Refining the Concept of Earthquake Precursory Fingerprint

Abstract

The recently proposed concept of “precursory fingerprint” is a logical consequence of the commonsense statement that seismic structures are unique and that their expected preshock behaviors, including precursory phenomena, are also unique. Our new prediction-related research strategy is conceptually based on the principles of (1) the uniqueness of seismogenic structures, (2) interconnected and interacting geospheres, and (3) non-equivalence of Earth’s surface spots in terms of precursory signal receptivity. The precursory fingerprint of a given seismic structure is a unique assemblage of precursory signals of various natures (seismic, physical, chemical, and biological), detectable in principle by using a system of proper monitoring equipment that consists of a matrix of n sensors placed on the ground at “sensitive” spots identified beforehand and on orbiting satellites. In principle, it is composed of a combination of signals that are emitted by the “responsive sensors”, in addition to the “non-responsive sensors”, coming from the sensor matrix, monitoring as many virtual precursory processes as possible by continuously measuring their relevant parameters. Each measured parameter has a pre-established (by experts) threshold value and an uncertainty interval, discriminating between background and anomalous values that are visualized similarly to traffic light signals (green, yellow, and red). The precursory fingerprint can thus be viewed as a particular configuration of “precursory signals” consisting of anomalous parameter values that are unique and characteristic to the targeted seismogenic structure. Presumably, it is a complex entity that consists of pattern, space, and time components. The “pattern component” is a particular arrangement of the responsive sensors on the master board of the monitoring system yielding anomalous parameter value signals, that can be re-arranged, after a series of experiments, in a spontaneously understandable new pattern. The “space component” is a map position configuration of the signal-detecting sensors, whereas the “time component” is a characteristic time sequence of the anomalous signals including the order, occurrence time before the event, transition time between yellow and red signals, etc. Artificial intelligence using pattern-recognition algorithms can be used to follow, evaluate, and validate the precursory signal assemblage and, finally, to judge, together with an expert board of human operators, its “precursory fingerprint” relevance. Signal interpretation limitations and uncertainties related to dependencies on sensor sensibility, focal depth, and magnitude can be established by completing all three phases (i.e., experimental, validation, and implementation) of the precursory fingerprint-based earthquake prediction research strategy.

1. Introduction

“Earthquake prediction is difficult but not impossible” [1]

The official and generally accepted and applied paradigm concerning precursor-based earthquake prediction that resulted from the “Nature debate” in the 1990s (e.g., [2]), which was then adopted by the USGS and reinforced in an influential book [3], is that it is “unpredictable in principle” due to its intrinsic nonlinear and chaotic character [4]. However, results obtained by a number of researchers worldwide during the last few decades [5,6] strongly suggest that this radically pessimistic paradigm is no longer sustainable. The achieved progress in this research field is illustrated by a number of well-documented “post-predictions” (i.e., the post-factum identification of relevant precursory signals), and even pre-event predictions by using various methodologies (e.g., [7,8,9,10,11,12,13]). However, the pre-event predictions were not popularized on social media nor delivered to authorities for warning purposes.

Multidisciplinary and multiparametric studies of short-term earthquake forecasting are now being undertaken. As such, the results published during the last few decades in numerous countries have intended to follow a “holistic approach” by recognizing that “a number of geophysical and geochemical measurements obtained by ground based or satellite-based approaches, ranging from ground associated deformation patterns to pre seismic alterations, may be related to stress variations in the lithosphere before to an eventual big earthquake” as stressed by [6]. For example, multiparameter data recording during the M7.8 and M7.3 earthquakes in Nepal included six different atmospheric and ionospheric physical parameters, all measured by satellite-held sensors [14]. On the other hand, ground-based seismo-geochemical research intending to detect “multi-anomalies” conducted in SW Taiwan [15] and Northern Italy [16]) presents another example of limited-range multiparameter monitoring of earthquake precursors. Multiparameter soil gas (such as noble gas Rn, Hg, H₂, CO₂, and He) monitoring in China [17] is another such example. A combination of satellite- and ground-based multiparameter research was also attempted recently, measuring just a small number of parameters by “hydrodynamic and magnetic monitoring” in Georgia [18].

