The great Petrinja earthquake – a year after
Josip Stipčević*, Iva Dasović*, Davorka Herak*, Marijan Herak*, Helena Latečki*, Marin Sečanj* and Bruno Tomljenović**.
* Department of Geophysics, Faculty of Science, University of Zagreb
** Faculty of Mining, Geology and Petroleum Engineering, University of Zagreb
A year has passed since Petrinja and its wider surroundings were hit by a devastating earthquake. In the period after the earthquake, a large number of citizens, emergency and rescue services helped the residents of Banija to overcome the severe consequences of the strong earthquake, which continued for months after the main earthquake in the form of daily aftershocks. During the same period, a large number of scientists and professional services were engaged in research into the causes and consequences of the earthquake itself. The Department of Geophysics of the Faculty of Science and the Faculty of Mining, Geology and Petroleum Engineering were among the first to rush to the devastated area to collect measurements. These measurements were of great importance for estimating the further course of the earthquake series near Petrinja and possible secondary earthquake effects (landslides, sinkholes/drooping, etc.). These early investigations helped to reduce the risk due to subsequent strong shaking and enabled the seismological and geological scientific community to obtain data that will provide answers in the future as to why and how often earthquakes occur in the area of Petrinja and Pokuplje. This is the first strong earthquake with an epicenter in Croatia in modern times allowing us to set up a dense network of extremely sensitive instruments and "monitor" this unique "experiment" in a natural laboratory. In the following text, we will go through the most important results derived from seismological, seismotectonic and geological surveys conducted in the last year.
A series of earthquakes with the epicentre in the vicinity of Petrinja began on Monday 28 December 2020 at 6:28 local time with an earthquake of magnitude ML = 5.1 (current magnitude MW = 4.9) which was felt in most of central Croatia. Its epicenter was southwest of Petrinja, near the town of Strašnik. Earthquakes of local magnitude 4.6 at 7:49 and magnitude 3.8 at 7:51 in the same epicentral area soon followed, as did a series of weaker earthquakes (Herak and Tomljenović, 2021). Unfortunately, as it turned out these moderately strong earthquakes were actually foreshocks because the next day, 29 December 2020, at 12:19, a strong earthquake of local magnitude 6.2 (MW = 6.4) occurred, with an epicenter also near Strašnik. According to the Seismological Survey at the Geophysical Department of the Faculty of Science, the earthquake was estimated to have had an intensity in the epicentre VIII EMS and is described as severely damaging (URL1). The earthquake was felt throughout Croatia and Slovenia and in much of Bosnia and Herzegovina, Serbia, Hungary, Italy, and even Austria and Slovakia (EMSC, 2021). Seven people lost their lives as a result of the collapsed houses. However, due to the earthquake the day before and the damage they caused, and the fact that the mainshock occurred in the middle of a sunny day, many people in the wider epicentral area were outside their homes, especially those living in the houses damaged in the foreshock.
Deployment of a temporary seismic network
The local network of instruments set up in this way dramatically improved the accuracy and reduced the instability of the earthquake foci locations, which enabled us to locate even the very weak earthquakes.
