InTech uses cookies to offer you the best online experience. By continuing to use our site, you agree to our Privacy Policy.

Earth and Planetary Sciences » Geology and Geophysics » "Advances in Geoscience and Remote Sensing", book edited by Gary Jedlovec, ISBN 978-953-307-005-6, Published: October 1, 2009 under CC BY-NC-SA 3.0 license. © The Author(s).

Chapter 34

Remote Sensing Rock Mechanics and Earthquake Thermal Infrared Anomalies

By Lixin Wu and Shanjun Liu
DOI: 10.5772/8292

Article top

Overview

RSRM experiment schemes to simulate different mechanisms of rock fracturing or tectonic EQ: a) uni-axially load on a standard cylinder rock sample; b) uni-axially load on a cylinder rock sample with a central hole; c) compressively-sheared load on a hexahedral rock sample; d) bi-axially load on three jointed rock samples to frictional sliding; e) bi-axially load on a damage rock sample with en echelon faults; f) bi-axially load on a damage rock sample with disjointed faults; and g) bi-axially load on three jointed rock samples simulating intersected faults.
Figure 1. RSRM experiment schemes to simulate different mechanisms of rock fracturing or tectonic EQ: a) uni-axially load on a standard cylinder rock sample; b) uni-axially load on a cylinder rock sample with a central hole; c) compressively-sheared load on a hexahedral rock sample; d) bi-axially load on three jointed rock samples to frictional sliding; e) bi-axially load on a damage rock sample with en echelon faults; f) bi-axially load on a damage rock sample with disjointed faults; and g) bi-axially load on three jointed rock samples simulating intersected faults.
The IRR image positive anomaly of “X”-shaped shearing fracturing of an uni-axially loaded marble sample.
Figure 2. The IRR image positive anomaly of “X”-shaped shearing fracturing of an uni-axially loaded marble sample.
The IRR image positive anomaly of “//”-shaped shearing-fracturing of an uni-axially loaded granite sample.
Figure 3. The IRR image positive anomaly of “//”-shaped shearing-fracturing of an uni-axially loaded granite sample.
The IRR image negative anomaly of “|”-shaped tensile fracturing of an uni-axially loaded granite sample.
Figure 4. The IRR image negative anomaly of “|”-shaped tensile fracturing of an uni-axially loaded granite sample.
The IRR image positive anomalies of a group of uni-axially loaded marble samples with a central hole.
Figure 5. The IRR image positive anomalies of a group of uni-axially loaded marble samples with a central hole.
The IRR image positive anomaly of the fracturing of a compressively sheared limestone sample (time in second).
Figure 6. The IRR image positive anomaly of the fracturing of a compressively sheared limestone sample (time in second).
The IRR image anomaly of the penniform-shaped fracturing of a compressively sheared marble sample.
Figure 7. The IRR image anomaly of the penniform-shaped fracturing of a compressively sheared marble sample.
The IRR image positive anomaly of the stick-slipping of symmetrical rock samples (time in second).
Figure 8. The IRR image positive anomaly of the stick-slipping of symmetrical rock samples (time in second).
The IRR image positive anomaly of the stick-slipping of asymmetrical rock samples (time in second).
Figure 9. The IRR image positive anomaly of the stick-slipping of asymmetrical rock samples (time in second).
The local IRR positive anomaly foretell the fracturing of en echelon disjointed faults (marble, time in second).
Figure 10. The local IRR positive anomaly foretell the fracturing of en echelon disjointed faults (marble, time in second).
The local IRR positive anomaly foretell the fracturing of collinearly disjointed faults (marble, time in second).
Figure 11. The local IRR positive anomaly foretell the fracturing of collinearly disjointed faults (marble, time in second).
The evolution of three indexes of IRR brightness temperature of loaded rock surface.
Figure 12. The evolution of three indexes of IRR brightness temperature of loaded rock surface.
Typical AIRT curve of uni-axially loaded rock.
Figure 13. Typical AIRT curve of uni-axially loaded rock.
The load-displacement curves of five tested rocks.
Figure 14. The load-displacement curves of five tested rocks.
The typical AIRT curve of uni-axially loaded rock sample (marble).
Figure 15. The typical AIRT curve of uni-axially loaded rock sample (marble).
The typical AIRT curve of compressively-sheared rock samples at different shearing angles.
Figure 16. The typical AIRT curve of compressively-sheared rock samples at different shearing angles.
The typical AIRT evolution of two kinds of disjointed faults bi-axially loaded.
Figure 17. The typical AIRT evolution of two kinds of disjointed faults bi-axially loaded.
AIRT of uni-axially loaded limestone sample.
Figure 18. AIRT of uni-axially loaded limestone sample.
AIRT short-dropping precursor for uni-axially loaded gabbro sample.
Figure 19. AIRT short-dropping precursor for uni-axially loaded gabbro sample.
AIRT fast rising precursor for uni-axially loaded marble sample.
Figure 20. AIRT fast rising precursor for uni-axially loaded marble sample.
AIRT dropping-to-rising precursor for compressively sheared loaded marble sample.
Figure 21. AIRT dropping-to-rising precursor for compressively sheared loaded marble sample.
The transient IRR thermograme of fracturing rock samples.
Figure 22. The transient IRR thermograme of fracturing rock samples.
The inside IRR isothermal filed on the fracturing zone of compressively sheared rock samples.
Figure 23. The inside IRR isothermal filed on the fracturing zone of compressively sheared rock samples.
The IRR mechanism related to the energy accumulation and consumption of a loaded rock.
Figure 24. The IRR mechanism related to the energy accumulation and consumption of a loaded rock.
The energy balance of a loaded rock sample in a closed independent system.
Figure 25. The energy balance of a loaded rock sample in a closed independent system.
Satellite TIR anomaly images before Dongsha Ms5.9 EQ (Sept 14, 1992).
Figure 26. Satellite TIR anomaly images before Dongsha Ms5.9 EQ (Sept 14, 1992).
The TIR images overlaid with detailed active faults before and after Zhangbei Ms6.2 EQ (Jan 10, 1998).
Figure 27. The TIR images overlaid with detailed active faults before and after Zhangbei Ms6.2 EQ (Jan 10, 1998).
The IRR anomaly features of simulated tectonic activities due to bi-axially load (Wu et al., 2007b).
Figure 28. The IRR anomaly features of simulated tectonic activities due to bi-axially load (Wu et al., 2007b).
TIR anomalies 0~11 days before Izmit Ms 7.8 EQ (Aug 17, 1999).
Figure 29. TIR anomalies 0~11 days before Izmit Ms 7.8 EQ (Aug 17, 1999).
The IRR anomaly and deformation concentration of a bi-axially loaded granite sample with disjointed faults.
Figure 30. The IRR anomaly and deformation concentration of a bi-axially loaded granite sample with disjointed faults.
Satellite TIR images one day before Hengchun Ms7.2 EQ (Dec 25, 2006).
Figure 31. Satellite TIR images one day before Hengchun Ms7.2 EQ (Dec 25, 2006).
Satellite TIR images before Wenchuan Ms8.0 EQ (May 12, 2008).
Figure 32. Satellite TIR images before Wenchuan Ms8.0 EQ (May 12, 2008).

Remote Sensing Rock Mechanics and Earthquake Thermal Infrared Anomalies

Lixin Wu1, 2 and Shanjun Liu2

1. Introduction

Rock fracturing is the cause of many geo-hazards including tectonic earthquake (EQ), rock burst, rock sloping and rock pillar failure. Radiation signals such as acoustic emission, radio frequency emission and electromagnetic (EM) radiation from loaded deforming rock, are able to provide useful information for monitoring, interpreting and predicting rock fracturing (Renata, 1977, Brady and Rowell, 1986, Yamada et al., 1989, Martelli et al., 1989). Based on thermo-elastic theory, thermo-elastic stress analysis (TSA) and stress pattern analysis by thermal emission (SPATE) were developed for the stress measurement of solid materials, including homogeneous metal, macromolecular and composite materials, respectively in 1960’s and 1970’s (Mounatin and Webber, 1978). Luong applied thermovision to study experimentally the damage processes of concrete and rock (Luong, 1990), but no reach to the remote sensing on geo-hazards.