One may note that all of these promising results were obtained in particular cases by using a limited set or one single type of measured parameter (e.g., atmospheric–ionospheric or various gas species), and none of them can be generalized in the form of a generally applicable seismic-prediction methodology. “Only few studies were truly multi-disciplinary”, as Matsumoto and Koizumi [19] stated. These authors used continuous multiparameter fluid sampling and measured the fluid geochemical composition and CO₂ isotopic signatures in three observation wells that were equipped with seismometers, groundwater level and temperature sensors, as well as tilt and strain meters. Although “multisensory” and “multiparameter” in its scope and design indeed, this approach is still limited to ground-based techniques and the monitoring of a single potentially precursory signal-carrying medium (groundwater).

The concept of “precursory fingerprint” was recently proposed and introduced [20] to consider and promote a new possible strategy in precursor-based earthquake prediction research. The basic line of reasoning in this proposal is that rather than searching for the Holy Grail of a universally valid and applicable prediction methodology and related monitoring technology and equipment, a less ambitious but more realistic goal should be pursued. This goal is based on the widely accepted finding that each seismogenic structure is unique. Therefore, it is reasonable to expect that their related precursory phenomena and signals—i.e., their precursory fingerprints—are also unique and cannot be replicated elsewhere.

The lessons learnt from the extremely meaningful succession of seismic events and their aftermath in China in the 1975–1976 period were not exploited in terms of earthquake prediction philosophy. The much-popularized world-first success of the Haicheng 4 February 1975 earthquake prediction that was based on outstandingly obvious and prominent precursory signals and evacuation of local residents resulted in the saving of many lives [21]. In contrast, the Tangshan earthquake striking one year later on 28 July did not exhibit such obvious precursors and resulted in more than 242,000 deaths (e.g., [22,23]). The evident conclusion that was not drawn from these events is that no two seismic structures are identical; therefore, no identical precursory behavior can be expected. If we are to extrapolate, we can conclude that no universal precursors do exist.

Recognizing this evidence, one may understand why the search for universal earthquake precursors failed so far despite the worldwide investigation efforts over the last half century. Researchers tackling the problems of precursor-based earthquake prediction topic should consider the need for a concept-based new paradigm-shifting strategy. Szakács’s paper [20] proposes such a possible approach based on the newly introduced concept of “precursory fingerprint”. This follow-up paper intends to detail and refine the precursory fingerprint concept in an attempt to develop the previous approach towards a realistic earthquake prediction research strategy.

2. Conceptual Background of the Precursory Fingerprint

The starting premise of any discussion related to precursor-based earthquake prediction is the explicit recognition of the existence of precursory phenomena—i.e., anomalous changes in the physical, chemical and biological parameters of the environment prior to seismic events. The majority of the seismological community acknowledges this, at least in theory (e.g., [1,6]) as stressed by Szakács [20]. Many scientists also acknowledge that these phenomena are detectable in principle using adequate monitoring techniques. Pulinets and Ouzounov [24], for example, persuasively argue that there are a number of ground-surface, atmospheric, and ionospheric precursors detectable in principle. They even propose a workable methodology based on remote-sensing technology. Likewise, a number of other researchers (e.g., [1,20,25]) express similar views supporting this approach.

