1 Introduction

The study area is approximately 55 km east of Bursa province, and the examined area is ~ 1024 hectares. This area is the district centre of Yenişehir (Bursa) in the middle part of the Yenişehir plain, close to the east, which consists of 52 existing map sheets with a scale of 1/1000 and 8 map sheets with a scale of 1/5000 (Fig. 1).

Fig. 1
figure 1

Study area and map sheets covering this area (1/29790) (MTA-H23 2014)

Full size image

As is known, it is important to define the physical conditions of urban and other building grounds and to design engineering constructions under these conditions in active earthquake zones. With developments in science and technology, the planning process of the environment in which people live nowadays is completed by the cooperation of many disciplines. Geophysical data are important in the planning of constructions and infrastructures for the effective use of urban and extra-urban areas, in the process of making and using them in accordance with the plan. Therefore, geotechnical studies should be completed with the results of geological–geophysical studies in urban planning. Thus, while geophysical studies make scientific contributions to regulating the interaction of super-infrastructure with the ground, they also affect the formation of healthy-safe urban development and a decrease in costs. When geophysical studies carried out to this end are examined, it is observed that more than one geophysical method is applied together in field studies. For example, different applications of geophysical seismic methods make use of S-wave velocity information in determining the dynamic-elastic properties of the ground. Many urban engineering problems are solved using these methods (Zhao et al. 1992; Goes et al. 2000; Sato and Fujimoto 2009; Kanbur and Kanbur 2009; Uyanık 2019; Uyanık 2020; Karslı 2018). Engineering studies are applied within the city and sometimes in restricted areas. In this case, it is important to obtain a sufficient spread length in the seismic refraction method to generate the seismic S-wave and reach the targeted exploration depth. Therefore, the seismic tomography method, which is a common application of seismic methods nowadays, is preferred. This method is one of the important methods, despite some difficulties in implementation. However, it has been observed in many studies that the use of surface waves provides advantages to overcome problems (Nazarian and Stoke 1985; Gucunski and Woods 1991; Socco and Stobbia 2004; Tokimatsu et al. 1992; Acarel 2003). Especially its contributions to seismicity, liquefaction, and risk problems in urban area grounds are important (Uyanık, 2019; Uyanık and Taktak 2009; Karslı 2018; Özçep 2018; Uyanık 2020). In ground investigations using the seismic surface wave analysis method, active source method applications are preferred as the source type. The most important reason for this is to eliminate the problems that can be caused by multiple signals since the place of the source is known. Moreover, data collection-processing stages are faster and easier (Xia et al. 2006). The spectral analysis method of surface waves is an example of active-source surface wave studies, adapted to the engineering use of the two-station method in seismology (Bergstrom 1999).

The active-source seismic MASW (Multi-channel Analysis of Surface Wave) method, which performs S-wave velocity analysis by using the dispersion properties of the Rayleigh surface wave, was used in the study area. There are many studies indicating that this method is successful in solving engineering problems (Richard 1977; Conyers and Goldman 1997; Cardimona et al. 1998; Cardimona 2002; Fernando and Afonso 2005; Dikmen et al. 2009; Cassidy 2009; Homan et al. 2010; Hamdan et al. 2010; Karslı 2018; Özçep 2018). The duration current resistivity (DCR) method has a duration current source and determines the resistivity and stratification of the ground. It is one of the most common geophysical methods, providing the ease of application, effective results, and enabling the evaluation of 1D (one-dimension) data collected in the field as 2D (two-dimension) and 3D (three-dimension) when necessary (Van Nostrand and Cook 1966; Zohdy 1974a, b, c; Başokur 1984, 2003, 2007, 2010; Telford et al. 1990; Candansayar and Başokur 2001; Dahlin et al. 2002; Dahlin and Zhou 2004). Furthermore, in the DCR method, 2D and 3D measurements can be collected directly by selecting the appropriate devices and equipment, and 1D measurements can also be derived from measurements of this dimension when necessary. According to the selection standards (ASTM D6422-99) of surface geophysics research methods of ASTM (American Society for Testing and Materials), the DCR method is a suitable and common geophysical method to examine the near-surface properties of the underground (ASTM D6429-99 1999). Therefore, vertical discontinuities of geological units can be investigated using the vertical electrical sounding (VES) technique of the DCR method. The apparent resistivity values ​​calculated from VES data can be evaluated with inversion, and the depth, thickness, and true resistivity values ​​of layers can be determined (Başokur 1984, 2003, 2010; Ardalı 2005). Thus, fractured, hollow regions, discontinuities and deterioration areas, geological stratification, high/low resistivity areas/roads, and groundwater level can be determined. Furthermore, horizontal changes can be observed and interpreted by evaluating the results of multiple VES points together with gridding or profiling techniques or directly examining the 1D inversion results. With these arrangements, low seismic velocity zones and seismic stratification observed in seismic results, resistant/less resistant areas observed in the electrical resistivity results, and resistivity stratification can be compared, and a better ground definition can be made. This is because the groundwater level in the environment is associated with decreasing Vs values, increasing conductivity or, conversely, decreasing resistivity values. This association is also used to distinguish environments with low and high porosity. In the MASW method application, profile length may need to be restricted due to a lot different environmental conditions. For example, there may be geophone number-device model, geophone spacing, topographic-geomorphological conditions, proximity to a residential area, traffic, etc., conditions. In this case, long profiles cannot be selected to obtain more depth information. The study area is also a region where agricultural activities are carried out. In this study, there were difficulties in taking measurements in an area where active agriculture was carried out. Therefore, a seismic profile length that determines the depth of the water-holding formation or the bedrock depth to a deeper level could not be selected in any area. However, with the VES method, changes in formation thickness and groundwater levels can be examined at a deeper research depth. This is due to differences in the application of the method. Therefore, the VES method is often preferred as the second method in such field studies. In this study, level maps were compared and interpreted for the same depth. Moreover, within the 50 m research depth, even though the bedrock depth could not be observed in the south of the study area, it was observed that the bedrock was deeper in level maps in the north (Bayramoğlu 2020). The changing in the formation boundary towards depth can be observed on the level maps. Therefore, the contributions of VES results to seismic and seismological interpretations are important.