In the experiments for investigating the mechanism of satellite thermal infrared (TIR) anomaly before tectonic EQ (Gorny et al., 1988, Qiang et al., 1990), it was discovered that there do exist TIR anomaly before rock fracturing (Geng et al., 1992). Later, it was furthermore discovered that there are obvious TIR features as precursors of rock fracturing, and that the loaded stress around 0.79 σc can be taken as a precaution index for the stability monitoring of loaded rocks (Wu and Wang, 1998). To explore the laws of infrared radiation (IRR) variation in the process of rock loading, deforming and fracturing, and to reveal the possible mechanism of satellite TIR anomaly before EQ, a large amount of IRR imaging experiments on rock loaded to fracturing were conducted in China (Wu et al., 2000, 2001, 2002, 2003, 2004a, 2004b, 2004c, 2004d, 2006a, 2006b; Deng et al., 2001, Liu et al., 2002). Hence, a new intersection discipline, Remote Sensing Rock Mechanics (RSRM), which takes Remote Sensing, Rock Mechanics, Rock Physics and Informatics as its foundations and serves for remote sensing on geo-hazards, was originated (Geng et al., 1992; Wu et al., 2000).

Based on retrospection to past experiments on RSRM, it was pointed out that there are two IRR anomalies, being IRR image anomaly and IRR temperature curve anomaly respectively, can act as rock fracturing precursors. The average IRR temperature (AIRT), being the integral reflection of surface IRR energy, is applied as a quantitative index to study the temporal evolution of IRR from loaded rock and to seek for the potential precursors of rock fracturing. The temporal evolution of AIRT are the comprehensive effect of a series of physical-mechanical processes inside a loaded rock, such as rock thermo-elastic acting, pore gas desorbing & escaping, fractures producing & extending, rock frictionating, heat transferring and environment radiation. The thermo-elastic effect and the frictional thermal are two of the main mechanisms of increased IRR from loaded rock. RSRM experiments had revealed the laws of changed IRR from loaded rock and provided scientific interpretations for the mechanisms of satellite TIR anomaly before tectonic EQs of Ms>5.5.

2. Remote Sensing Rock Mechanics Experiments

2.1. Experiment Methods and Tools

The typical RSRM experiment is comprised of a loader (uni-axial or bi-axial), an infrared imager and rock samples. As in Figure 1, a bi-axial loader was applied for loading along two directions, and an infrared imager was applied to detect the surface IRR from loaded rock. The maximum imaging rate of the imager is 60f/s, and the recording rate was usually set as 1f/s to record the IRR images continuously. Usually, tectonic EQ might be resulted from the suddenly fracturing of compressively-sheared crust rock, the suddenly breaking of faults at disjointed zones, the suddenly sliding of compressively-sheared faults or the stability losing of compressively loaded intersected faults. To simulate the different mechanisms of rock fracturing and EQ, several typical loading schemes were applied as in Figure 1.

media/image2.jpg

Figure 1.

RSRM experiment schemes to simulate different mechanisms of rock fracturing or tectonic EQ: a) uni-axially load on a standard cylinder rock sample; b) uni-axially load on a cylinder rock sample with a central hole; c) compressively-sheared load on a hexahedral rock sample; d) bi-axially load on three jointed rock samples to frictional sliding; e) bi-axially load on a damage rock sample with en echelon faults; f) bi-axially load on a damage rock sample with disjointed faults; and g) bi-axially load on three jointed rock samples simulating intersected faults.

2.2. Rock Fracturing Precursor: IRR Image Anomaly

2.2.1. Uni-axially loaded rock

Lots of rock samples made from coal, ironstone, sandstone, marble, limestone, granite, granodiorite, gabbro and gneiss were uni-axially loaded and thermal imaging detected. The sample size was standard of diameter and length, respectively, 50 and 100mm. It was discovered that the IRR images of the uni-axially loaded rock have different features for different fracturing pattern (Wu et al., 2006a). As in Figures 2~4, there are three fracturing patterns, “X”-shaped, “//”-shaped and “|”-shaped respectively, occurred in our experiments. The “X”-shaped and “//”-shaped positive IRR abnormal strips foretell the coming of “X”-shaped shearing fracturing and the coming of “//”-shaped shearing fracturing respectively, while the “|”-shaped negative IRR abnormal strip foretells the coming of tensile fracturing.

The “X”-shaped positive IRR abnormal strips generated with loading along the “X”-shaped shearing zone before peak stress, and got distinguished after peak stress, as in Figure 2. The rock sample got finally fractured along the “X”-shape shearing zone. The evolution of IRR abnormal strip had also reflected the fracturing being not symmetrical upper-and lower, in that the upper part was clear with higher temperature, while the lower part is fuzzy with lower temperature.

media/image3.jpg

Figure 2.

The IRR image positive anomaly of “X”-shaped shearing fracturing of an uni-axially loaded marble sample.

The “//”-shaped positive IRR abnormal strips generated with loading along the “//’-shaped shearing zone at the upper part of sample before peak stress, and got distinguished after peak stress, as in Fig 3. The evolution of the positive IRR image anomaly had also reflected the fracturing being not symmetrical in that the upper part of the IRR anomaly strip was clear with higher temperature, while the lower part was fuzzy excepting for the final fracturing near the bottom of the sample. Besides, there was strong IRR anomaly spot at the fracturing center for the intensive accumulation of mechanical energy and for the intensive generation of frictional thermal at the local central place.

media/image4.jpg

Figure 3.

The IRR image positive anomaly of “//”-shaped shearing-fracturing of an uni-axially loaded granite sample.

The “|”-shaped negative IRR abnormal strip generated with loading along the tensile fracturing zone of a rock sample before the peak stress, and got distinguished gradually at the peak stress and after fracturing, as the approximately vertical dark strip in Figure 4. The same phenomenon for a sandstone sample with a calcite vein was also reported (Wu, et al., 2000).

media/image5.jpg

Figure 4.

The IRR image negative anomaly of “|”-shaped tensile fracturing of an uni-axially loaded granite sample.

2.2.2. Uni-axially loaded rock with a central hole

More than 10 samples with a central hole, modeling the structure stability of loaded rock tunnels, made from marble and granite were infrared imaging detected. The rock samples had two kinds of shapes respectively being cylinder with diameter and length, respectively, 50 and 100mm, and regular block with thickness, width and length, respectively, 70, 35 and 100 mm. It was discovered that there were distinguished positive IRR image anomalies before rock fracturing, and the place of anomaly were exactly the coming fracturing place. As in Figure 5, the positive IRR image anomalies had reflected the two kinds of fracturing, respectively being diagonal fracturing (sample 1~5) and fork fracturing (sample 6). The IRR anomalies, along the fracturing planes and shaped as spots or strips, generated not only on rock surface but also on the hole’s surface (lateral sample 4 and 5). The temperature increment is 1~3oC and 4~8oC respectively for marble and granite samples.

media/image6.jpg

Figure 5.

The IRR image positive anomalies of a group of uni-axially loaded marble samples with a central hole.

2.2.3. Compressively sheared rock

More than 20 samples, size 7×7×7×cm3, made from sandstone, marble, limestone, granite and gneiss, were compressively sheared and infrared imaging detected. Three pairs of steel platens with shearing angle being 45º, 60ºand 70º respectively were applied. The loading rate was controlled as 2~5kN/s. It was discovered that the IRR temperature of rock surface changed with loading, and a strip-shaped positive IRR image anomaly generated along the central shearing plane before fracturing. With loading, the positive abnormal strip got more and more distinguished and migrated gradually from the upper end to the lower end of the sample, which foretold that the compressive-shearing fracturing was developing gradually from the upper end to the lower end of the sample along the central shear plane. Figure 6 shows the typical IRR image series of a compressively sheared limestone sample.

As a special geological phenomenon occurring with the formation of great fault, penniform-shaped fractures are a group of secondary fractures produced with the formation of great primary fracture (Nicolas et al., 1977). It happened to occur in our experiments that there were penniform-shaped fractures produced with a primary fracture in the compressive loaded rock samples, as in Figure 7. The IRR positive anomaly strips generated aside the primary IRR strip, passing through the central shearing plane, had reflected the penniform-shaped fracturing events.

media/image7.jpg

Figure 6.

The IRR image positive anomaly of the fracturing of a compressively sheared limestone sample (time in second).

media/image8.jpg

Figure 7.

The IRR image anomaly of the penniform-shaped fracturing of a compressively sheared marble sample.