The primary question is, however, whether these precursory phenomena, or at least a subset of them, manifest universally prior to earthquakes or if they are specific to particular seismic structures or events. The advocates of the dominant “impossible in principle” paradigm reject the first hypothesis on theoretical grounds, but they also deny the practical possibility of reliable precursor detection of any earthquake. Conversely, members of the “optimistic party” hypothesize the existence of precursors that are believed to occur universally. This premise is postulated to enable their detection in principle. In their holistic approach, Pulinets and Ouzounov [24] and Pulinets et al. [5] even envisaged a methodology based on satellite-borne sensors for the global detection of earthquake preparation areas. On the other hand, Szakács [20] is profoundly skeptical regarding the existence of universal precursors proposing, instead, an alternative research strategy based on the concept of the “precursory fingerprint” of individual seismic structures.

Having postulated the existence of precursory phenomena and of the related signals, we now turn to discussing the basic principles on which a possible research strategy focused on the concept of precursory fingerprint might be elaborated. Three principles are considered in this paper: (1) the principle of integrated and interacting geospheres (lithosphere, hydrosphere, atmosphere, ionosphere, biosphere), (2) the principle of the uniqueness of seismogenic structures and their precursory fingerprint, and (3) the principle of the non-equivalence of Earth surface spots in terms of signal receptivity.

The otherwise trivial and well-known principle of intercorrelated geospheres is at the very core of the approach of Pulinets et al. [5] and Pulinets and Ouzounov [24]. These authors originally valorized it in their quest for universal earthquake precursors and their detection. According to this principle, the Earth’s geospheres (comprising the lithosphere, hydrosphere, atmosphere, ionosphere, and biosphere) interact with each other in a complex way so that changes in one of them trigger changes in the others by way of propagation of the change-related signals from one geosphere to another across the entire Earth system. For example, changes in the earthquake preparation area in the lithosphere related to escalating and irreversible deformation processes leading to a seismic event, will propagate as signals of physical and/or chemical nature from the source to the surface. These signals will induce changes in the atmosphere, in turn propagating upwards and inducing further changes in the outermost geospheres, the ionosphere, and the magnetosphere. According to Pulinets and Ouzounov [24] and Pulinets et al. [5], the change-related signal carrier is the radon gas released from the rocks in the earthquake preparation zone of the lithosphere. This gas reaches the Earth’s surface where it ionizes the atmosphere, provoking an increase in temperature and a decrease in humidity locally. It then propagates upwards within the ionosphere inducing anomalies in its parameters such as, for example, the total electron content. Furthermore, it is imperative to stress the importance of alterations in the lithospheric physical fields (e.g., the electromagnetic field) which transcend those of radon. These alterations induce further changes within the coupled geospheres such as the hydrosphere (i.e., groundwater), biosphere (sensitivity of the living matter), and atmosphere in the form of induced precursory signals of various natures (physical, chemical, and biological). To summarize, the principle of coupled and interacting geospheres provides a robust theoretical foundation for precursor-based research strategies.

The principle of the uniqueness of seismogenic structures, as introduced and discussed by Szakács [20], and its logical consequence, the uniqueness of their precursory behavior, forms the basis of the precursory fingerprint concept. This concept originates from the empirical observation that different seismic events triggered at different locations, i.e., belonging to different seismic structures, are preceded and accompanied by different phenomena, as learnt from the 1975–1976 events in China, discussed in the Introduction. In essence, it states that there are no identical seismic structures; therefore, no identical precursors can be expected. In other words, each seismic structure has its own unique precursory fingerprint, distinct from any other structure. This principle forms the basis of a possibly paradigm-shifting research strategy whereby searching for the precursory fingerprint of individual seismic structures is considered a more realistic approach than searching for universal precursors [20].