As a result, geophysical studies for the purpose of the near-surface investigation were planned, and geophysical data were collected in an area of ​​approximately 1024 hectares of Yenişehir (Bursa) district, especially limited by the N-NW-SW side. Thus, earthquake, liquefaction, risk and hazard problems of the ground-construction interaction could be interpreted with level maps. To this end, geophysical data were collected using the seismic MASW method, one of the tomography applications of the seismic method, and the VES measurement technique, which can obtain 1D vertical changes in the DCR method as a drilling measurement. Geophysical data were evaluated using the appropriate software, and the geophysical properties of grounds were defined because a building site or ground investigation is a continuous research and interpretation process (Özçep 2018). Thus, risks/hazards against possible natural-artificial events on the grounds of the study area and its surroundings, in the existing buildings, in the constructions to be made and in its immediate surroundings were revealed. From the results of these studies, predictive results for urban areas and other building grounds were also prepared. The reason for this is to minimize the loss of life and property in possible natural-artificial events, to take the necessary precautions in advance, to prepare healthy-planned living-urban environments, and to reveal the study area in disaster-risk-prediction dimensions.

2 Geology

There are two geological units in the study area. The first one of these is the alluvial unit consisting of the Quaternary-aged gravel–sand–silt–clay mixture. This unit is observed throughout the study area. The second unit is the unit containing Neogene-aged conglomerate–sandstone–claystone. This unit outcrops mostly in the southern and northern borders of the Yenişehir basin. Accordingly, the study area generally consists of the Late Miocene-Pliocene aged Mesudiye formation (Tmpl) and Quaternary-aged alluvial (Qal) cover units (MTA-H23 2014) (Fig. 2).

Fig. 2
figure 2

The geology map of the study area (MTA-H23 2014)

Full size image

The Mesudiye formation starts with conglomerate and sandstone around the Mudanya district and continues with clayey limestones with lacustrine marl and claystone intercalations (Pehlivan et al. 2011). In the study area, similar rock assemblages, which are the lateral continuation of these units, were addressed under the same name. They spread over a wide area in the north of the Yenişehir plain. This unit, consisting of yellowish, grey, greyish, red, brown, medium-thick bedded, poorly sorted, polygenic pebble conglomerate, sandstone, also includes claystone–marl–clayey limestone intermediate levels. Limestones are white-yellow coloured and thin-medium bedded. Cross-bedding and channel fill structures are common in conglomerate–sandstone predominant levels. The Mesudiye formation overlies the basic units with an angular unconformity. The Çamlık formation, consisting of lacustrine units gradually transitive with it in horizontal and vertical directions, comes on the unit (MTA-H23 2014). According to Varol et al. (1997) and Pehlivan et al. (2011), the Mesudiye formation represents the river–lake environment in the Late Miocene-Pliocene. Alluviums are sediments consisting of gravel, sand, silt, and mud that developed in river beds, valley floors, and plains. These units, which spread widely along the Yenişehir plain, are composed of unbonded and in some places slightly bonded pebble, sand, silt, and mud that fill the plain. The unit, which unconformably overlies its predecessors, forms a thick river sedimentary sequence (MTA-H23 2014). Therefore, in this study, geophysical data were collected from alluviums and the Mesudiye formation by selecting appropriate geophysical methods for ground investigations. By analysing these data, the seismicity of the study area, geological layering according to seismic and electrical results and geophysical properties of these layers were determined. Taking into account the changes in the groundwater level determined according to electrical resistivity, the geological and geophysical results were examined together and comments were made. Thus, geological boundaries suitable and risky for construction and some suggestions were presented.

3 Analysis of earthquake data

If the earthquake records in the Kandilli Observatory and Earthquake Research Institute (KOERI) from 1900 to the present day are examined, it can be observed that the faults in the Marmara Region produce very large earthquakes. The most important fault affecting the study area is the North Anatolian Fault (NAF). This fault is one of the fastest moving and most active right-sided strike–slip faults in the world. Therefore, Yenişehir and its vicinity are tectonically in an active area where strike–slip faults are active, and they are controlled by right-lateral slip fault systems. In the region with high seismic activity, a large number of M ≥ 5 (M: Magnitude) destructive earthquakes were recorded, and the return periods are approximately 180–250 years (Doyuran et al. 2000). Selim et al. (2006) examined the seismological and geological properties of the instrumental period earthquakes in the study area. Accordingly, they determined that damaging and destructive earthquakes occurred on strike–slip and dip–slip faults, that the Yenişehir basin was approximately NW-striking, showed an ellipsoid-like geometric structure and was ~ 35 km long and 12–13 km wide. The Yenice–Gönen fault, one of the faults affecting the study area, is generally under the influence of a compression regime (Selim 2004). The other faults are the Manyas Fault, Uluabat Fault, and Edincik Fault, and these faults are still active in terms of generating earthquakes (Selim et al. 2006). The largest and most damaging earthquake in the investigation circle is the M = 7.4 earthquake dated 17.08.1999. It occurred ~ 70 km northeast of the study area (KOERI 2020). This earthquake was particularly effective in settlements near the southern coasts of the Marmara Sea, and surface deformations and groundwater outflows were observed in some places.

