Abstract The Charnath Khola is a large river crossing the Himalayan thrust system in the region devastated by the great M8.3 1934 Bihar ‐Nepalearthquake. Fluvial terraces are abandoned along the river and at the base of a ~20 ‐m high cumulative thrust escarpment. A trench across the fault scarp exposed Siwalik mudstone/siltstone overthrusting Quaternary units and three colluvial wedges inter ﬁngered with ﬂuvial sands. The 85 accelerator mass spectrometry radiocarbon dates, from detrital charcoals sampled in the trench, a river cut and river terraces, constrain the timing of the sedimentary processes following the last two major earthquakes, in 1934 and 1255 CE. Although several samples straddle the main earthquake horizon, associating it with the 1934 earthquake, based solely on radiocarbon ages, remains challenging. The 49 detrital charcoal ages found in the pre ‐earthquake and postearthquake units fall between 65 and 225 BP, a period with a ﬂat calibration curve. Many of these radiocarbon ages are suspected to include a part due to inbuilt time (i.e., age of the wood at the time of burning), transport time, and reworking processes, which are dif ﬁcult to resolve. Considering these ages at their face value could lead to dates older than the actual earthquake dates. We suggest that a part of this chronological bias is also related to a local postseismic aggradation pulse of 4 to 5 m of sediments, which is documented in the trench and terraces. This ﬂuvial sequence, hiding the most recent surface rupture, is likely related to landslide‐sediment deposition triggered by the 1934 Bihar ‐Nepalearthquake.
This study has combined multi-objective optimisation with a poste- riori preference articulation approach to support decision-makers (DMs) in a sudden-onset disaster response. Our research has targeted locating temporary relief distribution centres (RDCs), which has been referred to as one of the first critical decisions for an effective and efficient response [ 9 ]. A location-allocation model has been adapted to minimise uncov- ered demand, response time, and logistics costs, and this research offers an approach based on Monte Carlo Simulation to facilitate group decision-making by investigating the stability of the non-dominated alternatives. Our approach supports finding the tipping points which can assist DMs to focus group discussions and converge to a consensus quicker. The proposed methodology has been applied to a real dataset from UN WFP’s operations after the 2015 Nepalearthquake and then validated by representatives from humanitarian organisations (HOs) through a small experiment.
Soma Nath Sapkota 1* , Laurent Bollinger 2 and Frédéric Perrier 3
Large Himalayan earthquakes expose rapidly growing populations of millions of people to high levels of seismic hazards, in particular in northeast India and Nepal. Calibrating vulnerability models specific to this region of the world is therefore crucial to the development of reliable mitigation measures. Here, we reevaluate the >15,700 casualties (8500 in Nepal and 7200 in India) from the M w ~8.2, 1934, Bihar–Nepalearthquake and calculate the fatality rates for this earthquake using an estimation of the population derived from two census held in 1921 and 1942. Values reach 0.7–1 % in the epicentral region, located in eastern Nepal, and 2–5 % in the urban areas of the Kathmandu valley. Assuming a constant vulnerability, we obtain, if the same earthquake would have repeated in 2011, fatalities of 33,000 in Nepal and 50,000 in India. Fast-growing population in India indeed must unavoidably lead to increased levels of casualty compared with Nepal, where the population growth is smaller. Aside from that probably robust fact, extrapolations have to be taken with great caution. Among other effects, building and life vulnerability could depend on population concentration and evolution of construction methods. Indeed, fatalities of the April 25, 2015,
Table 4 , summarizes key points of this section. 5. Suggestions of propositions
Although this research is of Nepalearthquake responders, our ob- servations and ﬁndings of in-country transportation risks in HSC also provide general indications for other sudden onset natural disasters. Our rationale refers to, ﬁrst, the existence of several common char- acteristics in relief operations contexts after sudden onset natural dis- asters. Second, our interviews encompassed organizations of many diﬀerent sizes, capabilities, and infrastructures that work in various regions worldwide. And third, the context of Nepal entailed several challenges (aftershocks, landslides, landfalls, ﬂoods, monsoon season) for HOs and their supply chains. Hence, any robust solution that could deal with these challenges can be one of the best practices for future relief operations. In this section, we will discuss the implications of our ﬁndings in three propositions for the use of HOs and their decision makers.