2.2.4. Bi-sheared frictional sliding rock blocks

Ten groups of rock samples made from gabbro, granodiorite, limestone and marble were IRR detected in the process of bi-sheared frictional sliding or viscosity sliding. Each group, as in Figure 8 and 9, was comprised of three jointed rock blocks whose size respectively be 50×50×100mm3, 50×70×150mm3 and 50×50×100mm3 from left to right, and its friction area was constant, 50×100mm2. Four contact conditions, symmetrical (yes for rock property and for its smooth friction surface, as in Figure 8), uncertain symmetrical (yes for rock property but not for its coarse friction surface, as in Figure 9), unstable asymmetrical (yes for rock property but not for its staged friction surface) and stable asymmetrical (not for rock property but yes for its smooth friction surface), were designed and tested respectively (Wu et al., 2004b).

It is revealed that the evolution of rock surface IRR temperature field is not only correlated with rock stress, but also correlated with the features of friction surface and rock properties at both sides. General law lies in that the IRR at the place of stress concentration and strong friction zone is stronger than that at the place of stress relaxation and weak friction zone. In condition of friction surface be symmetrical, the IRR image is double butterfly-wings shaped, as in Figure 8. However, in condition of friction surface be uncertain symmetrical, unstable asymmetrical or stable asymmetrical, the temporal-spatial evolution of IRR anomaly is uncertain or unstable, as in Figure 9. The positive IRR anomaly spots, foretelling the evolution of stress, energy and viscosity-sliding process, may be beads-shaped, needle-shaped, suspended needle-shaped, strip-shaped, single butterfly-wings shaped or its evolution in order (Wu et al., 2004b).

media/image9.jpg

Figure 8.

The IRR image positive anomaly of the stick-slipping of symmetrical rock samples (time in second).

media/image10.jpg

Figure 9.

The IRR image positive anomaly of the stick-slipping of asymmetrical rock samples (time in second).

2.2.5. Bi-axially loaded rock

Infrared imaging detection on the rupturing of en echelon and collinearly disjointed jointed faults were done in the process of bi-axial loading. It was revealed that the IRR from loaded rock surface is correlated with loading stress, which could be divided into five stages as loading beginning, linear elastic, stress locking, stress unlocking and fracturing(Wu et al., 2004a). During the stress-unlocking stage, positive IRR anomaly strip generated at the disjointed zone, as in Figure 10 and 11. The positive IRR anomaly strip around the disjointed zone has general evolution features as: firstly, the strip gets enhancing; then, gets weakening (or ‘silence’); and finally, gets enhancing again. The re-enhancing of IRR anomaly strip after the weakening stage is a meaningful precursor foretelling the place of primary fracturing of faults or the coming epicenter of an EQ.

media/image11.jpg

Figure 10.

The local IRR positive anomaly foretell the fracturing of en echelon disjointed faults (marble, time in second).

media/image12.jpg

Figure 11.

The local IRR positive anomaly foretell the fracturing of collinearly disjointed faults (marble, time in second).

2.3. Rock Fracturing Precursor: IRR Temperature Anomaly

2.3.1. Quantitative index: AIRT

The evolution of surface IRR from loaded rock is the comprehensive effect of rock thermo-elastic acting, pore gas desorbing & escaping, fractures producing & extending, rock frictionating, heat transferring and environment radiation. Being the integral reflection of surface IRR energy, the average IRR temperature (AIRT) is selected as a quantitative index to study the evolution of IRR from loaded rock and to seek for rock fracturing precursors (Wu et al., 2006b).

The infrared imager detects and records the thermal images of loaded rock surface. The thermogram is comprised of a matrix of color pixels which representing the IRR brightness temperature of each pixel of the rock surface. For example, the imaging matrix of TVS-8100MKII infrared imager is 160×120. The IRR temperature of each pixel tends to fluctuate with time due to the instability of the detector unit and the influence of environmental radiation, and the IRR temperature of each pixel will not always be the same for the local difference of rock stress and rock strain. The maximum, minimum and average value of loaded rock surface IRR temperature, respectively being IRRT max, IRRT min and IRRT ave, could be quantitatively obtained from thermogram. The analysis revealed that IRRT max and IRRT min will not change obviously except that IRRT max might rise suddenly just before rock fracturing, while IRRT ave is to change stably with loading, as in Figure 12. The physical interpretation lies in that the surface IRRT ave is a general reflection of the energy balance inside the loaded rock.

media/image13.jpg

Figure 12.

The evolution of three indexes of IRR brightness temperature of loaded rock surface.

Hence, the IRRT ave of rock surface, denominated as AIRT, is selected as a quantitative index to study the precursors of rock fracturing and geo-hazards. The procedures for AIRT-based precursor analysis includes: 1) to define a unified boundary of the analyzed region (resampling region) for all thermogrames; 2) to resample the IRR temperature value from the data file of each thermograme in time order; 3) to calculate the AIRT of the resampling region of each thermograme; 4) to draw the AIRT–time curve of the rock sample; 5) to analyze the evaluation features of the AIRT–time curve and to identify the messages as a precursor of rock fracturing and hazard; 6) to compare with the qualitative image anomaly so as to analyze and to confirm the AIRT abnormal precursor.

2.3.2. Influence factors of AIRT curves

  • 1) Loading stages and rock deformation

The stress-strain curve is a basic method for describing rock deformation and for interpreting rock mechanical behaviors (Hudson and Harrison, 1997). Generally, the deformation process of loaded rock is divided into four stages respectively being stage-I of defects compaction, stage-II of linear elastic deformation, stage-III of plastic deformation and stage-IV of fracturing failure, as in Figure 13. The four characteristic points, E, Y, P and F are called as elastic-starting point, yield-starting point, peak-stress point and failure-impending point respectively.

Stage-I: the downward-concave curve section tells that there are some defects such as pores, fissures and joints inside the rock body, and that the defects is under compaction, which cause the stress to rise slowly. The more the defects, the severe the curve downward concave.

Stage-II: the curve section linearly developed tells that the compaction of defects has finished and the rock is undergoing elastic deformation. The higher the angle of the section line inclined, the stronger the rock.

Stage-III: the upward-concave curve section tells that there are new fractures developing inside the rock. The plastic deformation starts, and the new generated fractures together with the initial defects are possible to cause friction between its two side-faces.

Stage-IV: the curve section turning to drop tells that the fractures are getting wider, longer and to connect with each other. The rock is losing its strength and stability, and the final fracturing failure or rock hazard is impending.

For the difference between rock compositions, the details of stress-strain curve of different rock will be different. As to brittle rock, its stage-II is close to point P and its stage-IV will be cliff-shaped. Usually, most of the crust rocks are brittle. Five kinds of typical crust rock, granodiorite, gabbro, gneiss, limestone and marble had been tested in our experiments. The typical load-displacement curves of the tested rock samples are shown in Figure 14. It tells that all the tested rocks are brittle.

media/image14.jpg

Figure 13.

Typical AIRT curve of uni-axially loaded rock.

media/image15.jpg

Figure 14.

The load-displacement curves of five tested rocks.

  • 2) The influence of loading condition

The experiments discovered that the evolutions of AIRT have different laws in different loading condition, such as uni-axial loading (constant displacement controlled), compressively-sheared loading (approximately constant load controlled) and bi-axial loading (constant displacement controlled).

  • a) Uni-axial loading

As shown in the left-hand side of Figure 15, the surface facing to the infrared imager is to be detected, and a rectangle region close to the boundary of the rock sample is defined for data resampling and analyzing (Liu et al., 2002; Wu et al., 2002). Multiple experiments revealed that there was slight variation of AIRT at different deformation stage although the AIRT linearly increased with load and deformation. At stage-I, the AIRT will rise slowly or drop a little; at stage-II, the AIRT will rise stably; at stage-III, the AIRT will rise quickly than that in stage-I and stage-II. The right-hand side of Figure 15 shows the comparison of the evolution of AIRT and the load with rock deformation, which is rock displacement in generally, of a marble sample.

  • b) Compressive-shear loading

In condition of compressively shear, the rock sample will always get fracturing along the shearing plane, which locates near to the central plane of the loaded sample. To minimize data resampling work and to focus on the key region, a narrow rectangle along the shearing plane is defined as the resampling and analyzing region (Wu et al., 2004c). Multiple experiments revealed that the temporal evolution of the AIRT is different with the shearing angle. Three shearing angles (γ) being 45o, 60o and 70o respectively, are applied. As the shearing angle changes from 45o to 70o, the temporal evolution changes from monotonic rise, to drop-to-rise and to monotonic drop in order, as in Figure 16.

media/image16.jpg

Figure 15.