When earthquakes within the active earthquake zone borders of Turkey are examined, the risk class district of the earthquake hazard of Yenişehir (Bursa) district is between 0.2 and 0.3 g acceleration values (AFAD 2018). In this study, calculations were made using M ≥ 4.5 earthquake data that occurred in the area with a radius of 100 km with the centre in Yenişehir (Bursa) between 1900 and 2020 from the UDIM of the KOERI (UDIM 2020). The local magnitude calculated using these data is ML, and the magnitudes of earthquakes occurring in the instrumental period in the area with a radius of ​​100 km increase up to 7.4. In determining the magnitude–frequency relationship of this area, the environment constants a and b were calculated using the least-squares method. Earthquakes with magnitudes (ML) varying between 4.5 and 9 were selected and examined (Fig. 3). The foci and depths of these earthquakes are processed on the map in Fig. 3. Accordingly, the number of large earthquakes that will adversely affect the study area is quite high. Parallel to the magnitude of the earthquake, it is observed that both shallow and deeper earthquakes are effective in and around the study area.

Fig. 3
figure 3

M ≥ 4 earthquakes within a 100 km radius with the centre in Yenişehir between 1900 and 2020 (UDIM 2020)

Full size image

The earthquake risk analysis of the KOERI (2020) earthquake data was performed by the Probabilistic Earthquake Hazard Analysis (PEHA) and Poisson Probability Distribution (PPD) using the Excel-based Ground Geophysical Analysis© programme prepared by Özçep (2010) (Bayramoğlu 2020). Accordingly, earthquakes with the magnitude (ML) varying between 4.5 and 7.5 for a total of 120 years between 1900 and 2020 were classified, and their occurrence numbers were entered as data. The correlation coefficient R (Regression) between these input data and M-logN (N is the number) was calculated and shown in a statistical graph (Fig. 4). If the graph in Fig. 4 is examined, M-logN values ​​show a linear relationship inversely proportional to each other. According to this inverse proportion, it was determined that the greater the magnitude of the earthquake was, the lower the probability of the earthquake (linear) was in the area with a radius of 100 km, including the study area.

Fig. 4
figure 4

M-LogN relationship graph

Full size image

At the final stage, other earthquake parameters were also calculated using the results of the M-LogN graph. By distributing the result values ​​in Fig. 4 according to POD, the earthquakes that may occur, the magnitudes of these earthquakes (M), and the average recurrence periods (ARP) were calculated (Fig. 5 and Table 1). According to the ARP results in Fig. 5, as the magnitude of the earthquake increases, the probability of a larger earthquake occurs with a longer recurrence year. Furthermore, while the frequency of recurrence of earthquakes with M ≤ 5.5 within 10–50–75–100 years is higher, the frequency of recurrence of earthquakes with M ≥ 6 decreases (Table 1). In Table 1, it can be observed that the probability of earthquakes in 10, 50, 75, and 100 years for each earthquake with a magnitude of 6, 6.5, 7, and 7.5 and the same results are obtained from the ARP values corresponding to these earthquakes.

Fig. 5
figure 5

PEHA-PPD results calculated from the earthquake data

Full size image
Table 1 Magnitude and probability
Full size table

The ground motion acceleration (amax, g), which indicates the earthquake hazard and risk in a region, is one of the most important parameters of the earthquake effect in the design of earthquake-compatible buildings and in the calculation of the level of ground shaking that will affect buildings. M alone is insufficient in determining the earthquake hazard in a region and projecting earthquake-resistant buildings. The attenuation relationships of the ground motion, providing the largest acceleration (amax) value an earthquake will create at any point on the earth, are also required. Acceleration–distance attenuation relationships depend on the distance of the earthquake source to the area to be examined (Ulutaş et al. 2003). In the AFAD (2018) regulation, the acceleration of the study area in the Earthquake Hazard Map of Turkey ranges from 0.2 to 0.3 g. These accelerations should be followed in building. In this study, the PGA calculation was made from the acceleration–distance attenuation relationships. Accordingly, medium and large-scale earthquakes for a total of 120 years were considered as point sources, active fractures around the field were considered as linear sources, and the PGA values ​​that these earthquakes could cause in the study area were calculated (Fig. 6). Upon examining Fig. 6, the UDIM (2020) data were evaluated using the Ground Geophysical Analysis© programme, and the distance to the closest and active seismic source that was thought to affect the study area was calculated. To this end, the largest earthquake magnitude or the design earthquake magnitude (application example earthquake magnitude) determined for the intended engineering target on the site within the source zone (formed from the past to the present day) within the study area was selected. The possible PGA (amax) value for the study area was calculated by placing the determined input value in the decay equations. As the input data to calculate amax, the largest earthquake observed in UDIM is M = 7.4. This M value belongs to the 17 August 1999 Kocaeli-Başiskele earthquake, which was registered in UDIM with the earthquake code 19990817000137. Therefore, this earthquake record was used for the reference acceleration value in this study. The epicentre distance of this earthquake is ∆ = 25 km, and the focus depth is H = 15 km. The amax values ​​for the closest and active seismic source of an earthquake of this magnitude, which was thought to affect the study area, were also calculated using the formula of different researchers (Fig. 6). Accordingly, the four amax values calculated and the average of these values, average amax = 0.31 g, are presented in Fig. 6. Liquefaction and risk interpretation was made with the value of amax = 0.31 g. According to the ESC (European Seismological Commission) standards, the hazard level was determined to be in the high hazard class because the acceleration value was greater than 0.24 g.