The principle of the non-equivalence of Earth surface spots in terms of signal receptivity is invoked here in order to maximize the chances of identifying the precursory fingerprints of particular seismic structures to be used for prediction purposes. This principle emerges from the empirical observation that Earth’s surface areas are not equally suitable for the deployment of scientific equipment intended for lithosphere-originated signal detection. It is notable that that certain locations exhibit enhanced signal reception capabilities (i.e., being more “sensitive”) than others. This aspect related to seismic prediction was emphasized by Martinelli and Dadamo [26] and Martinelli [27]. For example, “site-specific precursors” were investigated by Vasilev et al. [28] and Bulusu et al. [29] for particular cases. The selection of the “optimal location for monitoring boreholes” is a necessity according to Zhang et al. [30]. The primary rationale for the non-equivalence of Earth surface spots is attributable to the fact that, in order to reach the surface, any possible precursory signal generated in the earthquake preparation volume within the lithosphere has to travel through a kilometers- or tens of kilometers-thick stack of rock bodies of various composition and structures separated from each other by a plethora of discontinuous surfaces, each of them able to reflect and refract physical signals of wavy nature, as well as to deflect, impede, or even block signal-carrying ascending fluids. In such circumstances, crust-transecting structures such as volcanic conduits and deep fracture zones may serve as privileged signal transmission paths. These structures can be regarded as inverted antennas functioning as wave-guides, resulting in minimized information energy loss and maximized signal to noise ratio at their surface intersection areas. These areas are therefore considered optimal for the location of monitoring equipment [31].

3. Precursory Fingerprints

3.1. The Concept

Following the work of Szakács [20], the precursory fingerprint of a specific seismic structure can be conceptually defined as a unique and characteristic assemblage of precursory signals of diverse natures (physical, chemical, and biological) that originate from earthquake preparation processes entering their final and irreversible phase, eventually leading to the onset of a seismic event. Its objective existence is a logical consequence of the principle of the uniqueness of seismic structures. Instead of being represented by one single type of pre-earthquake signals of a “key” or “diagnostic” precursor (e.g., [32]), the precursory fingerprint is rather an entity comprising multiple precursory signals of various types. Its predictive value depends on the ability to capture the inherent complexity of the system under investigation. The uncertainty related to “key” signal-based precursors is significantly reduced when combinations of multiple signal types are considered. It can thus be concluded that the search for the precursory fingerprint of particular seismic structures is only compatible with multidisciplinary investigations using multisensor and multiparameter monitoring methodologies. Consequently, a concept-based research strategy can be developed and implemented with the objective of identifying and examining known individual seismic structures across the globe. Such a possible research strategy was already outlined in the paper of Szakács [20] without, however, discussing the precursory fingerprint concept in detail.

Multidisciplinary approaches in the field of short-term earthquake precursor identification are increasingly being applied by researchers worldwide (e.g., [33]). The process entails the continuous monitoring of a range of measurable parameters that are indicative of various pre-seismic phenomena. These phenomena are regarded or have been proven to represent precursory signals of an impending seismic event of potentially destructive magnitude. Time series of the measured parameters are produced by sensors included in ground-based and/or satellite-held monitoring equipment in order to detect significant pre-earthquake signals. There are numerous examples of such multidisciplinary and multisensor applications in earthquake precursor research, some of them referred to in the Introduction. However, it should be noted that all of these systems are limited in their “multidisciplinarity” whether in terms of the sensor holders (i.e., ground-based or satellite-based monitoring equipment) or in terms of the precursory signal type that is being monitored (i.e., chemical, physical). For example, Pulinets and Herrera [34] explicitly stated that “our approach is based on the use of physical precursors”. Even in the case of combined monitoring (i.e., by both satellite- and ground-based sensors and/or of chemical and physical parameters), the range of “multidiscipinarity” remains constrained to a limited number of items. It is noteworthy that there are extremely rare cases of research that can be considered to be “truly multi-disciplinary” [19] and which apply monitoring of both physical and chemical parameters by sensors at both ground-based and satellite-based equipment. Symptomatically, although the potential predictive value of biological sensors is theoretically acknowledged (e.g., [12,32]), none of the earthquake prediction-related results published in the new era of “multidisciplinary” earthquake prediction research includes parameters related to potential pre-seismic signals emitted by living creatures. Furthermore, the monitoring systems that have been implemented at various seismogenic structures worldwide, the results of which have been reported over the past few decades, are characterized by a high degree of methodological heterogeneity. In essence, there are no two applied multidisciplinary multisensor monitoring systems that are alike. This circumstance precludes the comparison of the applied monitoring approaches, of the results, and of the evaluation of their applicability elsewhere. In summary, there is no concept-based, generally accepted, unitary, multidisciplinary, and multisensor earthquake precursor research strategy accepted and followed by the scholarly community concerned with seismic forecasting.