The typical AIRT curve of uni-axially loaded rock sample (marble).

media/image17.jpg

Figure 16.

The typical AIRT curve of compressively-sheared rock samples at different shearing angles.

The mechanism lays in that the ratio of compressive-stress to shear-stress along the shearing plane decrease with the rise of the shearing angle. The smaller is the shearing angle, the higher is the compression-shear ratio. In condition of 45o, the load-time curve developed in three stages, as stage A, B and C in Figure 16a, with load increasing from slow to rapid, and to slow again. In condition of 60o the load-time curve developed in two stages, as stage A and B in Figure 16b, with load speed changing from approximate constant to be decrease slightly. In condition of 70o, the load-time curve developed in two stages, as stage A and B in Figure 16c, with loading speed changing from slow to rapid.

The compressive action on loaded rock is to cause surface IRR temperature rise, while the tensile action on loaded rock is to cause surface IRR temperature drop. Actually, both compressive action and tensile action are to occur along the compressively sheared plane, and the detected surface IRR is the comprehensive effect of the two actions. It was reached that (Wu et al., 2004c): 1) in condition of shearing angle being 45o, the surface AIRT will rise monotonically with loading in that the temperature increment from compressive action and friction is stronger than the temperature decrement from tensile action in the whole loading process; 2) in condition of shearing angle being 60o, the surface AIRT will drop monotonically with loading in that the temperature increment from compressive action and friction is weaker than the temperature decrement from tensile action before stage-III (point Y in Figure 13, and point P in Figure 16b); with the friction effect getting strong in stage-III, the surface AIRT will get to rise in that the temperature from compression and friction get stronger than the temperature decrement from tensile action; 3) in condition of 70o, the surface AIRT will drop monotonically with loading in that the temperature increment from both compression and friction are weaker than the temperature decrement from tensile action.

  1. c) Biaxial loading

By using of bi-axial loading system and infrared imaging system, the IRR features of two kinds of disjointed jointed faults, respectively be collinearly and non-collinearly disjointed faults, were experimentally studied (Wu et al., 2004a). Since all the faults got fractured finally at the disjointed zone, a circle covering the disjointed region is defined as the resampling region, as in the left-hand side of Figure 17. It could be known from the right-hand side of Figure 17 that the IRR from loaded samples is related with load stress, and the evolution stage could be classified into five stages (I~V) relating with initial compacting, elastic deforming, stress blocking, stress deblocking and rock fracturing respectively. From stage-II to stage-IV, the evolution of AIRT has the features of rising to dropping, and to rising again.

media/image18.jpg

Figure 17.

The typical AIRT evolution of two kinds of disjointed faults bi-axially loaded.

  • 3) The influence of rock characteristics

It was discovered in our experiments that for the most of rock samples in condition of uni-axial loading, its AIRT approximately rose with loading. But there were a few abnormal samples made from limestone had shown AIRT features of dropping with loading, as in Figure 18. The cause lies in that limestone has much more pores than the other rocks. Usually, there are many gases, such as CH4, CO2, CO and O2 etc., enclosed inside the pores of rock body (Wang, 2003). With the decrement of pore volume due to the loading compaction and with the increment of fractures produced inside the limestone sample, the pore gases will get escaping. The escaping behavior of pore gas needs to absorb thermal energy from the rock sample. If the heat from compression and friction is lower than that absorbed by pore gases, the surface AIRT is to drop with loading. If look carefully at the load-displacement curves in Figure 14, it could be founded that the curve of limestone concaved downward the most at the compaction stage as compared to that of the other four kinds of rock, which means that there are more pores inside limestone than the others.

media/image19.jpg

Figure 18.

AIRT of uni-axially loaded limestone sample.

media/image20.jpg

Figure 19.

AIRT short-dropping precursor for uni-axially loaded gabbro sample.

2.3.3. The classification of precursors

Analysis to the evolution of AIRT curves discovered that a large amount of rock samples in condition of uni-axial loading, compressively-sheared loading and bi-axial loading had presented obvious precursors for rock fracturing. As referring to the general process of AIRT evolution, the AIRT anomaly precursors for rock fracturing and hazard could be classified as short-dropping, rapid-rising and dropping-to-rising respectively.

  • Short-dropping precursor

The AIRT curve rises with loading but has a short dropping at loading stage-IV; later, the AIRT curve will rise again. The bi-axially load on collinearly and non-collinearly disjointed faults had shown short-dropping precursors as in Figure 17, and the point P was suggested to be the precursor point of rock fracturing and rock hazard. Figure 19 shows another typical case of gabbro sample uni-axially loaded. Here, point A is the turning point of AIRT from rising to short dropping, and point P is another turning point from short dropping to rising again, which is suggested to be the precursor point of rock fracturing and rock hazard.

  • Rapid-rising precursor

The AIRT curve rises slowly with loading but turns to rise rapidly before rock fracturing, and the turning point is exactly the precursor point. Figure 15 and Figure 16a have this kind of precursor. Figure 20 shows another typical case of marble sample uni-axially loaded. Here, point P is the turning point of AIRT from rising slowly to rising fast, which is suggested to be the precursor point of rock fracturing and rock hazard.

  • 3) Dropping-to-rising precursor

The AIRT curve drops slowly with loading but turns to rise just before rock fracturing, and the turning point is exactly the precursor point. Figure 16b has this kind of precursor. Figure 21 shows another typical case of marble sample uni-axially loaded. Here, point P is the turning point of AIRT from dropping slowly to rising fast, which is suggested to be the precursor point of rock fracturing and rock hazard.

media/image21.jpg

Figure 20.

AIRT fast rising precursor for uni-axially loaded marble sample.

media/image22.jpg

Figure 21.

AIRT dropping-to-rising precursor for compressively sheared loaded marble sample.

2.3.4. The temporal features of precursors

The occurrence moments of the precursors of AIRT of totally 52 tested rock samples are listed in Table 1(Wu et al, 2006b). Although the loading conditions and the rock samples are different, the precursor occurrence moment, were very similar as 0.77~0.94σ c (σ p ). Here, σ c is the uni-axial compressive strength, and σ p is the peak stress. The precursor occurrence moment of uni-axially loaded or compressive sheared rock sample is 0.79σ c and 0.82σ c respectively. It is worthy to mention that the precursor occurrence moment of bi-axially loaded collinearly disjointed faults and echelon faults are much different. That for collinearly disjointed faults was close to the peak stress, 0.87σ p , while that for echelon faults was far away from peak stress, 0.77σ p . It provides an important evidence for the complexity of study on tectonic EQ prediction on shock time, based on satellite infrared remote sensing and referring to the seismogenic mechanism.

Loading conditionSum of tested samples, StSum of samples with precursors, SpThe ratio: (Sp/St)×100%Average of precursor occurrence moment
uni-axial loading22941%0.79σc
compressively-sheared loading70°7114%58%0.94σp
60°8674%0.82σp
45°11873%0.77σp
Bi-axial loading for collinearly disjointed faults22100%0.87σp
Bi-axial loading for en echelon faults22100%0.77σp

Table 1.

Statistic for precursor occurrence in different loading conditions.

2.4. Large IRR at Fracturing Centre

As the recording rate of the infrared imager applied was 60f/s, the transient IRR temperature at the fracturing center could be snapped. In condition of uni-axial loading, the fracturing center of a brittle rock is usually at the “X”-shaped fracturing center. It is discovered that the transient IRR temperature at the fracturing center is much higher than that on rock surface, as in Figure 22, and it is positively related with rock strength and rock deformation. For some compressively sheared hard rock samples made from gabbro and gneiss, the transient IRR temperature at the fracturing center is higher than 155oC, which is the upper limitation of the 2nd temperature range (72~155oC) of the imager applied.

In condition of high angle, 60ºand 70º, compressively sheared loading, the fracturing center is at the center of the fractured shearing zone. Since the ruptured upper block of rock sample was pushed apart from the steel platen immediately after the abrupt rupturing, usually be 1~2 s after the rupturing, the inside shearing zone got exposed to the imager immediately and the transient IRR temperature filed was snapped. It was discovered that the IRR temperature on the inside shearing zone is not only much higher than that of outside rock surface, but also inhomogeneous distributed, neither even nor centripetal, as in Figure 23. It means that much more mechanical energy had been converted into frictional thermal and IRR energy due to the intensive energy accumulation, the sufficient local deformation and the abrupt frictional sliding at the center of the shearing zone. In other words, the large IRR temperature at the inside shearing center had reflected the comprehensive effect of local concentrated energy conversion and frictional thermal.

media/image23.jpg

Figure 22.