Fig. 6
figure 6

PGA calculation and graph according to various formula

Full size image

As a result, it was determined that the groundwater level in the study area changed from about 5–6 m in alluvial units in the south to a depth of approximately 90 m in the Mesudiye formation in the north with a regular increase towards the north. Since the calculated average amax value was 0.31 g and it is known that alluvium consists of fine gravel sands, the risk class of liquefaction, especially in the south of the study area, was determined to be quite high. Therefore, areas with alluvial units in the south are the areas with the highest risk in the high hazard class. Additionally, the topography rises suddenly towards the north and the formation type also changes. Therefore, the decrease in the groundwater level towards the north and deeper is also evidence of the increase in Vs, Vp and ρ values in the northern area. This is showed by the fact that the bedrock depth starts from an average of 5 m and continues to > 10 m depth. In the alluvial area, it was observed that the bedrock was deeper. Therefore, in this study, despite the value of 0.31 g, it was interpreted that there were better ground conditions in the northern area.

4 Geophysical methods

4.1 Multi-channel analysis of surface wave (MASW) method

Nowadays, ground investigations are carried out according to the phase velocity of surface waves propagating at the ground–air interface and the characteristics of dispersion curves (Aki and Richards 1980; Shearer 1999). Seismic S-wave velocities (Vs, m/sec) can be calculated by the MASW method using the dispersion properties of the Rayleigh wave, one of the surface wave types. In the method, the source can be controlled, and the data collection-processing stages are fast and easy (Xia et al. 1999). Seismic data are collected with multiple (12 or more) receivers (geophones) placed on a linear line (Fig. 7). The geophone frequency is 4.5 Hz, and they are successfully recorded the surface waves (Park et al. 1999). The MASW data acquisition array is showed (Fig. 7). The MASW equipment is similar to the SRT (Seismic Refraction Tomography) method. The geophones are 14 Hz for the SRT method. Geophone spacing determines the resolution associated with the smallest wavelength (Dikmen et al. 2009). In the seismic study, shots in the offsets only at the beginning and end of the profile were made because of the large study area and the large number of profiles since the main purpose is to calculate the dynamic parameters of the study area ground. In addition, other reasons include the fact that the study area is in an agricultural area and the traffic of agricultural vehicles and people continues throughout the day.

Fig. 7
figure 7

The MASW data collection setup of the study area (rearranged from Özel and Darıcı (2020))

Full size image

As is known, the raw data reaching the seismic acquisition system appear inclined (Fig. 8a). In the example in Fig. 8, for the seismic MASW1 record of the study area, the depth-Vs values ​​of the layers were calculated by performing the inversion process after selecting the dispersion curve (Fig. 8b–c–d). The inversion process performed using a certain initial model was overlapped with the error rates of less than 5%, and the data processing stage was terminated, and 1D depth-Vs graphs were obtained (Fig. 8d). These seismic data processing stages were performed for the inversion processes of all MASW data, and the seismic data evaluation process was completed. However, although the error rates in geophysical level maps have increased to approximately 8.22%, it has been determined that this rate changes to around 6% on average when all level maps are taken into account (e.g. 6% on the Vs level map).

Fig. 8
figure 8

For the MASW1 sample a Raw data. b Dispersion curve. c Frequency–velocity relationship inversion. d 1B depth-Vs graph obtained from the inversion process

Full size image

There are different numerical techniques that evaluate seismic field effects in sedimentary basins, and researchers use them selectively according to the purpose in examining ground behaviour (Luzón et al. 2002). Thus, if the most appropriate algorithm for the seismic data to be processed is selected, the solution is approached and the inverse solution is completed successfully (Luzón et al. 2002; Wathelet 2008). As a result, the inversion of the dispersion curve of the Rayleigh wave phase velocity–frequency pairs is calculated by the iterative inversion of the dispersion curve, which requires the estimation of a certain density and Poisson's ratio values. This process is performed with the least-square (LS) approach. In the LS approach, while density and Poisson's ratio remain constant in each iteration step, Vs information converges reliably. The inversion process is terminated when the error rate is minimum (Park et al. 1999). Thus, the 1D data evaluation of the study area MASW data was completed, and 1D depth-Vs graphs were obtained. These graphs were combined and evaluated in the SeisImager software, and 2D depth-Vs sections were also obtained (Bayramoğlu 2020). For this purpose, layer thickness and Vs values were obtained by selecting them from the dispersion curve of the Rayleigh wave in the calculation. According to the 0.9–1.1 times difference between Rayleigh wave velocity and Vs, the reference point was accepted as 1 to 1 and the data were processed by selecting 10 layers.

At the final stage, seismic P-wave velocities (Vp, m/sec) by performing 1D analysis of seismic records obtained from the SRT method were also calculated using the SeisImager software, and 2D depth-Vp sections were also obtained (Bayramoğlu 2020). SRT measurements were obtained simultaneously on MASW profiles. Consequently, with the calculated Vp and Vs values, the values ​​of the dynamic parameters of the study area ground were calculated. The results obtained were transformed into level maps for three different depths using the Kriging method in the Surfer2013 software and made ready for 3D interpretation. In this software, after preparing Vp, Vs, G, σ, Ak depth level maps and a level map for Vs30, seismic data evaluation was completed.