When such an exceptional earthquake series occurs, the seismological community must respond quickly and set up a dense temporary network of instruments in the wider epicentral area in order to record and locate very weak earthquakes. This is extremely important especially for determining the depth of the earthquake foci – which is the least reliable parameter. Due to the Zagreb earthquake series that began in March 2020 and the deployment of a temporary network to monitor the aftershocks, Croatian seismologists were left without free instruments that could be placed in the Banija region. Seismologists from the Geophysical Institute of Andrija Mohorovičić, Department of Geophysics, Faculty of Science, University of Zagreb contacted the colleagues from the Istituto Nazionale di Oceanografia e di Geofisica Sperimentale (OGS) from Udine who made six seismographs available to deploy in the mainshock epicentral area. Five instruments were installed on 4 and 5 January 2021 in Hotnja, Sisak, Taborište, Novo Selo Glinsko and in Mečenčani (Petrinja-net, Figure 1; Stipčević et al., 2021). A seismograph from the temporary seismological station on the island of Vir was moved to Petrova gora on 4 January, with the cooperation of the Seismological Survey at the Geophysical Department of the Faculty of Science. The local network of instruments set up in this way dramatically improved the accuracy and reduced the instability of the earthquake foci locations (Figure 2), which enabled us to locate even the very weak earthquakes. The Seismological Survey soon received a valuable donation from the Government of the Republic of Croatia, which procured 40 quality instruments (20 seismometers and 20 accelerometers), so in the second half of January 2021 the installation of a temporary mobile network began (URL2). Since strong earthquakes are rare and no two earthquakes are equal, this event provided a unique opportunity allowing us to investigate this fault system and determine its seismotectonic characteristics.
Figure 1. Seismological stations in the wider vicinity of Petrinja. Stations that operated before the earthquake (Seismological Survey and Geophysical Institute, Geophysical Department of the Faculty of Science; ARSO, Ljubljana) are marked with green triangles. The red triangles show the stations set up on 4–5 January 2021 (Geophysical Institute, Zagreb and Istituto Nazionale di Oceanografia e di Geofisica Sperimentale – OGS, Udine). The blue triangles mark the stations with instruments from the donation of the Government of the Republic of Croatia, with the beginning of the installation on 18 January 2021 (Seismological Survey at the Geophysical Department of the Faculty of Science). All located earthquakes (28 December 2020 – 29 March 2021) are shown in black dots.
Figure 2. Time sequence of average errors (one standard deviation) of the locations of the hypocenters of the Petrinja earthquake series in the first three months (ML ≥ 0.75, Ndat ≥ 10). Empty blue rectangles indicate unreliability of focal depth, and solid red rectangles indicate unreliability of epicentre location (in kilometres). The points show the medial values for each day. There is a dramatic improvement in the accuracy of locations after the installation of a temporary network of six seismographs on 4–5 January 2021 installed jointly by Geophysical Institute of the Geophysical Department of the Faculty of Science and OGS from Udine.
The main earthquake was followed by numerous aftershocks, within a very short time interval, so earthquake recordings often overlap each other – this makes it difficult to accurately analyze and determine the location of the earthquake. Figure 3 shows earthquake records in the first 40 min after the mainshock: dozens of subsequent earthquakes followed, and when we take a closer look at 10 min window, we see even more very weak earthquakes. For some very weak earthquakes, it was not possible to determine the location because they were not strong enough to be recorded at at least three seismological stations, especially in the first few days.
Figure 3. a) Beginning (main earthquake on 29 December 2020 at 12:19, ML = 6.2) and the next 40 min of the Petrinja earthquake series for the five stations (STA1,… STA5) closest to the epicentre. A red rectangle indicates the 10 min time window shown in more detail in section b). For each station, three components of motion are shown, from top to bottom in order Z (up-down), N (north-south), E (east-west). The amplitudes are scaled so that weaker earthquakes are seen, so for stronger aftershocks the records look truncated.
Seismological features of the Banija earthquake series
...9350 earthquakes were located in the period between 28 December 2020 and 29 March 2021.