25 April 2015 NepalEarthquake
Landslide Intensity (revision 2.0 - 7 May 2015)
Description of landslide features: This map has been compiled from optical satellite imagery across the area that experienced shaking during the earthquake, available up to 7 May 2015. Approx. 3,300 landslides have been identified and mapped marking the landslide location. All landslides shown are either new landslides triggered by the earthquake, or those which have been reactivated by the earthquake. The main map shows landslide distribution. The purpose of this inventory and map is to describe the overall spatial distribution of landsliding triggered by the earthquake, and not for site-specific assessment. Image qulaity is low in steep terrain meaning precise landslide locations may be inaccurate by up to 100 m. Key rivers, valleys and roads are labelled, and the yellow star indicates the epicentre of 25 April 2015 M7.8 earthquake. Landslide are now data available via: https://data.hdx.rw- labs.org/group/nepal-earthquake.
3 National Seismological Centre, Department of Mines and Geology, Lainchaur, Kathmandu, Nepal
Accepted 2016 September 22. Received 2016 September 21; in original form 2016 March 31
S U M M A R Y
The depth of 61 aftershocks of the 2015 April 25 Gorkha, Nepalearthquake, that occurred within the first 20 d following the main shock, is constrained using time delays between teleseismic P phases and depth phases (pP and sP). The detection and identification of these phases are automatically processed using the cepstral method developed by Letort et al., and are validated with computed radiation patterns from the most probable focal mechanisms. The events are found to be relatively shallow (13.1 ± 3.9 km). Because depth estimations could potentially be biased by the method, velocity model or selected data, we also evaluate the depth resolution of the events from local catalogues by extracting 138 events with assumed well-constrained depth estimations. Comparison between the teleseismic depths and the depths from local and regional catalogues helps decrease epistemic uncertainties, and shows that the seismicity is clustered in a narrow band between 10 and 15 km depth. Given the geometry and depth of the major tectonic structures, most aftershocks are probably located in the immediate vicinity of the Main Himalayan Thrust (MHT) shear zone. The mid-crustal ramp of the flat/ramp MHT system is not resolved indicating that its height is moderate (less than 5−10 km) in the trace of the sections that ruptured on April 25. However, the seismicity depth range widens and deepens through an adjacent section to the east, a region that failed on 2015 May 12 during an M w 7.3 earthquake. This deeper seismicity could reflect a step-down of the
1%. We find that most surface loading phenomena are either too small, or have the wrong polarity to enhance winter seismicity. We consider enhanced Coulomb failure caused by a pore-pressure increase at seismogenic depths as a possible mechanism. For this to enhance winter seismicity, however, we find that fluid diffusion following surface hydraulic loading would need to be associated with a six- month phase lag, which we consider to be possible, though unlikely. We favor instead the suppression of summer seismicity caused by stress-loading accompanying monsoon rains in the Ganges and northern India, a mechanism that is discussed in a companion article. Citation: Bollinger, L., F. Perrier, J.-P. Avouac, S. Sapkota, U. Gautam, and D. R. Tiwari (2007), Seasonal modulation of seismicity in the Himalaya of Nepal, Geophys. Res. Lett., 34, L08304, doi:10.1029/ 2006GL029192.
(2007). Hence, we suggest that along its southern termination the Tsetserleg rupture likely did not reach the surface, although it propagated a longer distance than what it indicated by surface rupture solely.