The transient IRR thermograme of fracturing rock samples.

media/image24.jpg

Figure 23.

The inside IRR isothermal filed on the fracturing zone of compressively sheared rock samples.

Hence, it could be deduced that in condition of great tectonic stress, large deformation and/or abrupt frictional sliding, large temperature as high as hundreds or thousands of degree Celsius is possible inside crust rocks. The high temperature could cause partial melt of crust rock, which provides a scientific explanation for the existence of pseudotachylyte in some lager faults (Nicolas et al., 1977, Sibson et al, 1980) and for the failure-generated EQ lights (Martelli et al., 1989). Besides, we can deduce that the continuous shearing deformation or the abrupt fracturing of highly loaded rock/coal body in a coal mine is possible to cause local sheared heating of temperature hundreds of degree Celsius, which might be a potential ignition of local methane (the minimum ignition temperature is 595oC).

3. Remote Sensing Rock Mechanics Model (RSRM-model)

3.1. Thermo-Mechanical Coupling Effect

3.1.1. Thermo-mechanical coupling in a loaded solid

The heat production inside a loaded solid is called as thermo-mechanical coupling effect. According to the material features and the different deformation stages of a loaded solid, the thermo-mechanical coupling is classified as thermo-elastic, thermo-plastic and thermo-viscous respectively for elastic deformation, plastic deformation and viscous deformation. Generally the rock is a hard brittle solid, its plastic and viscous deformation could be ignored, and the thermo-elastic effect and the frictional thermal are the two chief mechanisms of surface IRR from loaded rock. Kelvin coined the thermo-elastic theory in 1853 that the changed physical temperature of a loaded component is correlated to its changed stress as follows:

ΔT/T=K0Δσ
(1)

Here: T is the absolute temperature of a loaded component (K); ΔT is the changed temperature (K); K0 is the thermo-elastic factor (MPa-1); and Δσ is the changed sum of three principal stresses ( σi,i=1,2,3 , MPa).

As for an isotropic linear elastic solid loaded bi-axially with a free surface, the surface physical temperature variation is tightly correlated with the sum of two principal stresses ( σi,i=1,2 ):

ΔT=α/ρCp[TΔ(σ1+σ2)]
(2)

Here: T is the surface absolute temperature of a loaded solid (K); ΔT is the changed temperature (K); α is the factor of linear expansion (K-1); ρ is the solid density (Kg m-3); Cp is thermal capacity of solid at normal atmosphere (J Kg-1 K-1); σ1 and σ2 are the two principal stresses (MPa). The thermo-elastic factor K is defined as K=α/ρCp .

For the mechanism of stress measurement with TSA and SPATE, the relationship between the stress increment and the IRR signal based on equation (2) is as follows (Mounatin and Webber, 1978):

Δ(σ1+σ2)=AthΔSΔS=Δ(σ1+σ2)Ath1
(3)

Here: A th is a comprehensive factor called as corrective factor, which is a function of solid surface emissivity, solid surface physical temperature, solid thermo-elastic factor and three parameters related to the IRR detector, unit in MPa U-1. ΔS is the increment of thermo-elastic voltage signal detected (U).

3.1.2. Changed IRR temperature of loaded rock

If the slight change of rock surface emissivity, rock thermo-elastic factor and the physical parameters of IRR detector during rock loading could be ignored, and if the changed rock surface physical temperature cannot be ignored due to the thermal exchange and the frictional thermal, the relationship between A th and changed rock surface physical temperature could be expressed as Ath=βT1 . The detected IRR signal S is a direct representation of surface IRR temperature, i.e., ΔIRRT=γΔS . Hence, the following equation for changed IRR temperature could be deduced:

ΔIRRT=γβ1TΔ(σ1+σ2)
(4)

Here: ΔIRRT is the changed IRR temperature (K); β is a constant correction factor related to rock surface’s emisivity, rock thermo-elastic factor and three parameters of IRR detector, in unit MPa K U-1; γ is a transfer factor between detected voltage signal and IRR temperature (K U-1).

It means in equation (4) that the changed IRR temperature of rock surface is a direct reflection of the changed sum of the two principal stresses. If no frictional thermal produced and the thermal exchange is stable, the IRR temperature of rock surface is to rise with loading, and the spatial-temporal evolution of surface IRR image will be stable. If there is no frictional thermal produced but the thermal exchange is unstable, the surface IRR temperature will be unstable, and the spatial-temporal evolution of surface IRR image will also be unstable. If there is frictional thermal produced inside and conducted to rock surface, both the thermal exchange and the surface IRR temperature of rock surface will be unstable, and the spatial-temporal evolution of surface IRR image will get complicated, as in Figures 2~5, 10, 11.

Especially, in condition of compressive shearing and fictional sliding, there is a large amount of frictional thermal produced in the friction zone, which is to cause the rise of physical temperature in friction zone. The rock surface temperature will rise if the thermal conduction from the friction zone can reach to rock surface. Hence, the IRR image anomaly will be distinguished and sometimes be large as a combined effect of rock stress and frictional thermal, as in Figures 6~9.

3.2. Remote Sensing Rock Mechanics model

3.2.1. Energy-Related IRR from Loaded Rock

As a relatively independent closed system comprised of loader head, rock sample and environmental air, as in Figure 24, the rock deformation, rock fracturing and rock hazard are all of a complex process of energy input and consumption. If the possible chemical reactions inside a loaded rock can be ignored, the inputted energy of a loaded rock will include the mechanical work from loader and the heat input through positive thermal exchange from loader head and environmental air. The energy consumption by the loaded rock is much more complex including the energy accumulation in rock and the energy dissipation from rock.

media/image46.jpg

Figure 24.

The IRR mechanism related to the energy accumulation and consumption of a loaded rock.

The energy accumulation in a loaded rock includes the positive elastic-plastic deformation energy of rock (the positive change of oscillation and rotation energy of mineral molecules), the surface energy of new produced fractures or fissures, and the friction actions between mineral molecules, grains, joints, fissures and fractures inside the rock as well as thus produced frictional thermal. The energy dissipation from loaded rock includes the negative thermal exchange with the loader head and/or environmental air (i.e., heat output), the kinetic energy of departed fragments of fractured rock, the light radiation, acoustic emission, radio frequency emission and IR & microwave radiation.

The thermal exchange and the friction action are to change the heat state of a loaded rock, and the rock surface physical temperature is a direct index reflecting the heat state of the loaded rock. Stephen-Boltzmann law states that the IRR strength (radiation flux density) of any material, at temperature above absolute zero degree, is biquadratic to its surface physical temperature. Crystal Physics states that the energy jump of molecules oscillation and/or rotation due to the change of molecules distance, resulting from deformation, is an important mechanism of electromagnetic radiation. Hence, rock surface IRR is a comprehensive effect of rock deformation and rock surface thermal state. Rock surface IRR temperature could be a detective index reflecting rock surface physical temperature and rock surface deformation field, which implicating the complex physical-mechanical process inside the loaded rock.

In spite of thermal exchange and plastic deformation, the thermo-elastic effect and the frictional thermal are two of the main mechanics of changed IRR from loaded brittle rock. In the stage of elastic deformation, the thermo-elastic effect is the main cause; while in the stage of plastic deformation or fracturing, the friction-thermal effect plays a great role. At the moment of rock fracturing or hazard, the fraction-heat effect gets more distinguished. The frictional thermal effect depends on two factors being frictional force (decided by normal stress and frictional coefficient) and frictional speed respectively. The larger the frictional force and the quicker the frictional speed, the more the frictional heat.

3.2.2. RSRM-model based on independent system

The rock sample, load header and environment air could be taken as a closed independent system, as in Figure 25, and the energy of the loaded rock sample is in a balance state as follows:

media/image47.jpg

Figure 25.

The energy balance of a loaded rock sample in a closed independent system.