In contrast, the proposed precursory fingerprint concept [20] allows for the establishment of a novel earthquake forecast-related research strategy of potentially paradigm-shifting relevance. Building on two easily understood principles (i.e., the principle of geospheres coupling and the principle of non-equivalence of Earth surface spots in terms of signal receptivity), this concept could provide a sound conceptual framework for future investigation efforts opening new perspectives. In particular, the advantages of adopting the proposed new conceptual framework based on the precursory fingerprint concept as a possible multidisciplinary research strategy are manifold. Chief amongst these is the fact that it considers (1) all possible types of precursory signals (physical, chemical, and biological) as opposed to a select few or a very limited number of signals of only physical and/or chemical nature, as the so far published results of multidisciplinary approaches show, (2) all signal-propagation media (lithosphere, hydrosphere, biosphere, atmosphere, and ionosphere, and (3) both ground-based and satellite-based monitoring systems, equipment and sensors. Indeed, a preliminary outline of a research strategy based on the precursory fingerprint concept was already proposed consisting of three stages—experimental/learning, validation and implementation [20].

3.2. Finding Precursory Fingerprints

We start by postulating that the signals that compose the precursory fingerprint of particular seismic structures are, in principle, detectable using adequate monitoring equipment. Such equipment would follow the time-dependent variation of a number of parameters, which are related to possible precursory signals of various natures (physical, chemical, biological). The purpose is to detect their pre-seismic anomalous behavior. Such a monitoring system can be depicted as a matrix of n sensors (Figure 1) installed both on the ground at pre-selected locations, designated as “sensitive” spots in the local area of the targeted seismic structure, and mounted on selected orbiting satellites periodically overflying the same area.

In the experimental phase of the research, the sensors are designated to monitor as many virtual precursory processes as possible. This is achieved by means of continuous measurement of the signal-relevant n + x parameters. It is reasonable to hypothesize that such a multisensor and multiparameter monitoring system will record a combination of signals emitted by “responsive sensors”, in addition to other “non-responsive sensors”, of the sensor matrix (Figure 2).

It is important to note that sensor “responsivity” is relative and contingent on factors such as (1) the sensitivity/detection limits and signal/noise accuracy, both probably improving in time with technological progress, and (2) the earthquake magnitude.

Each measured parameter has a threshold value T, which is pre-established by experts in the particular fields to which the monitored precursor belongs. In addition, there is an uncertainty interval (T ± x), which serves to discriminate between the background and anomalous values (Figure 3).

As shown by He et al. [35], the current measuring state of the responsive sensors can be visualized on the command board of the monitoring system in a manner analogous to traffic light signals (TLS): green for below T − x values, yellow for T ± x values, and red for above T + x values) (Figure 2 and Figure 3). Figure 4 shows a hypothetical configuration of an active sensor matrix at a certain time. The system itself has the capacity to be configured to display visually the following: background (green), uncertain (yellow), and alarm (red) levels of attention. A machine learning system can be constructed to monitor the changing time-dependent configuration of the system emitting pertinent messages whenever significant changes occur as recorded by the sensor matrix.