Detailed mapping of the 1905 T-B EQs rupture based on the HRS imagery provides us with a revised set of parameters for the coseismic ruptures. The average horizontal slip and rupture length are, respectively, 2.34 ± 0.42 m over at least 114 km for the Tsetserleg rupture, 6.37 ± 0.95 m over ~388 km for the Bulnay rupture, and 2.90 ± 0.63 m over ~80 km for the Teregtiyn rupture. For the main Bulnay rupture, more speci ﬁcally, we obtain comprehensive slip distribution, geometric segmentation, and the type and size of steps and bends associated with the segment geometry. This leads us to estimate the moment magnitude for the major fault segments. Our results show that beyond local apparent complexity, the surface rupture associated with the sequence is consistent at the rupture scale. Actually, we could demonstrate that location of off-fault branching, mostly along the northern side of the rupture, is directly related to the direction of rupture propagation, enhanced by some structural imprints in the regional geological units. Along the Bulnay rupture, off-fault damage related to rupture propagation is mostly found along the western part of the rupture, while the eastern part of the rupture is more localized. Accordingly, the average slip along the western part of the rupture is smaller by ~2 m compared to average slip along the eastern part. It is found, however, that where it is possible to measure slip on secondary deformation, such as along the western part of the rupture, the slip reaches 1 to 2 m, which added to the slip on the main rupture matches average slip along the eastern part of the rupture. Talking this into account, the slip distribution is rather even along the entire rupture. Because the difference in slip along the main rupture is associated with the difference in off- fault damage, we suggest that any affect of the 9 July Tsetserleg rupture on the variability of the Bulnay slip distribution is minor. Instead, we suggest that local variability in the geological units along the western part of the rupture strongly favors bimaterial behavior, which enhances off-fault damage, and consequently impacts the slip distribution. Hence, although detailed geologic mapping of areas affected by large earthquakes is not always available, our results show that it proves to be a useful addition for understanding earthquake rupture patterns. Unlike variation of frictional properties or stress heterogeneities on the fault plane, geological information is more directly observable and it should be more systematically incorporated into earthquake source process analyses.
The aim of our work was to retrieve the main parameters of the rupture process of this earthquake from seismograms recorded at local and regional distances (20–300 km). To eliminate path and site e ffects from the seismograms, we compared the main shock recordings at each station with those of the largest aftershocks nearby. We used a combination of techniques, including pulse-width measurements and cross-correlation of velocity traces, comparison of P-wave displacement pulses, and empirical Green’s function deconvolution, to retrieve the apparent duration of the rupture process as seen at each station. Our results demonstrate that, in the absence of on-scale data, P-wave pulse-width measurements on clipped signals can be misleading if the rupture process is complex. In the case of the Annecy earthquake, comparisons of on-scale P-wave displacement seismograms and the empirical Green’s function deconvolutions show that the rupture process consisted of at least two subevents separated by 0.2–0.3 s, and with a total duration of about 0.5 s. The systematic azimuthal dependence of both the shape and duration of the apparent source-time function is consistent with a nearly unilateral propagation of the main rupture phase in a southeast direction along the fault plane and parallel to the direction of slip. An isochron analysis reveals that the first subevent occurred slightly to the northwest of the nucleation point but that the second subevent was located further to the southeast, thus confirming the overall rupture directivity towards the southeast. An interpretation of our results in light of the previously documented aftershock distribution and of observations of ground cracks in the epicentral area suggests that the main shock occurred on the Vuache Fault, and that rupture in a northwest direction was inhibited by a right-lateral stepover in the fault. Accordingly, the vast majority of the subsequent aftershocks, which include several magnitude 3–4 events, occurred on a fault segment that is slightly offset from the inferred surface trace of the Vuache Fault and that was activated by the main shock.
[ 19 ] However, the half-year delay between rainfall at the
surface and increased pore pressure at seismogenic depth would be a surprising coincidence. In addition, given the wide range of hypocenter depths (5 to 15 km), the phase delay between forcing at the surface and earthquake trig- gering should vary a lot. The locations of microearthquakes are insufficiently well resolved to test this possibility, and there have been insufficient larger events with better depth resolution to subject this to a rigorous test. However, no noticeable depth dependence with time has been observed in the data. Consequently, the diffusion of pore pressure, although possible, does not appear to us probable. We favor the direct mass loading induced by the summer monsoon as the dominant mechanism driving the seasonal variations of seismicity reported in this study.