ΔM+ΔT=ΔD+ΔF+ΔH+ΔG
(5)

Here: ΔM ——The inputted mechanical energy from loader, J; be positive;

ΔT ——The inputted thermal energy from loader and environment air, J; be positive; ΔD ——The produced deformation energy of rock sample in elastic and plastic state, J; be positive; ΔF ——The sum of consumed rock fracturing energy and formed fracture surface energy, J; be positive; ΔH ——The heat energy increment of rock decided by its physical temperature, J; be positive if temperature rise or be negative if temperature drop; ΔG ——The energy consumed by the desorbing and escaping of pore gas in rock samples, J; be positive.

From equation (5) we have:

ΔH=(ΔM+ΔT)(ΔD+ΔF+ΔG)
(6)

The change of heat energy will result in the change of physical temperature,surface radiation energy and IRR temperature of loaded rock samples. Referring to Stephen-Boltzmann law, the AIRT is a direct index of rock radiation energy, and the change of AIRT ( ΔAIRT ) must have certain a relationship with the physical temperature (T) of loaded rock sample as:

EIR=f(AIRT)=εσT4
(7)

Here: E IR ——The radiation energy of a loaded rock, J;

AIRT ——The surface average IRR temperature of a loaded rock, K; ε ——The radiation factor of rock sample, 0ε1 ; σ ——The constant of Stephen-Boltzmann, σ=5.6679×108 , J×m-2×K-4

Thermo-elastic effect is the basic mechanism of changed IRR from a loaded rock sample. Moreover, the desorbing and the escaping of pore gas, the expanding of initial fissures or joints, the friction between fissures, fractures, joints and grains, the thermal transfer between the rock, loader header and environment air, and the radiation from the environment all are to have thermal effect on loaded rock samples. The thermal state is comprehensively affected by the six factors which are rock stress, pore gas desorbing & escaping, rock fracturing, heat transferring, rock frictionating and environment radiation respectively. The equation could be expressed as follows:

ΔAIRT=f[Δ(σ1+σ2),ΔG,ΔF1,ΔH,ΔF2,ΔE]=f1(t)+f2(t)+f3(t)+f4(t)+f5(t)+f6(t)
(8)

Here: ΔAIRT ——The detected change of AIRT , K; be positive if rise, be negative if drop;

f1(t) ——IRR temperature change due to thermo-elastic effect ( Δ(σ1+σ2) , K; be positive or negative; f2(t) ——IRR temperature change due to pore gas adsorbing & escaping ( ΔG ), K; be negative; f3(t) ——IRR temperature change due to the production of new fractures and the expansion of initial fissures, joints and new produced fractures ( ΔF1 ), K; be negative; f4(t) ——IRR temperature change due to frictional thermal ( ΔF2 ), K; be positive; f5(t) ——IRR temperature change due to heat transfer ( ΔH ), K; be positive or negative. f6(t) ——IRR temperature change due to environment radiation ( ΔE ), K; be positive.
  • 1) Thermo-elastic effect: f1(t)

Referring to thermo-elastic theory and equation (4), f1(t) could be calculated as:

f1(t)=γβ1TΔ(σ1+σ2)
(9)

In condition of uni-axially compressive loading, the load is to cause temperature rise in that σ2 is constant zero and the positive σ1 will linearly increase with loading. If Δσ1 is positive, the f1(t) will be positive; If Δσ1 is negative, the f1(t) will be negative. Hence, the f1(t) will be positive before the compressive stress peak, and will turn to be negative after the compressive stress peak.

In condition of uni-axially tensile loading, the load is to cause temperature drop in that σ2 is constant zero and the negative σ1 will linearly increase with loading. If Δσ1 is positive, the f1(t) will be positive; If Δσ1 is negative, the f1(t) will be negative. Hence, the f1(t) will be negative before the tensile stress peak, and will turn to be positive after the tensile stress peak.

In condition of compressively-sheared loading, the rock sample will be compressed by the normal component of the load. However, as to the central shearing plane, it will suffer not only compressive stress but also shearing stress. The σ1 refers to the positive compressive stress normal to the shearing plane, while the σ2 refers to the shearing stress which is actually negative tensile stress along the shearing plane. Hence, the ΔAIRT is decided by the sum of compressive stress and the tensile stress. If Δ(σ1+σ2)0 , the f1(t) will be positive; If Δ(σ1+σ2)0 , the f1(t) will be negative. As to Figure 1c, it’s easy to know that Δ(σ1+σ2) will be positive if the shearing angle γ450 , Δ(σ1+σ2) will be zero if the shearing angle γ=450 , Δ(σ1+σ2) will be negative if the shearing angle γ450 .

  1. 2) Pore gas desorbing & escaping effect: f2(t)

Any rock has pores of different size more or less inside. Some rock, especially the sedimentary rock, may has certain gas, such as CH4, CO2, CO and O2, enclosed in the pores or/and absorbed on the pore surface (Wang, 2003; Yang et al., 1999). Usually, most of the pores are enclosed and the gas molecules stay inside both in free gassy state and absorbed state. If the rock is loaded and suffers deformation, the volume of pores will decrease which results in the escape of gas from the pores. Once the load and deformation cause the enclosed pores getting fractured, the gassy molecules will escape firstly and the absorbed molecules will get desorbed to be gassy molecules and escape later. Both desorbing and escaping actions need to make use of heat energy from the rock, and thus will result in the AIRT drop of rock surface. Hence, the f2(t) is always negative, and the more the pores and gas enclosed, the more the negative effect of f2(t) .

  • 1) Fracture effect: f3(t)

With loading and deforming, the rock is to get fractured. The new produced fractures together with the initial fissures and joints will extend both in width and length. The production of new fractures needs to consume energy, and the extension of fissures, joints and fractures also needs to consume energy. Hence, the f3(t) is always negative, and the more the fractures produced and fissures, joints and fractures extended, the more the negative effect of f3(t) .

  • 2) Frictional thermal effect: f4(t)

With loading and deforming, the friction action is to occur between rock fissures, rock joints, rock grains, and new produced fractures. The friction action could be interpreted as: 1) at the beginning stage of loading, the friction may only be resulted from between rock fissures and between rock joints; 2) later, rock deformation increases with loading, new fractures are produced, and the friction between rock grains and between new produced fractures will join in; 3) finally, at the ending stage of loading, the rock deformation and fractures will be sufficiently developed, and the frictions between rock grains and between new produced fractures will be the chief contributors to the frictional thermal. In a word, the f4(t) is always positive no matter what is the principle friction factor. The more the friction, the more the positive effect of f4(t) .

  • 3) Heat transfer effect: f5(t)

In the process of loading and inside the respectively independent loading system, the heat exchange is inevitable between the rock sample and the load header, the shearing platen or cushion-blocks, the surrounded atmosphere, etc. If the current temperature of rock sample is higher than the others, the heat of rock sample will be transferred out to whose temperature is lower. If the current temperature of rock sample is lower than the others, the heat of the others will be transferred into rock sample. Hence, the temperature of rock sample is a dynamic balance behavior between the heat transferred in and the heat transferred out. If the heat transferred in is more than that transferred out, the f5(t) will be positive; otherwise, it will be negative.

  • 4) Environment-radiation effect: f6(t)

The IRR detected by infrared imager includes not only the direct radiation from rock surface itself, but also the reflected radiation from environment. In laboratory, the chief environment radiations are the scattering sunshine, the moving human bodies and the illumination lamp. For the uncertain change of scattering sunshine, the movement of human bodies before the loaded rock sample, and the fluctuation of illumination light, the environment radiation effect on rock sample will be random. Hence, sometimes f6(t) may be positive, but sometimes f6(t) be negative. To eliminating the environment-radiation effect, the human bodies inside the laboratory were not permitted to move during testing process, the illumination lights were turned off, and the windows as well as its curtains were closed. Furthermore, some experiments were conducted in the evening so as to avoid the scattering sunshine completely.

3.3. Experiment Interpretation with RSRM-Model

Due to the comprehensive effects of f1(t) ~ f6(t) , the evolution of AIRT will be complex and will result in different possibility for abnormal AIRT precursors in different rock loading conditions, which including the loading scheme, rock type, and environment parameters. The following discussions are based on that f5(t) and f6(t) can be ignored.

3.3.1. Uni-axial loading experiments

For uni-axially loaded rock, f4(t) will take place only after that the rock has sufficiently deformed and the fractures have sufficiently developed.