3.3. Expected Complexity of the Precursory Fingerprints

The precursory fingerprint of a given seismogenic structure can be viewed as a complex particular configuration of “precursory signals” (i.e., anomalous values of a number of measured parameters) showing a unique and characteristic pattern of the monitored physical, chemical, and biological signals recorded in the target area. Figure 4 shows a hypothetical pattern of measured parameter values, of which 21 are physical, 14 chemical, and 14 biological, with 44 recorded by ground-based equipment and 5 by satellite-borne sensors. Such sensor configurations constitute the “composition” or “pattern” component of a precursory fingerprint. Visually, it would appear as a particular arrangement of the anomalous signals from the responsive sensors on the master board of the monitoring system, which can be re-arranged, after a series of experiments, in a spontaneously readable and understandable pattern (for example, featuring the letter “V” for the Vrancea seismic nest in Romania).

The complex nature of the precursory fingerprints implies that, in addition to the “pattern”, they are also characterized by spatial and temporal components. The “space component” of the precursory fingerprint can be defined as the spatial configuration on the map of the anomalous signal-emitting sensors physically placed in the field at various “sensitive” spots within the monitoring area, whose actual location and extension are experimentally established in advance. The spatial distribution pattern of responsive (at various levels—green, yellow, and red) and non-responsive sensors (Figure 5) could be significant by itself when analyzing the “space component” of the fingerprint. The satellite-based monitoring equipment of sensors can be considered as representing the z dimension of the monitored 3D space.

The “time component” of the precursory fingerprint can be defined as the characteristic time sequence of the anomalous signals of different types (physical, chemical, biological) recorded by the sensors. It includes their order, timing before event (lead time), duration, transition time between the anomalous (yellow to red) signals (Figure 6), and any other significant time-related parameters that may be identified during experiments.

Fingerprint complexity can thus be identified as a key ingredient of earthquake prediction research. It can be expected that the higher the complexity of the identified precursory fingerprint, the higher its specificity to the targeted seismogenic structure and implicitly, the higher its actual predictive value. This general theoretical consideration does not imply that “simple-structure” precursory fingerprints are necessary absent and will not be found. In contrast, the “degree of complexity” itself can be viewed as part of the precursory fingerprint paradigm in earthquake prediction.

4. Discussion

It is evident that the practical implementation of any precursory fingerprint-based earthquake prediction research faces numerous technical and resource-related challenges. The operation of a monitoring system comprising numerous—e.g., tens of—sensors which, in turn, monitor an even higher number of measured parameters, is a prerequisite for the experimental (i.e., learning) phase of the project [20]. The continuous supply of numerical data at a sampling frequency of minutes, hours, and days for time intervals of months or years similarly requires a complex, sophisticated, and integrated data management system, including acquisition, format, transmission, storage, processing, systematization, and evaluation. Data volume management is thus a first-order challenge of such an endeavor. It appears impossible to address this issue without leveraging contemporary machine learning and artificial intelligence technologies that utilize pattern-recognition algorithms in order to follow, evaluate, validate the precursory signal assemblage and finally, enable the determination, together with an expert board of human operators, of its “precursory fingerprint” relevance.

Another category of challenges pertains to the evaluation of the predictive significance of the precursory signals that have been recorded. The assemblages of signals potentially considered as components of the precursory fingerprint of a particular seismogenic structure are inherently subject to uncertainties related to several factors. These include the sensor sensitivity and detection limits (e.g., [36]), sensor location (e.g., [27]), focal depth, and magnitude of the seismic events (e.g., [37]). Each of them and all together may influence the resulting precursory signal pattern (i.e., its composition and space and time “dimensions”), which needs further evaluation.

The quality and sensitivity of sensors are of paramount importance in detecting subtle changes in the physical and chemical environment of the precursory signal value. Signal-to-noise relationships are particularly critical in this respect; hence, the most advanced and currently available sensors should be used in the investigations. The focal depth-dependency of the popping up signal assemblage must also be given due consideration (e.g., [37]). The same sensors may record a number of anomalous precursory values in the case of shallow focus earthquakes and less or even none for deeper events related to the same seismic structure. Furthermore, the distinctive structure-specific precursory fingerprint pattern may manifest only above a certain magnitude threshold or, alternatively, its manifestation may vary according to the magnitude of the event.

