Artificial levees along alluvial rivers are major components of flood risk mitigation. This is especially true in the case of Hungary, where more than one-third of the country is threatened by floods and protected by an over 2940-km-long levee system. Most of the levees were built in the 19th century. Since then, several natural and anthropogenic processes, such as compaction, erosion, Etc., could contribute to these earth structures' slow but steady deformation. Meanwhile, as construction works were scarcely documented, the structure and composition of artificial levees are not well known. Therefore, the present analysis aimed to use different geophysical techniques to validate their efficiency in mapping structural differences, possible compositional deficiencies, potential defects and sections where elevation decrease and compare the compositional and structural variations of two very different levee sections along a 24 km section of the River Tisza and a 24 km section of the River Maros. Investigations were conducted by real-time kinematic GPS (RTK-GPS), Ground penetrating radar (GPR), Electrical Resistivity Tomography (ERT) and drillings. Onsite data acquisition was complemented with an analysis using a Persistent Scatterer Synthetic Aperture Radar (PSI) to assess general surface deformation. The higher frequency 200 MHz GPR data have shown that levee structures can significantly vary even in a few km on sections with the same construction history. Based on electrical resistivity tomography results with a precise analysis of grain size and their related physical parameters used for monitoring the materials of two different levee sections along the Tisza and Maros rivers, we noticed that the main components of investigated Tisza levee section are medium and fine silts, however, the situation of the investigated Maros levee section shows more variation of different materials which are fine, medium, and coarse silt, moreover, fine, medium, and coarse sand. The investigated section of the Tisza levee showed low resistivity values, indicating the fine-grained materials' conductivity. In contrast, the investigated section of the Maros levee showed high resistivity values, indicating the resistivity nature of higher grain size sediments forming this section, especially noticed on the protected side of the levee. It was possible to capture structural changes and resolving the thin layers by 1 m electrode spacing ERT profile. In turn, at a larger spacing it was possible to get information on the sedimentary base below the levee body. The selected levee section could be assessed in terms of its structure and composition and major units within the levee body and their composition could be resolved by the applied methods. In general, there is a similarity in the materials and their resistivity range which form the core of Tisza and Maros levees, however, the situation on their both sides is not the same. Regarding the analysis of different physical properties of the two levee systems like resistivity, porosity, density, water content, grain size, and saturated hydraulic conductivity, the materials of the Maros levee could be distinguished well and showed more variation when it is compared to the materials of Tisza levee. It means that the physical properties of levee materials are very important, and they are recommended when carrying out further levee investigations. From the physical properties mentioned above, it was found that some of them show a connection with resistivity except hydraulic conductivity parameter that did not show a direct connection, however the latter could exhibit the aquitard nature of Tisza levee materials and the non-aquitard nature of Maros levee materials which illustrates the difference in levee composition in terms of flood risk or flood safety. Based on height measurements, the mean elevation of the levee crown decreased by 8 cm in a 40-year time span. However, elevation decrease could reach up to 30 cm at some locations. Sections affected by structural anomalies, compositional changes, and increased surface subsidence are especially sensitive to floods when measurement results are compared to flood phenomena archives. GPR profiles showed several anomalies, including structural and compositional discontinuities and local features. They were classified into six types regarding to the flood risk; tensile cracks (enables piping, leading to levee breach or mass failure, cracks might close when the levee gets wet), remarkable changes in dielectric permittivity (enables seepage, leading to mass failure), animal burrows (enables piping, leading to levee breach or mass failure), layer deformation (results in height decrease, overtopping), paleo river channel (enables seepage below the levee, leading to water upwelling and the development of sand boils), sudden change in stratification or dipping layers (enables contour line seepage, leading to mass failure).
