WIHG studied this disaster using seismological data, satellite imagery and numerical modelling and estimated that \(27\times 10^{3}\ m^{3}\) of rock and glacier-ice collapsed from the steep north face of the Ronti Peak (Shugar et al., 2021; Tiwari A., et al., 2022). The main rock-fall followed by a noteworthy sequence of small events recorded by nearby seismic stations. The main event appears to have been initiated by precursory signals for nearly 2:30 h. The seismic data of three nearby stations within 50 km distance also distinguished debris flow and hitting obstacles from other seismic sources. The proximal high-quality seismic data allowed estimation of debris-flow speed and reconstruction of the complete chronological sequence, from the initiation of the nucleation phase to the occurrence of the rock-fall.

8.2 Earthquake Precursory Studies

Vijaya Kumar et al. (2020) recognized pre- and co-seismic signatures in MT data for the Mw 4.6 earthquake that occurred on 2007-11-24 in the Koyna-Warna region. Wavelet analysis of the MT time series data shows significant enhancement at 3–6 Hz frequency band in the scalogram during the earthquake in comparison with pre- and post-time. The spectral polarization ratio technique was implemented on these events to identify the precursory signatures. A few days before the earthquake, a significant anomaly was identified for most of the earthquakes using this technique. Akilan et al. (2021) studied changes in the zenith total delay (ZTD) and total electron content (TEC) associated with the 2015 Nepal earthquake (Mw 7.8). They analyzed the ZTD derived from GPS data received at HYDE, LCK3 International GNSS Service (IGS) stations and MSUN operated by CSIR-NGRI. The decrease in ZTD and TEC suggests hydration in the atmosphere by the joining of ions in the atmosphere during earthquakes. Almost the same results regarding the changes in ZTD and TEC were obtained for another large event (the 1999 Chamoli earthquake of Mw 6.8) in the Himalayan region, although they differ slightly in intensity depending on the observation distance.
    Probabilistic analysis was performed on the seismic data of 100 years (1918–2018) for forecast of probable future earthquakes above Mw \(\ge\) 5.0 in NE India (20°–30°N and 86°–98°E) and its vicinity to ascertain mean occurrence period E(t) for earthquakes of Mw \(\ge\) 5 (Chetia et al., 2019a). Here, Kolmogorov–Smirnov statistics constrained by Weibull distribution has been utilized to achieve the best fit on the dataset. E(t) is found to be \(\sim\) 74 days with 50% probability. Similarly, cumulative probability function indicates a time of 140 days with 80% probability, while 400–500 days of recurrence time period is embedded with 90–100% probability for an earthquake of Mw \(\ge\) 5.0 to recur following the occurrence of the last earthquake.
    Mechanical deformations from within the earthquake preparation zones are believed to cause seismo-electromagnetic (SEM) emission in ultra-low frequency (ULF) band, i.e. between 0.001 and 10 Hz. The 3-component ULF induction coil magnetometer data from Multi-parametric Geophysical Observatory (MPGO), Tezpur were used to study SEM emissions employing both polarization ratio analysis and fractal analysis during the campaign period of Apr.20 – Sep.3, 2019 (Dey et al., 2021). The findings show candidate SEM emissions, in the form of enhancements in SZ/SH, associated with all the seven credible events, even as nine enhancements could not be attributed to immediately adjacent credible events.
    WIHG operates an MPGO at Ghuttu (Tehri, Uttarakhand) for earthquake precursory studies. Inert gas (soil Radon) and magnetic field changes before the occurrence of small-to-moderate magnitude earthquakes provide evidence for short-term (months-to-days, hours) earthquake precursors. However, this remote site in the Higher-Himalaya, away from human generated noise, indicates that there are different background natural noises to make the data very complex. There is a very high hydrological effect in the gravity and radon emanation (Shukla et al., 2020; Chauhan et al., 2021). After removing the background noises, anomalous changes are reported in case of 19-20 moderate magnitude local earthquakes. Earlier, this observatory has also reported precursory changes during the Mw 7.8 Nepal earthquake of 2015.
    ISR has established three MPGOs in Gujarat for precursory studies. Soil radon (Rn-222) data of the Badargadh station were used to identify precursory signal of two earthquakes of M3.7 and M4.2 which occurred on March 26, 2011, and May 17, 2011, respectively through advanced processing of the time series (Sahoo et al., 2020, 2021). The ultra-low frequency geomagnetic variations were observed before the Dholavira earthquake (M 5.1) of 2012-06-20 in the Kachchh region (Joshi and Rao., 2021).
    The apparent resistivity imaging at MPGO, Tezpur, operated by CSIR-NEIST, was carried out since 2016-08-31 for fixed interval time of 3 days to investigate precursory signatures prior to earthquake events. Anomalies in apparent resistivity prior to earthquake events are observed. These are not influenced by the rainfall activity in the region during the investigation period. Weibull distribution technique with observed apparent resistivity data is adopted to look for and revalidate the observations. The Weibull parameters, i.e. K and m, are estimated to be 0.26 and 1.04, respectively. No precise relation between magnitude and earthquake precursory time is found. Weibull probability distribution indicates that the probability of an earthquake occurrence exceeds the measure up to 80% (9 days) after the precursory signal is observed (Chetia et al., 2020). Similarly, temporal variability of the soil radon emanations measured at the MPGO was scrutinized using singular spectrum analysis (SSA) (Chetia et al., 2019b). The study concludes that SSA eliminates diurnal and semidiurnal components from time series of soil radon emanation for better correlation of soil radon emanation with earthquakes.
    Mukherjee et al. (2021) have proposed a novel approach for Earthquake Early Warning (EEW) System Design using deep learning Techniques. The method converts a seismic signal into audio signals and then uses popular speech recognition techniques. Both Convolutional Neural Network (CNN) and a Long Short-Term Memory (LSTM) network have been trained.