At loading stage I and stage II, the rock surface AIRT are decided by f1(t) and f2(t) . If the loaded rock is igneous rock or metamorphic rock, f2(t) is very rare since no gas absorbed in its pores usually, and the AIRT will rise with loading. If the loaded rock is sediment rock and with gas closed and absorbed in pores, f2(t) is inevitable, and the AIRT will rise if f1(t)|f2(t)| , or be constant if f1(t)=|f2(t)| , or drop if f1(t)|f2(t)| .

At loading stage III, fractures get sufficiently developed and pores get seriously damaged. The f3(t) begins to has more and more effect on evolution process of AIRT. AIRT will rise if f1(t)|f2(t)+f3(t)| , or be constant if f1(t)=|f2(t)+f3(t)| , or drop if f1(t)|f2(t)+f3(t)| .

At loading stage IV, the friction action starts and f4(t) begins to have more and more effect on the evolution of AIRT. AIRT will rise if [f1(t)+f4(t)]|f2(t)+f3(t)| , or be constant if [f1(t)+f4(t)]=|f2(t)+f3(t)| , or drop if [f1(t)+f4(t)]|f2(t)+f3(t)| . Since f2(t) is very rare for igneous rock or metamorphic rock, the rise speed of AIRT of loaded igneous rock or metamorphic rock will get fast at stage IV, and the speed turning point is suggested to be the precursors point.

3.3.2. Compressive shearing experiments

For compressively sheared rock, not only f1(t) but also f4(t) is decided by shearing angle ( γ ). If γ450 , f1(t) will be positive for Δ(σ1+σ2)0 , and the friction action will be much strong in that σ1 , which is normal to the friction plane, is large. If γ=450 , f1(t) will be zero since Δ(σ1+σ2)=0 . If γ450 , f1(t) will be negative for Δ(σ1+σ2)0 , and the friction action will be much week since σ1 is slight.

3.3.3. Biaxial loading experiments

For bi-axially loaded rock, f1(t) will always be positive. As to bi-axially loaded en echelon faults, collinearly and non-collinearly disjointed faults, f2(t) , f3(t) and f4(t) will occur simultaneously, and the evolution of AIRT will be fluctuated. If [f1(t)+f4(t)]|f2(t)f3(t)| , AIRT will rise; if [f1(t)+f4(t)]|f2(t)f3(t)| , AIRT will keep in the same level, and if [f1(t)+f4(t)]|f2(t)f3(t)| , AIRT will drop. Usually, the fact is that [f1(t)+f4(t)]|f2(t)f3(t)| at the fracturing stage, and there is a short drop of AIRT, which is called as ‘silence’ before EQ (Ohtake et al., 1981). However, f4(t) will be an important factor at the later stage of loading for the concentrated formation of fractures and friction in the disjointed zone, and the final state of AIRT will be rise for [f1(t)+f4(t)]|f2(t)f3(t)| .

4. Earthquake Thermal Infrared Anomalies

The prediction of EQ is very difficult, but it's not impossible. A number of signs warning of EQs, such as foreshock activities, peculiar animal behaviour, increased low frequency EM-noise, concentrations of radon in water and air, ionosphere and magnetosphere perturbation, radio frequency emissions, terrestrial gas emanations, EQ clouds, and satellite TIR anomalies, have been proposed and reported during the past centuries. Satellite TIR anomaly was firstly reported in 1989, and had been repeatedly verified in the world during the past 20 years. It is becoming a prospecting space observation technology for seismic activity monitoring and for EQ predicating.

4.1. General Features of EQ TIR Anomaly

Gorny (1988) firstly reported that there were large area of TIR anomalies in METEO satellite remote sensing images, spatial resolution being 5 Km and wave length being 10.5~12.5μm, before many moderate-strong tectonic EQs in the mid-Asia and the east-Mediterranean region. Tronin (1996) analyzed 10000 about TIR images of channel AVHRR-2 of NOAA in ten years for the mid-Asia, and reached that there existed average anomaly, 1~5℃, before the EQs at this region, and that there was obvious statistical relations between the EQ and TIR anomaly. Qiang (1990), Cui (1999), Liu (1999), Xu (2000), Zhang(2002), Ouzounov(2004), Arun (2005), Liu (2007), Wu (2008) also reported that there occurred TIR anomalies in satellite images (NOAA, FY, MODIS) days before more than 100 EQs in Asia (China, India, Iran, Japan, Kamchatka, Pakistan, Turkey) and Europe (Italy, Greece, Spain). Analysis to all the reported cases, it was uncovered that satellite TIR anomaly before EQ has the following features generally (Wu et al., 2009):

  1. Temporal features: Satellite TIR anomaly usually appears 1~26 days before shock, and reaches to its peak 1~2 days before shock, and will disappeared soon after shock.

  2. Spatial features: The spatial distribution and geometrical shape of TIR anomaly is tightly related with tectonic structures such as plate borders and active faults. With the EQ impending, the TIR anomaly will move to or extend to gradually the coming epicentre along the structures.

  3. Temperature features: the temperature of the TIR anomaly is usually 2~6℃ higher than that of outside or surrounding the TIR anomaly.

  4. Magnitude features: there are positive relations somewhat between the TIR anomaly energy (the anomaly area times the anomaly temperature) and the magnitude of future shock.

4.2. Anomalies Interpretation with RSRM-Model

The RSRM experimental results are applied to analyze satellite remote sensed TIR anomalies before several strong EQs in Asia. Referring to the seismogenic mechanisms, the satellite TIR anomalies are in good accordance with the detected IRR anomalies from rock fracturing and seismogenic experiments with fault system being simulated with disjointed faults and intersected faults.

4.2.1. Dongsha Ms5.9 EQ 1992 in Taiwan, China

Dongsha Ms5.9 EQ occurred in Taiwan on Sept 14, 1992. The NOAA satellite images show that there was TIR anomaly appeared along the regional faults and downfaulted basins before shock as in Figure 26 (Wu et al., 2006c). There was an isolated and spoon-shaped high temperature area to the southwest of Taiwan Island 25 days before shock (Aug 19, 1992). The head of ‘spoon’ locates in the downfaulted basin-Ⅳ and the handle of ‘spoon’ distributes along fault-6. Satellite TIR image on Aug 22 shows that the high temperature on ‘spoon head’ diminished, but the TIR anomaly on ‘spoon handle’ became wide. Later, the anomaly moves gradually to the epicenter. TIR image on Sept 9 shows that a large area of high temperature had appeared around the coming epicenter, and the maximum temperature (in brown color) appeared south-close to the coming epicenter. It indicated that the TIR anomaly was consistent with the regional tectonic structures (faults and basins) in spatial position and geometry.

media/image167.jpg

Figure 26.

Satellite TIR anomaly images before Dongsha Ms5.9 EQ (Sept 14, 1992).

4.2.2. Zhangbei Ms6.2 EQ 1998 in China

Zhangbei Ms6.2 EQ occurred in China at 11:50 am, Jan 10, 1998. With the overlay of the active fault system investigated and deduced, in red lines, on night time NOAA satellite infrared images, as in Figure 27, it was discovered that (Wu et al., 2007b): 1) 18 days before shock, Dec 28, 1997, the TIR images on land and sea surface are basically normal, and the contours of TIR temperature field of Bohai Bay appeared to be along the coastline; 2) 5 days before shock, Jan 5, 1998, there occurred a positive TIR anomaly strip from Bohai Bay to Zhangbei passing through Beijing, its temperature was 3 C higher than that of outside, and the contours of TIR temperature field of Bohai Bay got offset the coastline; 3) 1 day before shock, Jan 8, 1998, the positive TIR anomaly strip had got wider and its temperature was 7 C higher than that of outside, and the contours of TIR temperature field of Bohai Bay got in accordance with the positive TIR strip; 4) 1 day after shock, Jan 11, 1998, the pattern of TIR images on the surface of land and sea turned to be normal again.