According to preliminary data (analyzed by M. and D. Herak), 9350 earthquakes were located in the period between 28 December 2020 and 29 March 2021 (Figure 4). Of that number, 6374 were determined with a standard location error of less than 1 km (Figure 5). Most of the epicenters are located in a very narrow, well-defined area along Hrastovička gora (shown in Figure 10), along the well-known Petrinja fault trending northwest-southeast. One smaller group of earthquakes is located east of the main group – in the area between Petrinja and Mošćenica. A smaller group stands out west of the main group – around Velika Solina. From Figures 5 and 6 we can see that the main earthquake had a focal depth of about 6–7 km. The longitudinal profile (A in Figure 6) shows that a relatively small number of subsequent earthquakes occurred in the part of the fault surface around the main earthquake and between the depth of approximately 10 km and the surface, because most of the collected tension was released from that section. Most of the subsequent earthquakes occurred at depths between 10 and 18 km, below the part of the fault that was activated in the main earthquake. Transverse profile B clearly shows that the fault is vertical. The strongest subsequent earthquake occurred near Župić on 6 January 2021 at 18:01, and was of local magnitude 4.9 (MW = 4.8). In the first 13 days of the series, there were a total of ten earthquakes with ML ≥ 4.0, and 76 earthquakes ML ≥ 3.0 (Dasović et al., 2021). Statistical analysis of the magnitude of subsequent earthquakes in the first three months of the series indicates that the slope of the Gutenberg-Richter relation has value of b = 0.9, which means that for each unit of decrease in magnitude the number of aftershock earthquakes increases 7.9 times.
Figure 4. Epicentres of all 9350 located earthquakes in the first three months of the earthquake series, between 28 December 2020 and 29 March 2021. The color of the circles shows the magnitude of the earthquake according to the scale on the right.
Figure 5. Epicentres of 6374 earthquakes (28 December 2020 – 29 March 2021) located with a standard location error of less than 1 km. The color of the circles depends on the depth of the focus according to the scale on the right side of the image.
Figure 6. Depth sections through the earthquake foci cloud. The top figure shows the epicenters of reliably located earthquakes with magnitude ML ≥ 1.5 for which there is data of at least 10 occurrence times at various stations, with azimuth window gap less than 90° and standard epicenter error of less than 1 km. Profiles A and B are shown in the figures below (depth is on the y axis and distance from the beginning of the profile is on the x axis). The color scale on the side of each graph shows the local magnitude of the earthquake, which is added to the size of the circles, so the large purple circle represents the strongest, main earthquake.
The earthquake focal mechanism of the main earthquake (Figure 7) shows that it is a subvertical (almost vertical) right fault, strike-slip with a horizontal displacement along the northwest-southeast fault surface. This means the southwestern fault wing shifted to the northwest and northeastern to the southeast. The obtained focal mechanism coincides well with the solutions obtained by global seismological centers done with the inversion of seismic moment tensors, with the spatial distribution of earthquake foci outlining the activated part of the fault, as well as with the preliminary results obtained by InSAR method (Figure 9). The focal mechanisms for ten strongest events (ML ≥ 4.2) mostly indicate earthquakes with a horizontal displacement along the vertical fault (strike-slip), especially for the strongest foreshock and aftershock, and mainshock (Figure 8). But three earthquakes show a different type of focal mechanism, the so-called reverse faults when the fault wing above the fault surface of an oblique fault moves upwards in relation to the fault wing below the dip surface. Most of the 25 mechanisms we have calculated so far are strike-slip (as well as the main earthquake), but many, especially outside the main earthquake group, have formed on reverse faults. From the focal mechanisms we can determine the direction of the highest pressure in the stress field and these earthquakes show that in this area there is compression (compaction) of the Earth's crust in the direction of SSW-NNE.
Figure 7. Displacement mechanism in the epicenter of the main earthquake (December 29, 2020, ML = 6.4). The stereographic projection on the lower focal sphere shows the orientations of the first P-wave displacement at a total of 201 seismological stations: red crosses refer to the first upward displacements (compression), while white circles show the first downward displacements of the soil (dilatation). Of the two fault lines (red and blue line) which are equivalent solutions, the causal fault is the one of the NW – SE (right fault, blue arch).
Figure 8. Focal mechanisms for ten ML ≥ 4.2 earthquakes from the Petrinja earthquake series. "Beach balls" with red fields indicate a horizontal shift along an almost vertical fault (e.g. strike-slip, for the strongest previous, main and strongest subsequent earthquake), while the blue "beach balls" mark the reverse (dip-slip) mechanisms in which the upper fault wing of an oblique fault moves upwards in relation to the lower wing. The green lines next to each diagram indicate the direction of the axis of maximum pressure, i.e. tectonic compression - all mechanisms indicate the compression in the direction of SSW-NNE.