4.2 Vertical electrical sounding (VES) method

The DCR method, one of the electrical methods, is based on measuring the potential difference (ΔV, Volt) that the current (I, Ampere) sent into the ground with two electrodes from the surface will create in the other two electrodes by taking advantage of the different electrical properties of the underground geological structures. With these potential differences, it is aimed to determine the parameters corresponding to the geological structure. To this end, apparent resistivities are calculated from the potential differences measured in the field. These data are evaluated with an inversion process in suitable software, and the thickness, depth, and real resistivity values ​​of geological layers are calculated. One of the electrical methods used for this purpose is the application of the VES technique, one of the 1D measurement techniques in the DCR method. The VES technique is a common method because it is easy to apply in the field, its devices and equipment are economical, the data can be evaluated quickly, and the knowledge and experience accumulation is high (Van Nostrand and Cook 1966; Zohdy 1974a, b, c; Telford et al. 1990; Candansayar and Başokur 2001). In practice, a power source (battery), a current meter, a voltage difference meter, and 4 electrodes are sufficient for the measuring setup. Upon examining Fig. 9a, the current is applied to the ground with the A and B electrodes fixed at two points on the earth, and the voltage difference between the M and N electrodes fixed at the other two points is measured. 1D electrical VES data were collected by selecting the Schlumberger array to examine the vertical changes at the planned VES points in the study area (Fig. 9a). Apparent resistivity (ρa, Ohm m)–distance (ab/2, m) curves were prepared by processing the 1D apparent resistivity values ​​calculated from the measured potential differences logarithmically (Van Nostrand and Cook 1966). The apparent resistivity general equation in Fig. 9b was used in calculating the apparent resistivities. In this equation, I refers to the current (Ampere), ΔV refers to the voltage difference (Volt), and K refers to the geometric factor. Then, 1D depth (m)-resistivity (ρ, Ohm m) graphs were obtained by the inversion of the ρa curves in the IPI2WIN + IP programme. Thus, geological layering and the depth-resistivity values of these layers were calculated separately for each VES point.

Fig. 9
figure 9

For the Schlumberger array a VES measurement setup (rearranged from Özel and Darıcı (2020)). b apparent resistivity calculation general equation

Full size image

At the final stage, the results of 1D depth-resistivity graphs were combined in the Surfer2013 programme and transformed into resistivity level maps for three different depths to be used in 3D interpretation. Consequently, 1D inversion of VES data was mostly completed with rates of approximately 5–6%. These errors increased to approximately 9% when determining level maps. This may be related to the slightly greater distance between some VES points. Therefore, this error rate can be interpreted as acceptable since it is below 10%. Moreover, a groundwater level map was prepared from the VES data, and comments were made by associating the Vs30 results with other calculated seismic parameters and geological units.

5 Geophysical study and data processing

5.1 MASW and VES studies

In this study, more than one geophysical method was selected in field applications to define the study area, and the results obtained were interpreted together. At the first stage in the field, MASW data and then VES data were collected and evaluated. The results of these two methods were compared with each other and with the analysis results of earthquake data. In MASW measurements, a SARA-DOREMI engineering seismograph with 12-channel signal accumulation was used (Fig. 10a). In MASW applications, according to the measurement setup in Fig. 11, the geophone spacing of 3 m and the offset interval of 9 m were selected. The power source is a 12 V battery, and the sledgehammer method was selected to generate a seismic signal (Fig. 10b–c). To this end, a 9 kg sledgehammer and 4.5 Hz natural frequency vertical geophones were used as a receiver. In the records, the seismic MASW data were recorded in the SEG2 format by selecting the recording length of 2 s and the sampling interval of 1000 Hz. Since the study area was in an active agricultural area, there were noise sources in taking measurements. These were noises caused by agricultural processing-product transportation vehicles, other vehicles and curious people. Therefore, in data collection, seismic profile lengths were limited to 33 m due to noise and limited profile locations. Thus, data could be collected without any problems, preventing any delays in the project due to the size of the area and difficult conditions. Seismic measurements with straight-reverse-medium shot techniques were recorded carefully by hitting at least 3 times, when the traffic was under control. Due to the difficult conditions in the measurement area, it was constantly monitored by monitoring the screen that 3 stacks were sufficient for each stroke. A total of 40 seismic MASW profile measurements were collected. In VES applications, a DSI-made electrical resistivity device and a 12 V battery as a power source were used (Fig. 12a). Potential electrodes and current electrodes are stainless steel electrodes with a circular cross section (Fig. 12b–c). For the VES application, in the Schlumberger electrode array, openings were made as AB/2 = 50 m, and 20 VES measurements.

Fig. 10
figure 10

For the MASW measurement a Sara-Doremi seismograph. b Measurement in the natural area. c) Measurement in the city

Full size image
Fig. 11
figure 11

MASW measurement distances selected for the application

Full size image
Fig. 12
figure 12

The appearance of the electrical resistivity device and equipment

Full size image

The profile locations (40 profiles) for MASW measurements were determined to represent the study area (Fig. 13). To determine the discontinuities, water content, and resistive/less resistive properties of the general units of Yenişehir (Bursa) district, the VES study was conducted at 20 points in the study area (Fig. 13). Upon examining Fig. 13, it is observed that the location selections of VES points mostly cover the north-northwest and the south-southeast of the district centre in a small amount by taking the evaluation results of the seismic MASW measurements into account. The reason for this is that the measurement areas in the existing settlements are limited and urban infra-superstructures (power lines, water, sewage, natural gas lines) affect the measurements.

Fig. 13
figure 13

MASW and VES locations in the study area (prepared in the GoogleEarthPro (2020) software)

Full size image

5.2 The results of MASW data and the level maps

As is known, seismic velocities vary depending on the density of the ground and dynamic elastic parameters, and these parameters vary depending on the lithology of rocks and grounds. After collecting seismic data, first, seismic velocities and seismic dynamic-elastic parameters from these velocities were calculated in the evaluation of the seismic data. With seismic velocities (Vs, Vp, Vs30), the maximum shear modulus (Gmax), Poisson's ratio (σ), and ground amplification (Ak) values were calculated for each layer. Afterwards, the level maps of S and P-wave velocities for 0–5, 5–10, and > 10 m depths, the level maps of ground dynamic parameters, and the Vs30 level map were prepared, and grounds were examined in 3D (Figs. 14, 15, 16 and 17). A study in this setup was also conducted to prepare electrical resistivity level maps and a groundwater level map from the results of 1D VES data (Figs. 19 and 20).