4.1 Direct, Peer Level and Indirect Exposure
In the 1999 follow-up round of the Nepal Social Network Survey, respondents were asked to ”list the full names of those women in your village with whom you have discussed family planning in the last six months”. This piece of information is used to define whether a link exists and to trace the network architecture for each of the three villages. The links are assumed to be reciprocal for two reasons: first, given the high specificity of the question posed to respondents, and given that interaction among women in the village is frequent and complex, women are more likely to forget about having discussed FPC issues with one of their friends rather than misreporting to have had discussions if this was not the case. Second, women could mention no more than 5 partners, therefore some well-connected women may have been forced to mention only their principal discussion partners rather than all of them. I thus assume links to be unweighted and undirected, and every time a woman mentions another one I draft a communication link between them. 1 With this procedure 310 links among the 337 women are identified overall. The resulting network is rather sparse, with 71 isolated women out of 337, an average number of links equal to 1.8. The geodesic distance (that is, the number of steps in the shortest path between two women) has a mean value of 4.45 and a maximum value of 11 steps. For a graphical representation of the three villages’ networks see Figure A.1, Appendix.
sition of the forest. Trees can be heavily lopped. Köhlin and Parks (2001) also discuss the implications of tree species choice in reforestation campaigns in India where plantations can target trees producing fodder and firewood or belonging to species producing good timber but that are not useful as household fuel. Dif- ferences in the quality of the wood biomass can actually have non trivial impact in terms of respiratory health for households as explained by Jagger and Shively (2014) in Uganda. The fact that collection times rose 12% in Nepal could reflect a process of deforestation which is not picked up by aerial satellite images or by a progressive change in tree species, something hard to measure by large spectrum remote sensing. More detailed on-the-ground studies are needed to evaluate this possibility. Some of the rise in collection times can however be explained by the growing role of community forest groups. Note also that this issue does not affect the second and third main findings described above.
Laboratory. Here we only address the part of the Kaik ōura rupture that concerns the Jordan-Kekerengu- Papatea triple junction.
Based on satellite imagery and ﬁeld observations, we ﬁrst deﬁned the fault geometry in the triple junction area and discretized the domain by using an unstructured triangular mesh around the prescribed faults (Figure S4). The mesh size is adaptively controlled to be ﬁner close to the fault to optimize trade-off between the numerical accuracy and computational cost. We then de ﬁned the initial stress state σij uni- formly in the medium. The angle of σ1 , the maximum principal compressive stress was set to N107°, to be both compatible with the sense of slip on the different faults activated during the earthquake, in a range of N105° –N115°, and with regional focal mechanisms (Townend et al., 2012). It is assumed that the material around the faults has been previously damaged (i.e., weakened) and therefore is less compe- tent than the rest of the material in the model. The areas of weakened material are highlighted in yellow in Figure S4a. The introduction of this weakened material area also restricted unrealistic crack propagation at the edge of faults, which could be generated due to the simple friction model (friction slip-weakening law) used here. The FDEM allows for tensile, shear, and mixed-mode crack represented as the break of cohesion at the boundary of the ﬁnite elements. In other words, each boundary of a ﬁnite element is a potential failure plane. To avoid numerical bias in the orientation of cracks, the orientations of the poten- tial failure planes are kept isotropic (Figure S4b).
zones appear as a complex system where ﬂuid circulation, crustal permeability, and possibly earthquake occurrence might be interrelated dynamically [Manning and Ingebritsen, 1999; Ingebritsen and Manning, 2010; Manga et al., 2012].
The Himalayas offer a natural laboratory where this essential coupling can be studied. High seismic activity is concentrated on a midcrustal ramp located below the Main Central Thrust (MCT) zone on the Main Himalayan Thrust accommodating the 2 cm yr 1 convergence between India and Southern Tibet [Avouac, 2003; Ader et al., 2012], where ﬂuid occurrence might explain the high electrical conductivity observed by magnetotelluric sounding [Lemonnier et al., 1999]. Seasonal variations of seismicity [Bollinger et al., 2007] and deformation [Bettinelli et al., 2008; Chanard et al., 2014] can be related to surface hydrological forcing. Evidence of CO 2 release exists in the MCT zone of central Nepal. First, high alkalinity of hot springs up to
- Clustered seismicity develops at the intersection between the megathrust and contacts between Lesser Himalayan tectonic slivers.
- Seismic swarms migrate at 50 m/day at the transition zone.
- The local earthquakes reveal ramps and flat geometry of the megathrust highlighting a new segmented fault model for West Nepal.