Figure 9. Earthquakes of magnitude ML ≥ 2.0, on satellite interferogram (DInSAR, M. Govorčin, personal communication). Areas that shifted to the east during the main earthquake are marked in green, and those that have shifted to the west are marked in purple. The largest displacements were about 40 cm. The dark blue line indicates the approximate position of the causal fault (right fault with displacement by extension).
The epicentral area belongs to Hrastovička gora, the most pronounced morphological structure near Petrinja, with the highest peak Cepeliš at 415 m above sea level (for more on geomorphological analysis, see Bočić, 2021). According to surface geological data (Pikija, 1987), Hrastovička gora is an asymmetric anticline trending northwest-southeast. The hinge and steeper northeastern limb of the anticline are locally built of Upper Cretaceous and Eocene rocks unconformably covered by younger Miocene and Pliocene deposits (sedimentary filling of the Sava Depression), which are largely exposed in the gently dipping southwestern limb of the anticline. The northeastern limb of the anticline is bounded by the Hrastovica fault (part of the Petrinja fault system) striking northwest-southeast, which was characterized as a reverse fault with vergence (tectonic transport) to the northeast, based on the seismotectonic profile (Figures 10 and 11). Due to the neotectonic activity of the Petrinja fault system, i.e. several seismic cycles during the Pliocene and Quaternary, the sedimentary deposits exposed at Hrastovička gora were uplifted in relation to identical deposits northeast of Hrastovica fault and in relation to the youngest Plio-Quaternary alluvial deposits of Kupa river system.
Figure 10. Simplified geological map of the Petrinja epicentral area (modified according to Pikija, 1987) showing the earthquake locations of the Petrinja seismic series. Data on earthquake epicenters were taken from the Croatian Earthquake Catalog of the Seismological Survey at the Geophysical Department of the Faculty of Science for the period from 25 December 2020 to 15 February 2021.
Figure 11. Seismotectonic profile through the Petrinja epicentral area showing the foci of the Petrinja seismic series. Data on earthquake foci were taken from the Croatian Earthquake Catalog of the Seismological Survey at the Geophysical Department of the Faculty of Science for the period from 25 December 2020 to 15 February 2021. The spatial position of the profile is shown in Figure 10.
Coseismic and secondary effects of earthquake
... in the wider epicentral area a large number of secondary earthquake effects, i.e. cosmic deformations on the surface such as liquefaction, cracks, landslides and sinkholes was observed.
... the greatest natural hazard was the occurrence of dropout sinkholes in the villages of Mečenčani and Borojevići...
Due to the strength of the earthquake and specific geological setting, in the aftermath of the shaking in the wider epicentral area a large number of secondary earthquake effects, i.e. cosmic deformations on the surface such as liquefaction, cracks, landslides and sinkholes was observed (e.g. Tomljenović et al., 2021; Pollak et al., 2021). Liquefaction, which represents a sudden loss of strength of saturated incoherent soils, was observed in the wider epicentral area around Petrinja, Sisak and Glina, in the alluvial deposits of the Kupa, Sava rivers and their tributaries. However, the most susceptible to liquefaction were sandy soils deposited in the floodplains of the Kupa and Sava rivers. The occurrence of liquefaction in the form of sand “volcanoes” (Figure 12a) and liquefaction cracks on the surface was most pronounced in agricultural areas along the Kupa and Sava rivers (Figure 12). At the same time, in urban areas it caused damage to transport infrastructure, buildings and embankments of the Sava and Kupa due to subsidence, sinking and lateral spreading of the soil (Figure 13a).