8.3 Ionospheric Seismology

Ionospheric seismology refers to the study of events of earthquakes and tsunamis through co-seismic ionospheric waves produced by the dynamic coupling between the Earth’s surface and the atmosphere. The ionosphere is a highly dynamic ionized region of the Earth’s upper atmosphere extending from \(\sim\) 60 km to \(\sim\) 1000 km. The origin of any perturbations in the ionospheric electron density can be traced to various sources either from above (e.g. solar, geomagnetic, etc.) or below (e.g. lower atmospheric, earthquakes, tsunamis, volcano eruptions, etc.) the ionosphere. In particular, short-period acoustic waves and long-period gravity waves emitted by earthquakes and tsunamis propagate upward in the region of exponentially decreasing atmospheric neutral density, and thus, their amplitude increases with atmospheric heights. On arrival at ionospheric heights, the waves redistribute ionospheric electron density and produce electron density perturbations known as co-seismic/co-tsunami ionospheric perturbations, respectively. Recently, it has been suggested that ionospheric signals produced by earthquakes and tsunamis can be inverted to infer the seismic source characteristics of large earthquakes and to envisage the propagation time and amplitudes of tsunami waves. It seems possible, but there are difficulties in terms of non-tectonic forcing mechanisms that act upon the ionospheric perturbation evolution at ionospheric altitudes (Bagiya et al., 2019).
    A simple and direct 3D model is developed to estimate the combined effects of nontectonic forcing mechanisms of i) orientation between the geomagnetic field and tectonically induced atmospheric wave perturbations, ii) orientation between the GNSS satellite line of sight (LOS) geometry and coseismic atmospheric wave perturbations, and iii) ambient electron density gradients on the manifestations of Global Positioning System (GPS) – Total Electron Content (TEC) measured near field co-seismic ionospheric perturbations. This model can compute the nontectonic effects at various ionospheric altitudes depending on the propagation characteristics of seismo-acoustic rays (Fig.12). Further, this model is tested on earthquakes occurring at different latitudes. It is presumed that this model would induce and enhance a proper perception among the researchers about the seismic source characteristics derived based on the corresponding ionospheric manifestations (Bagiya et al., 2019).
    Further, GPS-TEC observations provide adequate information on the spatial and temporal characteristics of earthquake induced ionospheric perturbations. However, one of the major limitations of this technique is the lack of altitude information of the recorded ionospheric perturbations. GPS derived TEC is an integrated quantity; hence it is difficult to relate the detection of ionospheric perturbations in TEC to a precise altitude. Using the modelled propagation of acoustic rays in space and time and their interaction with satellite-station line of sight (LOS) geometry, a novel method has been developed to infer the detection altitude of ionospheric perturbations observed through GPS-TEC. This modest method has been further upgraded to identify the distinct seismic sources that evolved along an extended rupture varying simultaneously in space and time akin to the seismic rupture of the Mw 9.0 2011-03-11 Tohoku-Oki earthquake (Bagiya et al., 2020).
    In addition to the transient perturbations, prolonged ionospheric oscillations following large earthquakes (Mw > 8.0) have also been found to provide seismic source information from the ionosphere. Such ionospheric oscillations related to the earth-atmospheric resonance frequencies of \(\sim\) 3.7 and \(\sim\) 4.4 mHz following the 2012-04-11 Sumatra doublet (Fig.13) and 2011-03-11 Tohoku-Oki earthquakes were scrutinized (Nayak et al., 2021, 2022). The Earth’s background free oscillations at \(\sim\) 3.7 and \(\sim\) 4.4 mHz resonantly couple with the atmospheric acoustic modes and thus energy cross-talk between the earth-atmosphere system is maximum at these frequencies. Our studies emphasized that resonant ionospheric signatures during the Sumatra doublet event were related to the seismic source. Therefore, resonant co-seismic ionospheric signatures could provide additional information on the low frequency features of seismic ruptures.