4.1 The effect of ΔCFS imparted by the 2017 Poso and the 2018 Palu-Donggala earthquakes on the Poso fault
In explaining the influence of the 2017 Poso and the 2018 Palu-Donggala earthquakes into the Poso fault, we estimated the accumulation of ΔCFS imparted by those two events (Figure 4 and 5) and displayed their cumulative values on multi-segments of the Poso fault in Figure 6. The distribution of the cumulative ΔCFS along the Poso fault exhibits high negative value in the north (segment 1- segment 4) of up to -20 kPa and moderate negative value in the center (segment 5 – segment 11) of up to -5 kPa. To analyse more detail the effect of large inland earthquakes on each segment of Poso fault, we investigate their values on each parameter of and dip. When we used =0.4 and 0.1 with dip=25° (Figure 4a and 4b) for ΔCFS calculation of the 2017 Poso earthquake, at segment 10, it exhibits similar positive ΔCFS on each depth. However, when we used dip=45°, it produced negative ΔCFS > ~-2 kPa at 20 km – 25 km depth (Figure 4c and Figure 4d). This pattern is also found on other segments, such as segment 8, 9, 11 and 12. Daniarsyad et al. (2021) modeled the ΔCFS distribution of the 2017 Poso earthquake and found positive ΔCFS > ~10 kPa in segment 1 – 10, it relatively consistent with the results in this study where we also found positive ΔCFS of ~18 kPa in segment 5 – 10 while other segments exhibit negative ΔCFS.
For the ΔCFS calculation of the 2018 Palu-Donggala earthquake using =0.4 and 0.1 with dip=25° (Figure 5a and 5b), segment 1 – 12 exhibits high negative ΔCFS > -10 kPa, remarkably similar to the pattern following dip=45° (Figure 5c and 5d). Only Figure 5c that exhibits smaller negative ΔCFS than any other combined parameters. However, despite the dominantly negative ΔCFS imparted by the 2018 Palu-Donggala earthquake, the positive ΔCFSs are identified when using =0.4 and dip=45° (Figure 5d) at lower depth on southern part of Poso fault (segment 13, 15, 16, 18, and 19) with ΔCFS > 0.1 kPa. In comparison with another study produced by Gunawan et al. (2020), their results showed positive ΔCFS > ~20 kPa in the north of the Poso fault and negative ΔCFS in the central up to south of the Poso fault > ~-20 kPa. This difference is significantly caused by multi-strike parameter used in this study.
The cumulative ΔCFS displayed in Figure 6 also yields almost similar pattern for different parameters of and dip. The ΔCFS results in Figure 6a and 6b exhibit similar pattern with high negative ΔCFS > ~-20 kPa on the north and low negative ΔCFS > ~-5 kPa. Only when using dip=45° (Figure 6c and 6d), some segments experiences positive ΔCFS, for instance, for =0.1 (Figure 6c), the positive ΔCFS of ~5 kPa recorded on segment 17 at 5 km depth and for =0.4 (Figure 6d), the positive ΔCFSs are identified in segment 9 at 20 km depth and segment 10 at 5 km depth of around 1 kPa. Based on cumulative results of those two large events, it can be seen that the influence of the 2018 Palu-Donggala earthquake is much stronger than the 2017 Poso earthquake. This is certainly caused by the difference in magnitude where the 2018 Palu-Donggala earthquake produced Mw 7.5, almost an approximate logarithmic divergence of one unit in contrast to the magnitude produced by the 2017 Poso earthquake. The cumulative ΔCFS results also transferred the negative ΔCFS in almost the entire segment of the Poso fault (from ~-5 kPa to ~-10 kPa) meaning that their effects lengthen the Poso fault to advance towards its critical condition.
4.2 ΔCFS time function on the Poso fault
In modeling the ΔCFS time function on the Poso fault, we combine the results of ΔCFS due to large inland earthquakes and the tectonic stress rate as this model has been presented in other stress evolution study on Sumatran Fault (see Rafie et al., 2023), SW Japan (Hori and Oike 1999; Shikakura et al. 2014; Mitogawa and Nishimura 2020) and Palu Koro fault (Liu and Shi 2021).
We model the ΔCFS time function in range of 20 years, from 2010 – 2030 where the 2017 Poso and the 2018 Palu-Donggala occurred on that period of time. We used the ΔCFS model resulted from using =0.4 and dip=25° at 10 km depth. The ΔCFS time function model is shown in Figure 7, other models that using different combination of and dip can be seen in Figure S2, S3, and S4. The tectonic stress rate of 2.4 kPa on the Poso fault estimated in this study is roughly similar to the result obtained by Liu and Shi (2021) where their estimation the tectonic stress rate around Sulawesi Island to be 3.6 kPa. Also, this value is similar to that obtained in SW Japan’s Nankai Trough forearc by Shikakura et al. (2020) and Mitogawa and Nishimura (2020) as well as in Sumatran Fault with the case of no sliver movement (see Rafie et al. 2023), which suggesting a low rate of tectonic loading.
Based on Figure 7, high negative ΔCFSs due to the 2018 Palu-Donggala earthquake are concentrated around northern part of the Poso fault at segment 1 – 12 ranging from -20 kPa to -10 kPa and low negative stress in the south at segment 13 – 20 with ΔCFS > -3 kPa. The 2017 Poso earthquake transferred low negative stress as low as > ~-3 kPa in the north (segment 1 – 4) and south (segment 12 – 20) of the Poso fault, while high positive ΔCFSs of > ~10 kPa are found in the center of the fault (segment 5 – 11). Due to high stress release by the 2018 Palu-Donggala earthquake in comparison with the 2017 Poso earthquake, its coseismic stress change drop the stress accumulation on segment 1 – 4, 6, and 11 – 12, so that it inhibits these segments to bring forward to their critical condition. For segment 5 and 7 – 10, positive ΔCFS resulting from the 2017 Poso earthquake and negative ΔCFS arising the 2018 Palu-Donggala earthquake exhibit nearly equivalent magnitudes, so that their effects cancel each other out. Other segments on the south (segment 13 – 20) experience the lowest coseismic ΔCFSs by these two large earthquakes
From these analyses, we infer that the effect of large earthquakes sufficiently controls the stress accumulation within the northern to central segments of the Poso fault, whereas the southern portion of the Poso fault is more distinctly governed by tectonic stress rates (see Figure 8). For overall ΔCFS time function models along the Poso fault displayed in Figure 7, it is evident that the southernmost segment of the Poso fault exhibits the highest stress accumulation, suggesting a potential advancement towards critical conditions. However, determining the specific segments on the Poso fault that may have advanced to critical thresholds remains difficult to constrain due to the paucity of historical large earthquake and the limited understanding of fault strength characteristics along the Poso fault. Furthermore, in the absence of large earthquakes on this fault, it suggests the potential presence of creeping, akin to those observed along the Aceh fault. Despite its high stress accumulation, the Aceh fault experiences creep and has yet to produce earthquake with M>7.0 (Rafie et al. 2023; Ito et al. 2012; Tong et al. 2018). Nonetheless, further investigations are requisite to substantiate this assertion, particularly through geodetic and geological observations.