Figure 12. Photographs of liquefaction: a) in the form of sand “volcanoes” in village of Brest on the left bank of the Kupa river, b) in village of Nebojan on the right bank of the Kupa river, c) in village of Palanjek on the left bank of the Sava river and d) satellite image taken from Google Earth showing the liquefaction along the abandoned tributary of the Sava River near village of Budaševo.
Also, a number of new and reactivated landslides were recorded (slides, lateral spreading and rockfalls). Landslides have been recorded in loose materials (e.g. road embankments) damaging a number of roads in the wider epicenter area, but also in marl and clayey soil, such as the most dangerous landslide in the village of Prnjavor Čuntićki (Mihalić Arbanas et al., 2021) about 10 km south of Petrinja. Due to the reactivation of this landslide and its further activity in the months after the main earthquake, it was necessary to evacuate residents from their homes located on the landslide and near vicinity. In addition to landslides, immediately after the mainshock, minor occurrences of rockfalls from steep slopes or rock cuts were recorded, as well as lateral spreading that is directly related to liquefaction. Due to the lateral spreading subvertical cracks several meters deep were formed, most often on the Sava and Kupa river embankments and in their vicinity (Figure 13).
Figure 13. Photographs of almost vertical cracks caused by lateral spreading on a) the Sava embankment and b) agricultural land in the village of Palanjek.
In addition to the previously described secondary deformations on the surface caused by shaking, the greatest natural hazard was the occurrence of dropout sinkholes in the villages of Mečenčani and Borojevići, which are still opening after stronger aftershocks (over 140 such sinkholes have been recorded so far). Dropout sinkholes are formed in the Holocene alluvial-proluvial deposits of the Sunja River and its occasional torrential tributaries. In the valley where the villages of Mečenčani and Borojevići are located, these deposits cover the karstified base built of Baden carbonate rocks. The formation of dropout sinkholes is a long-term process associated with suffusion, during which fine-grained soil and sediments in the top layer are washed away by groundwater into fissures and underground caverns, which ultimately leads to sudden collapse of the cover deposits (Figure 14). This otherwise lengthy process was significantly accelerated by a tremor during the Petrinja earthquake series, with dropout sinkholes damaging several residential buildings. Given that the earthquake series is still ongoing, there is still the possibility for the formation of new dropout sinkholes, and this poses a significant risk to the local population.
Figure 14. a) Sketch of the dropout sinkholes formation process (modified according to USGS, 2021, URL3) and b) photograph of the largest dropout sinkholes in Mečenčani village with a diameter of more than 20 m and a depth of about 12 m.
Although preliminary, the results clearly show the extent to which the Earth's crustal structure and soil properties of an area affect the propagation of seismic waves. Thus, for example, in valleys and lowlands the waves remain trapped in softer alluvial sedimentary deposits, between the surface and the solid rock mass in the bedrock.
To simulate the Petrinja earthquake that occurred on 29 December 2020 at 12:19:54, MW = 6.4, a new 3D model of velocity and density of the central Croatia was created. The source is represented by a simplified point model, which assumes that the accumulated energy is released in a very small space, specifically the point. This is, of course, only an assumption introduced for the purposes of performing a mathematical calculation, while in reality the source of the earthquake is a fault area about 15 km long and 10 km wide. For the next iteration of the simulation, the plan is to use a source model of finite dimensions that will better describe the fault displacement field that generates seismic waves. Such simulations contribute to a better understanding of the effects of earthquakes (both current and historical) and improve the assessment of seismic hazard in the study area.
