultet Sveučilišta u Zagrebu, Varaždin, Hrvatska
* Dopisni autor.
Sažetak
Varaždin County is underlain by Neogene and Quaternary clays and marls that are prone to landslides triggered by
intense rainfall. In this study, a landslide risk assessment was carried out for nine municipalities during 2023–2025 using
open-source QGIS software and Sentinel-2 imagery. The analysis integrated static factors (slope, aspect, land use) with the
dynamic factor NDVI (Normalized Difference Vegetation Index) through Multi-Criteria Decision Analysis (MCDA) and a
weighted linear combination. Slope was assigned the highest weight (40%), while NDVI contributed 20%. The results reveal
clear spatial contrasts: mountainous municipalities (Donja Voća, Klenovnik, Bednja, Lepoglava, Ivanec) show high to very
high landslide risk, whereas lowland municipalities along the Drava River (Cestica, Petrijanec, Maruševec) remain largely
stable. Temporal differences are small, with 2023 showing the highest landslide risk and 2024 the lowest. NDVI-driven
fluctuations reflect vegetation dynamics but are less influential than topography. Overall, the findings demonstrate the value
of open data and free GIS tools for regional hazard mapping. The produced maps provide useful guidance for spatial planning
and disaster risk management, while also pointing to the potential of integrating additional datasets in future studies to further
improve the robustness of the approach.
Ključne riječi
remote sensing; QGIS; landslide risk; slope; aspect; land use; NDVI
Hrčak ID:
341612
URI
Datum izdavanja:
18.12.2025.
Posjeta: 180 *