Fig. 14
figure 14

Depth–velocity level maps a S wave velocities (Vs). b P wave velocities (Vp)

Full size image
Fig. 15
figure 15

Depth-dependent level maps of seismic Gmax values

Full size image
Fig. 16
figure 16

Depth-dependent level maps of σ values

Full size image
Fig. 17
figure 17

Depth-dependent level maps of Ak values

Full size image

Upon examining the seismic velocity level maps in Fig. 14a–b, the borders of two dominant geological formations in the area could be distinguished. While the formation border is ~ 1050 m in the north, in the west–east direction and on the 1st level map, it was determined that this border shifted further south in depth on the 2nd and 3rd level maps and changed up to ~ 2679 m on the 3rd level map. These borders were also observed in electrical resistivity level maps. It is observed that the seismic Vs and Vp values ​​decrease towards the south in alluvial units and increase with the beginning of the transition to the Mesudiye formation towards the north and in depth (Fig. 14a–b). Although the Vs values in Fig. 14a increase in depth (> 345 m/sec), it is observed that these values ​​are on average < 345 m/sec higher in the Mesudiye formation and on average < 229 m/sec lower in alluviums (Fig. 14a). Lower Vs values, especially in areas with alluvial units, are related to the groundwater level being very close to the surface and porous units. The presence of groundwater affects the fact that the velocities are not too high in the Mesudiye formation, which has higher velocities and mostly continues as a subunit and deeper. Likewise, the velocity changes in the level maps of the Vp values ​​in Fig. 14b depend on the same effects. However, the Vp values are on average > 1300 m/sec higher in the Mesudiye formation and on average < 807 m/sec lower in alluvium (Fig. 14b).

According to the evaluation results of seismic dynamic parameters;

The shear modulus (Gmax, kg/cm2) was calculated using the formula of Gmax = d.Vs2/100 (Bowles 1988). It shows the stiffness and shear resistance of the ground and is important in predicting ground-induced earthquake damages. According to the calculated Gmax values, the grounds in the south (MASW 1–4–5–6–8–9–10–11–12–13–14–15–16–17–18–20–24–25–26–27–28–29–36) were determined in the loose ground in the first 5 m on average (average 455 kg/cm2) and in the medium loose ground after an average of 5 m (average 856 kg/cm2) class. In the south, the grounds on the southwest edge of the area (MASW 2–7–19–21) were determined in the loose ground (average 404 kg/cm2) in the first 10 m on average in the south–north direction, and the later units were determined in the medium loose ground (average 931 kg/cm2) class. To the north, the grounds (MASW 3–22–23–30–31–32–33–34–35–37–38–39–40) were completely classified as the medium loose ground (average 1478 kg/cm2) class (Figs. 13, 15).

If the Gmax level maps in Fig. 15 are examined, it is observed that the ground is more solid towards the north and the ground strength increases in direct proportion to the depth. Towards the south, varying ground properties are observed due to alluvium. Considering the geology of the study area, it was determined that alluvial units decreased towards the north, the Mesudiye formation was entered, so the average Gmax values ​​were > 1428 kg/cm2 on average and the seismic S-wave velocities were < 347 m/sec on average, they were > 455 kg/cm2 and < 229 m/sec on average in alluvial units (Figs. 14a, 15). It is observed that the resistivities corresponding to these values ​​are higher than > 73 Ohm m on average in the Mesudiye formation and lower than < 30 Ohm m on average in the alluvium (Fig. 19).

Poisson’s ratio (unitless, σ) is the ratio of transverse compression to longitudinal extension and was calculated using the formula of σ = (Vp/Vs)2)/(2(Vp/Vs)2) (Bowles 1988). This ratio changes between 0 and 0.5 in loose, highly porous (porosity), and fractured-cracked environments. According to the calculated σ values, the study area ground was determined to be highly porous (0.42) and saturated with water in general (Figs. 13, 18). These results are consistent with the loose ground classification in the Ed and K results and geological units.

Fig. 18
figure 18

a Vs30 map. b Risk map

Full size image

According to the σ level maps in Fig. 16, σ values decrease towards the north and in depth. It is observed that while the porosity and water saturation rates decrease towards the north, these rates increase towards the south. Furthermore, the porosity is constant, while the water content in the pores decreases deeper in the north. The reason why the units in the study area usually have the same σ value is that the units are generally porous and these pores (spaces) are also filled with water. Therefore, it was observed that deeper σ values ​​decreased more, resulting in tighter ground conditions and low resistivities and seismic velocities corresponding to high porosity environments (Figs. 14, 16, 19).