Although preliminary, the results clearly show the extent to which the Earth's crustal structure and soil properties of an area affect the propagation of seismic waves. Thus, for example, in valleys and lowlands the waves remain trapped in softer alluvial sedimentary deposits, between the surface and the solid rock mass in the bedrock (see the time simulation 12:20:20 in Figure 15). As a result, the duration of the tremor is prolonged and the amplitude of the waves increases (e.g. Zagreb has been shaking for a little over a minute!). On the other hand, in mountainous areas, where solid rock reaches the surface, a rapid passage of the wave front can be seen, so the time of the tremor itself is somewhat shorter. Thus, for example, on Medvednica the shaking lasted only between 30 and 45 s due to different soil properties (solid rock which, unlike sediments, does not amplify waves), it was not as pronounced as in Zagreb. It is interesting to note that seismic waves are channeled in some parts (e.g. the area near the town of Zaprešić), which also increases the strength of the shaking. This is a direct consequence of the geological structure - in the case of Zaprešić, between the Samobor hills and Medvednica, there are deposits of sedimentary rocks in which the waves remained trapped, and the amplitude of the shaking increased. This resulted in increased and prolonged tremors in the area, which caused great damage to some buildings, although the location is relatively far from the epicenter of the earthquake.
Figure 15. Representation of the horizontal E-W (east-west) component of the ground velocity obtained by numerical simulation of the Petrinja earthquake shown as six snapshots from simulation. This simulation was made as part of the research work of PhD student Helena Latečki under the guidance of Assistant professor Josip Stipčević from the Faculty of Science and in collaboration with Dr. Irene Molinari from the National Institute of Geophysics and Volcanology (INGV), Italy. As part of this research, the resources of the Isabella computer cluster maintained by the University Computing Center of the University of Zagreb (Srce) were also used.
The complete simulation of seismic wave propagation, for all three components, partially shown in Figure 15 can be viewed at the following links: N–S component, E–W component i vertical component Z
In order to be able to analyze this series of earthquakes in more detail, it is necessary to carefully analyze at least one whole year: locate all earthquakes that can be located, determine the focal mechanism of those that are somewhat stronger. Only then will it be possible to draw some detailed conclusions about what exactly happened and how to describe this fault system more accurately. Machine learning methods can help to automatically recognize thousands of very small earthquakes among several terabytes of collected data, which give a clear insight into the structure of activated faults. In addition, a multidisciplinary approach to seismology, geodesy, and geology can help distinguish activated faults and their interactions (e.g., stress-transfer) over a wider epicentral area. These studies can provide better input data for modeling the observed fields of intensity, displacement, velocity and soil acceleration during earthquakes in the wider epicentral area (including the effects of surface soil layers, realistic fault and fault models, etc.). Such models make it possible to determine realistic seismic scenarios for future earthquakes, thus improving deterministic modeling of earthquake effects and seismic hazards (see previous chapter!). Research using engineering seismology and geology, as well as geotechnical methods, will provide better knowledge of local soil properties that significantly affect the frequency and amplitude of earthquake waves, and can have devastating effects on buildings, not only in the epicentral area, but far from it - just as it was in Zaprešić. Certainly, there is the collaboration of seismologists and builders that must exist in order to build a society more resilient to earthquakes. Last but not least, we need to work hard to popularize and promote seismology and raise awareness of the dangers of earthquakes in order to learn to live with them so that their consequences are not disastrous.
In this analyses earthquake records of the Croatian Seismograph Network were used, jointly maintained by the Seismological Survey and the Andrija Mohorovičić Geophysical Institute of the Department of Geophysics, Faculty of Science, University of Zagreb – we would like to thank the Seismological Survey for sharing their data.
Bočić, N. (2021): Structural-geomorphological aspects of the Petrinja earthquake M6.2 (Croatia) – preliminary considerations, Hrvatski geografski glasnik, 83/1, 5−24, https://doi.org/10.21861/HGG.2021.83.01.01.
Dasović, I., Herak, D., Herak, M., Stipčević, J. (2021): Niz potresa na Baniji, Vijesti Hrvatskog geološkog društva, 57/2, 4–7, http://www.geologija.hr/wp-content/uploads/2021/10/web-vijesti-57_2.pdf.