All these dependencies and uncertainties must be considered and evaluated when assessing the genuine predictive value of any identified or, at the very least, interpreted precursory fingerprint. The establishment of a reliable prediction methodology is only possible after completing the entire three-stage investigation strategy involving experimental/learning, validation, and implementation phases [20] and applying what is learnt from one phase to the next. This necessitates a wide-scale interdisciplinary and international cooperation effort, along with the allocation of adequate financial and human resources. Nevertheless, such an endeavor appears as a more realistic approach to the earthquake prediction conundrum than the pursuit of universally valid precursors and the development of related methodologies and equipment.

In summary, following the successful identification of a significant number of precursory fingerprints for various seismogenic structures worldwide, it can be expected that a certain fingerprint typology will emerge enabling the systematization and classification. Such an evolution would allow optimization of the precursory signal systems, thereby enhancing their effectiveness and credibility.

5. Conclusions

Starting from the now worldwide shared postulate that precursor-based earthquake prediction is not impossible in principle, three principles may be considered as constituting the conceptual background of a virtually paradigm-shifting research strategy based on the recently introduced concept of a “precursory fingerprint” [20]: (1) the uniqueness of seismogenic structures, (2) the system of integrated and interacting geospheres, and (3) the non-equivalence of Earth surface spots in terms of precursory signal receptivity.

Following this original proposal, the precursory fingerprint concept was developed, refined, and discussed in detail by envisaging a sensor system that would continuously monitor a large number of time-varying physical, chemical, and biological parameters. These parameters would be measured using ground-based and satellite-held instruments targeting particular seismogenic structures, whose unique precursory fingerprint is investigated.

The precursory fingerprint of a given seismic structure is thought to be an assemblage of precursory signals (seismic, physical, chemical, biological) that are, in principle, detectable using monitoring equipment arranged as a matrix of n sensors placed on the ground at “sensitive” spots and on orbiting satellites. The system functions by means of a combination of signals emitted by “responsive sensors”, in addition to “non-responsive sensors”. The purpose is to monitor as many virtual precursory processes as possible by continuously measuring their relevant time-varying parameters. Each measured parameter possesses a threshold value, as determined by experts, and an uncertainty interval (“warning level”), which serves to differentiate between background and anomalous values, visualized similarly to traffic light signals (green, yellow, and red) to denote the respective levels of risk.

The precursory fingerprint, understood as a unique and distinctive configuration of “precursory signals”, may possess a multifaceted complex structure comprising various components. The “pattern component” is defined as the particular arrangement of the responsive sensors on the master board of the monitoring system, (green, yellow, and red, meaning background, warning, and anomalous parameter levels, respectively). This arrangement eventually can be re-arranged to form a spontaneously visible and easily comprehensible pattern. The “space component” is represented by the configuration of the signal-detecting sensors (responsive and non-responsive) in terms of their location on the map. The “time component” is defined as the time sequence of the anomalous signals, including their order, lead-time before the event, duration, transition time between yellow and red signals, etc.

The enormous volume of data resulting from years-long acquisition of time series requires the use of machine learning and pattern-recognition algorithms and artificial intelligence tools as well as sophisticated automated data managing systems, including acquisition, format, transmission, storage, processing, systematization, and evaluation. Uncertainties related to signal interpretation pertaining to sensor sensitivity, focal depth, and magnitude threshold have to be evaluated and interpreted by trained professional experts in all pertinent research domains.

While the above discussed research philosophy and strategy may appear utopic, it is nevertheless possible to follow and implement through large-scale and properly funded interdisciplinary international cooperation. Such cooperation could be initiated a number of well-known seismic structures worldwide in order to identify their earthquake precursory fingerprints. The experience learned from this initiative could then be applied to other structures at a later date, to enable a comprehensive research program with potentially paradigm-shifting effects to the research community at large.