Fig. 19
figure 19

Ground electrical resistivity level maps

Full size image

Ground amplification (Ak, unitless) provides the relationship of the site with the level of hazard. This important seismicity parameter is calculated with the equation of Ak = 68.Vs30−0.6 (Midorikawa 1987). The spectral ground amplification values were calculated with the Vs30 velocities of the study area, and the ground was classified according to the ground seismic amplification-hazard level assessment table of TDY (2018) (Fig. 17). Accordingly, Fig. 17 is compared with the Vs30 level map (Fig. 18a). It was determined that while Vs30 values increased by varying between ~ 207 and 347 m/sec towards the north, the amplification rates, on the contrary, decreased by changing from 2.69 to 2.14. Accordingly, the amplification class of the areas in the north corresponding to the Mesudiye formation was classified as A (low) and the average site amplification value was 2.21, while the class of the areas with alluvium was B (medium) and the average site amplification value was 2.51. These results are also consistent with the resistivity changes in the resistivity level maps (Fig. 19). Especially the fact that the amplification values of alluvial areas are B-medium and the calculated liquefaction risk is high is a risk-hazard prediction that should be taken into consideration. To this end, a risk map was prepared by considering the Vs30, Ak and groundwater level maps (Fig. 18b). The risky areas showing the hazard class in the map in Fig. 18b were divided into three regions as lower risk, medium risk, and higher risk areas. According to this risk rating, high-risk areas are observed in alluvial units. It was determined that the study area was at higher risk towards the south and these areas were in the high hazard class for housing. Therefore, the fact that Ak values ​​are B-medium in high hazard class grounds in a possible event suggests that the expected liquefaction risk and effects in a possible earthquake will increase more. As it is known, Ak values are also obtained by calculating from Vs values with the MASW method. Furthermore, the H/V ratio can also be calculated by the microtremor method, which generates natural source data with multiple random sources on the surface (Sanchez-Sasma et al. 2011). Additionally, the depth is not known exactly from the H/V ratio calculated from microtremor data. However, since Ak calculated from the V value uses the Vs value at the exact desired depth, it shows its effect better for the desired depth (Kawase et al. 2019). Generally, if the Ak value is to be calculated for an area, the H/V or Vs30 value is used. However, in specific projects, the exact amplification effect for the depth at which the foundation will sit is calculated by the Vs value at that depth. In this article's project, Ak was calculated using Vs. A level map was also prepared by processing Ak values and contributed to the interpretation.

As a result, the changes observed in the results of all calculations and level maps were examined together with resistive–non-resistive areas and low seismic velocity areas. Therefore, increases in seismic velocities and electrical resistivity values ​​(especially with the beginning of the transition to the Mesudiye formation) were observed, whether in the horizontal or vertical direction, towards the tight and tighter or the region with the decreased porosity. It was observed that these increases decreased between 5 and 10 m vertically, increased again from about 10 m in depth, and increased with the beginning of the transition to the Mesudiye formation towards the north in the horizontal direction. The formation border was clearly observed in the west–east direction and in depth.

5.3 The results of VES data and the level maps

In collecting VES measurements, the opening orientation of the Schlumberger array was generally preferred in the north–south direction since the formation change was north–south oriented. After the field studies, the inversion of 1D VES curves in the IPI2WIN + IP programme was mostly terminated with RMS errors varying between 5 and 6%, when the convergence was provided. Afterwards, level maps were prepared with the layer thickness, depth, and resistivity of the 1D depth-resistivity graphs obtained from the inversion. These maps were transformed into depth-electrical resistivity level maps in the Surfer2013 software for three different depth intervals (Fig. 19). If the horizontal and vertical resistivity changes in Fig. 19 are examined, it is observed that the resistivity in the grounds generally increases towards the north and in depth. It was observed that the electrical resistivity values on the map of the top 0–5 m varied between approximately 14–70 Ohm m, the electrical resistivity values on the map of 5–10 m in the middle varied between 6 and 46 Ohm m, and since the groundwater level in this depth range according to the upper level was approximately 5–6 m, the resistivity was lower. In the last map, it was observed that the electrical resistivity values ​​of the units towards 10 m and deeper varied between approximately 8–50 Ohm m and they had values close to the upper level values. The reason for this is that almost the same units continued at depth and these units also contained water. While the formation border in these maps was ~ 1050 m in the west–east direction at the approximately surface level in the north, if the 2nd and 3rd level maps were examined, it was determined that this border shifted towards the south under the same conditions and reached ~ 2679 m in the 3rd level map. In other words, the transition border from alluvial units with the resistivity lower than < 30 Ohm m on average in the level maps in Fig. 19 to the Mesudiye formation with the resistivity higher than > 73 Ohm m on average is also observed in Fig. 20. In Fig. 20, the varying depths of groundwater levels within the geological formation can be easily observed. Therefore, it is observed that the groundwater depths become deeper (up to about 90 m) with the beginning of the transition of the ground to the Mesudiye formation, which is more compact and has a high resistivity, towards the north. Moreover, it was measured in the study area that the elevation difference towards the north and the Mesudiye formation increased with a low inclination and reached approximately 40 m elevation levels. This is also a reason for determining the groundwater depth at deeper levels in the Mesudiye formation (Fig. 20).

Fig. 20
figure 20

Groundwater level (depth) map

Full size image

As a result, the results of seismicity, seismic, and electrical applications revealed that the high-risk areas were mostly in alluvial units where the groundwater level in the south was closer to the surface (at the depth between approximately 5–25 m) (Figs. 18a–b, 19). In other words, the ρ–Vs–Vs30 level maps were compatible, the ground was filled with tighter units with a higher resistivity towards the north, and the common features of the Mesudiye formation, the borders of the two dominant formations in the area, and the groundwater level depths were clearly determined. Two separate alluvial units were also distinguished from the calculated electrical resistivity values ​​for the exploration depth, and it was determined that the degree of grading (from a silty-clayey sand unit to a pebbly sandy unit) in these units was different in some places (Bayramoğlu 2020). Among these units, it was revealed that the upper unit had an average resistivity range of 12–40 Ohm m and there were alluvial units with the depths continuing up to 20 m deep from the surface. The lower unit is the sandstone–claystone unit with a resistivity range of 30 Ohm m increasing up to 3300 Ohm m towards the Mesudiye formation (Bayramoğlu 2020). Because the recorded signal and calculated physical quantities were different, the research depth was determined up to 20–40 m in the seismic study and up to 50 m in the electrical study, according to the 1D-2D results. Furthermore, considering the earthquake data results and seismic activity studies, due to the fact that the ground is more solid and sensitive towards the north, the ground amplification values ​​are lower in the north inversely with other dynamic parameters, the groundwater level increases towards the south, and the ground consists of silty-clayey sand and pebbly sandy units more in depth, it was suggested that liquefaction might be higher than expected in the southern grounds during a seismic activity. Therefore, it is important that the location selections in urban housing are preferred more on the northern sides of the formation border determined in the geophysical level maps and the high-risk border drawn in the risk rating map.