The European-Mediterranean Seismological Centre (2021): M6.4 CROATIA on December 29th 2020 at 11:19 UTC, Special report, https://www.emsc-csem.org/Earthquake/264/M6-4-CROATIA-on-December-29th-2020-at-11-19-UTC. [4. ožujka 2021.]
Herak, M., Tomljenović, B. (2021): Seizmotektonske značajke epicentralnih područja, koseizmičke deformacije na površini i seizmološka opažanja serija potresa kod Zagreba i Petrinje 2020./21. godine, Znanstveno vijeće za zaštitu prirode Hrvatske akademije znanosti i umjetnosti, online predavanje, 5. veljače 2021.
Mihalić Arbanas, S., Arbanas, Ž., Krkač, M., Bernat Gazibara S., Damjanović, V., Sinčić, M., Jagodnik, P., Jagodnik, V. (2021). Potres u Petrinji M6,2 od 29.12.2020: Izvješće o inženjerskogeološkoj i geotehničkoj prospekciji, Rudarsko-geološko-naftni fakultet, Sveučilište u Zagrebu, https://www.rgn.unizg.hr/images/izdvojeno/UNIZR-RGNF_UNIRI-GF_potes_Petrinja_M6.2_ig_i_gt_prospekcija.pdf.
Pikija, M. (1987): Osnovna geološka karta SFRJ 1:100.000, list Sisak, L 33-93. Geološki zavod, Zagreb; Savezni geološki zavod, Beograd.
Pollak, D., Gulam, V., Novosel, T., Avanić, R., Tomljenović, B., Hećej, N., Terzić, J., Stipčević, J., Bačić, M., Kurečić, T., Dolić, M., Bostjančić, I., Wacha, L., Kosović, I., Budić, M., Vukovski, M., Belić, N., Špelić, M., Brčić, V., Barbača, J., Kordić, B., Palenik, D., Filjak, R., Frangen, T., Pavić, M., Urumović, K., Sečanj, M., Matoš, B., Govorčin, M., Kovačević, M.S., Librić, L. (2021): The preliminary inventory of coseismic ground failures related to December 2020 – January 2021 Petrinja earthquake series. Geologia Croatica, 74/1, 189–208, https://doi.org/10.4154/gc.2021.08.
Stipčević, J., Poggi, V., Herak, M., Parolai, S., Herak, D., Dasović, I., Bertoni, M., Barnaba, C., Pesaresi, D., (2021): First results from temporary deployment of small seismic network following the Mw=6.4 Petrinja earthquake, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-16579, https://doi.org/10.5194/egusphere-egu21-16579, prezentacija.
Šikić, K. (2014): Osnovna geološka karta Republike Hrvatske 1:100.000 list Bosanski Novi, L 33-105. Hrvatski geološki institut, Zagreb.
Tomljenović, B., Stipčević, J., Sečanj, M. (2021). Izvješće o zabilježenim pojavama koseizmičkih površinskih deformacija na području Pokuplja i Banovine nastalih potresnom serijom od 28.12. 2020 do 5.01.2021. Rudarsko-geološko-naftni fakultet, Sveučilište u Zagrebu, https://www.rgn.unizg.hr/hr/izdvojeno/2790-izvjesce-o-zabiljezenim-pojavama-koseizmickih-povrsinskih-deformacija-na-podrucju-pokuplja-i-banovine-nastalih-potresnom-serijom-od-28-12-2020-do-5-01-2021.
[URL1] Seismological Survey (2021): Godina dana od razornog potresa kod Petrinje, https://www.pmf.unizg.hr/geof/seizmoloska_sluzba/potresi_kod_petrinje/2020-2021.
[URL2] Seismological Survey (2020): Mobilna mreža seizmoloških postaja, https://www.pmf.unizg.hr/geof/seizmoloska_sluzba/mobilna_mreza.
[URL3] U.S. Geological Survey’s Water Science School (2018): Sinkholes, https://www.usgs.gov/special-topics/water-science-school/science/sinkholes.