The corrosion classification of the study area was also made using BS1021 and TS5141EN12954 standards according to the calculated electrical resistivity values. As is known, corrosion is the loss of the initial volume and raw properties of the materials with their abrasion. Since the calculated resistivity values ​​of corrosive areas generally varied between 8 and 30 Ohm m, the corrosion degree of the ground at depths varying between 5 and 25 m was classified as medium corrosive–corrosive. This result also corresponds to the geological unit in which groundwater was present because it was determined in general that the groundwater level varied between 5 and 45 m, especially in the alluvial southern part of the area. In this case, it is important to select corrosion-resistant materials (to be long-term and safe) to be used in buildings that will be constructed in the south of the study area. Therefore, the compatibility of the results of seismicity, seismic, and electrical applications was effective in determining the risk areas, and the risky areas could be determined with precise borders (Fig. 18b). According to these borders, life-threatening and economic losses will be reduced by preferring construction in low hazard class areas. Therefore, it is important and recommended to determine the residential area according to these borders.

6 Conclusions

In this study, the earthquake, MASW and DES (provided from the KOERI) data of the area covering Yenişehir (Bursa) district centre and the Vs, Vp, Vs30 values ​​of the grounds, geological stratification, dynamic-elastic parameters of the layers (Gmax, σ, average Ak) values, the resistivity of the layers, groundwater levels, probable earthquake magnitudes and recurrence probabilities expected from the earthquake data, average amax and risk-hazard conditions in terms of liquefaction–seismicity were determined. The formation transition borders in the horizontal–vertical directions, loose-porous areas corresponding to low seismic velocity-low resistivity areas, and the liquefaction risk and hazard class in these areas were determined with the level maps, which are the main objectives of the study. According to all geophysical calculations:

  • 1. According to the calculated Vs-Vp-Gmax-σ values ​​and all the level maps prepared, it was observed that the seismic velocity and dynamic parameter values ​​increased with the beginning of the transition from alluvial units classified as quite loose and porous to the Mesudiye formation, which has a tighter ground characteristic, towards the north. On the contrary, it was observed that Ak values ​​decreased. While the Vs30 values ​​increased towards the north by changing between 217.6 and 319.1 m/sec, the ground amplification decreased from 2.69 to 2.14. The amplification classification of the Mesudiye formation region was A (low) and Ak = 2.21 on average, and the amplification class of the area with alluvium was B (medium) and Ak = 2.51 on average. It was observed that the units forming the Mesudiye formation were more compact than the alluvial units, the study area was generally highly porous (0.42) and saturated with water, and these characteristics increased mostly towards the south.

  • 2. It was determined that low resistivity areas in alluvial units in the south ranged between 8 and 30 Ohm m, the groundwater level depths varied between ~ 5 and 25 m (deeper the Mesudiye formation in the north). However, the resistivity values ​​changing towards the depth showed also that the grounds were in the medium corrosive–corrosive class and there were different gradations in the alluvial units in some places. Therefore, it is important to select building materials resistant to corrosion.

  • 3. The formation boundary was observed on all level maps and depths. This boundary is also compatible with the risk map. On the level maps, it was observed that the seismic velocities (approximately Vs > 270 and Vp > 1300 m/sec) and electrical resistivity values ​​(approximately > 53 Ohm m) increased towards the solid region in the horizontal and vertical directions. Especially with the beginning of the transition to the Mesudiye formation, these increases were observed to decrease between 5 and 10 m vertically, but they increased again from 10 m to deeper levels. While the formation border length was ~ 1050 m approximately on the surface in the west–east direction in the north, it was determined that this border shifted further south under the same conditions more in depth on the 2nd and 3rd level maps and reached ~ 2679 m on the 3rd level map.

  • 4. From the PEHA, PGA, and PPD results, it was calculated that the M-logN values showed an inversely proportional linear relationship with each other, as the magnitude of the earthquake increased, the probability of a larger earthquake occurred with a longer recurrence year, the frequency of recurrence of earthquakes with M ≤ 5.5 within 10–50–75–100 years was higher, and the frequency of recurrence of earthquakes with M ≥ 6 and the average recurrence periods also decreased. For M = 7.4, the average amax = 0.31 g was calculated, and according to the average amax, Vs30, Ak, and groundwater information, the study area ground had a liquefaction risk and was in the high hazard class. This situation involves more risk and higher hazard further south and was accepted as a hazard prediction that requires attention. Since the high risk of liquefaction in alluvial areas is related to the low values ​​of Vs velocities and other calculated parameters, it is important to consider this in housing by taking into account that the expected liquefaction risk and effects in a possible earthquake in the study area will increase more.

Consequently, geophysical studies were effective in determining risky areas with precise borders, and it was determined that the location selection in construction should be made towards the north. Therefore, it is important to consider the results of this study within the disaster-risk-prediction relationship. While these studies provide scientific contributions to regulating the construction–ground interaction, they will also affect the selection of the right building material and the number of floors in higher risk alluvial areas, healthy and safe urban development and reducing costs. Therefore, if the results of these and similar studies, which provide preliminary information, are taken into account, significant contributions will be made to the protection of safe housing area borders.