There are several disjointed active faults along the positive TIR strip, including a possible uncovered great active deep fault going from Bohai Bay to Zhangbei. Besides, to the southwest of Zhangbei, there are two small active faults pointing to Zhangbei. If extend the three faults respectively towards Zhangbei, it could be found that Zhangbei is exactly the intersection point. Hence, the tectonic background around the epicenter is basically comprised of three intersected active faults as two groups, i.e., the primary fault is the independent one from Bohai Bay to Zhangbei, and the secondary two act as another group. The two groups of faults split basically the regional crust into three geo-blocks (geo-block A, B and C1+C2) as the RSRM experiment model in Figure 28b (rock block A, B and C), and the secondary two faults act as the two sides of an acute wedge-shaped geo-block (geo-block C2) as the RSRM experiment model in Figure 28d (rock block C). The geometric features of the faults and the spatio-temporal features of TIR anomaly were similar. Therefore, the mechanical mechanism of Zhangbei EQ should be classified to be the failure of an intersected active fault system, which with an acute wedge-shaped geo-block being its secondary active object.

media/image168.jpg

Figure 27.

The TIR images overlaid with detailed active faults before and after Zhangbei Ms6.2 EQ (Jan 10, 1998).

media/image169.jpg

Figure 28.

The IRR anomaly features of simulated tectonic activities due to bi-axially load (Wu et al., 2007b).

4.2.3. Izmit Ms 7.8 EQ 1999 in Turkey

Izmit Ms 7.8 EQ occurred in Turkey on Aug 17, 1999. In the epicenter zone, there were two en echelon tectonic faults, and the epicenter is 45km about to the west of the disjointed place of the two en echelon faults, as in Figure 29. With thermal image of Aug 1&2 being the reference, the differential thermal images from Aug 6 to Aug 26 were obtained, and it was revealed that there were NOAA satellite TIR anomaly at the disjointed zone of two en echelon faults since Aug 6, 11 days before shock (Tronin, 2000), as in Figure 29. The RSRM experiments on the simulation of tectonic EQ due to the fracturing of disjointed faults had revealed that there were IRR anomaly increment and concentrated deformation at the disjointed zone, as in Figure 30. Obviously, the spatial features of NOAA TIR anomaly before Izmit Ms 7.8 EQ were the same as that of IRR anomaly before the fracturing of disjointed faults.

media/image170.jpg

Figure 29.

TIR anomalies 0~11 days before Izmit Ms 7.8 EQ (Aug 17, 1999).

media/image171.jpg

Figure 30.

The IRR anomaly and deformation concentration of a bi-axially loaded granite sample with disjointed faults.

4.2.4. Hengchun Ms7.2 EQ 2006 in Taiwan, China

Hengchun Ms7.2 EQ occurred in Taiwan at 12:26pm, Dec 26, 2006. TIR images from stationary satellite FY-2 showed that there was TIR anomaly nearby the coming epicenter. The TIR anomaly appeared to the east of Philippines six days before shock, and moved gradually toward west to Philippines. Later, the anomaly changed its direction to the north, and the temperature increased 10oC about one day before shock. Figure 31 shows that the satellite TIR images around Hengchun on Dec 25, 2006 (Liu et al., 2007). A disjointed thermal strip (in orange color) appeared on the southwest of Taiwan at 1:00am, and developed gradually to be an X-shaped thermal anomaly zone at 10:00am. The epicenter located closely to the cross point of the X-shaped zones. The evolution process of satellite TIR images was very similar to that of X-shaped thermal IRR anomaly in RSRM experiments as in Figure 2.

4.2.5. Wenchuan Ms8.0 EQ 2008 in China

Wenchuan Ms8.0 EQ occurred in China at 14:28 am, May 12, 2008. The epicenter locates at the transferring zone of Tibetan Plateau to Sichuan Basin, which is only 92Km to the NW of Chengdu, the capital of Sichuan province. Analysis to FY-2C TIR images, as in Figure 32, it is discovered that there was an high temperature strip with length of 3 000 km appeared on April 23, 20 days before shock, which started from India Plate and developed to northeast along the east front of Tibetan and Loess Plateau (Wu et al., 2008). The cause might be the great accumulation of fictional sliding stress along the east foreland of Tibetan and Loess Plateau due to the subduction of India Plate to Euro-Asia Plate. The east foreland of Tibetan and Loess Plateau act as the fictional sliding plane between the west part of China (including Tibetan and Loess Plateau) and the east part of China (including the North, Middle and Southwest China). The evolution of satellite TIR images was similar to that of detected strip-shaped IRR anomaly in the fictional sliding experiments as in Figure 8 and Figure 9, which shows the evolution of TIR anomaly along the sliding plane.

media/image172.jpg

Figure 31.

Satellite TIR images one day before Hengchun Ms7.2 EQ (Dec 25, 2006).

media/image173.jpg

Figure 32.

Satellite TIR images before Wenchuan Ms8.0 EQ (May 12, 2008).

5. Future Researches

5.1. On RSRM Experiment

As a detectable remote sensing signal related with rock stress and physical temperature, IRR is a meaningful index for studying rock load, rock deformation, rock fracturing and rock hazard. The temporal evolution of surface IRR from loaded rock is the comprehensive effect of rock thermo-elastic acting, pore gas desorbing & escaping, fractures producing & extending, rock frictionating, heat transferring and environment radiation. The IRR image anomaly referring to the spatial-temporal evolution of IRR from loaded rock is an important precursor for rock fracturing, and will be meaningful for the predication of geo-hazards including tectonic EQ. For the practical application of RSRM, deeper research on IRR imaging detection quantitatively and specially on rock stress and rock hazard for experimental rock mechanics, rock engineering, tectonic activity and strong EQ is demanded.

The mechanism of experimental detected IRR anomaly can be theoretically interpreted by taking the load header, the rock sample and the environment to be a closed independent system in energy balance state. There are two of the main rock-physics mechanisms, respectively being thermo-mechanical coupling and frictional thermal due to tectonic stress, rock fracturing and fictional sliding, for the change IRR from loaded rock samples (Wu, et al., 2006c). Besides, positive hole (P-hole) activation due to piezoelectricity was suggested to be another mechanism of IRR from loaded quartz-bearing igneous rock (Freund, 2002), such as granite, basalt, diorite, and gabbro.

To search for scientific interpretation on the relations among satellite TIR anomaly, rock stress and experimentally detected IRR anomaly, the EM transferring process from underground rock body to satellite sensors, through lithosphere, earth surface coversphere (including soil layers, water bodies and vegetations), the atmosphere and lithosphere, should be systematically studied. Pulinets pointed out that the incubation of an EQ is to disturb the ionosphere (Pulinets, 1998), and it was suggested that lithosphere-atmosphere-ionosphere (LAI) coupling is the mechanism of satellite TIR anomaly before strong EQ (Molchanov et al., 2004). Nevertheless, the action of earth surface coversphere on the transferring and the magnifying of EM signals from underground loaded rock to atmosphere should not be ignored, even if its physical mechanism are not clear. For the scientific interpretation of satellite TIR anomaly before strong EQ, the lithosphere-coversphere-ionosphere (LCA) coupling is the key, while for the scientific interpretation of ionosphere anomaly, the lithosphere-coversphere-atmosphere-ionosphere (LCAI) coupling should be focused. However, present experiments on LAI, LCA and LCAI coupling are rather insufficient. Future experiments specially designed to uncover the mechanisms & laws and to construct the models & quantitative equations of LAI, LCA and LCAI couplings are expected.

5.2. On EQ Thermal Infrared Anomaly

Although there are uncertain influences from meteorological variation, satellite TIR anomaly has quite different identification features from that of unseismology-resulted TIR anomaly. Satellite TIR remote sensing is becoming a prospecting technique for monitoring tectonic activities and for predicting strong tectonic EQ, which provide a negativism to that EQ cannot be predicted. Anyway, the practical predication of EQ is not so easy. The regional tectonic background and the active fault system have extremely important affects on the incubation of EQ and the TIR anomaly. Especially, the intersected faults, compressively-sheared faults, and disjointed faults are to control the location and the routing of the spatial evolution of satellite TIR anomaly, and the brightest spot of TIR anomaly along the fault, or at the intersection point, or at disjointed zone of faults might foretell the possible epicenter (Wu et al., 2007b).

First of all, massive observation information including crust stress, land deformation, atmosphere components, underground water, surface and near-surface temperature, satellite remote sensing TIR anomaly and EM disturbance in ionosphere should be integrated together for data fusion and cross checking to analyze comprehensively the tectonic activity and rock fracturing process. A grid-based distributed database and analysis tools on TIR remote sensing images, with global and regional tectonic structures being its background, should be set up to assist the extraction of EQ TIR anomaly in different temporal and spatial scale. Besides, a quantitative model for tectonic activity analysis and for EQ magnitude predication based on TIR anomaly should be developed.

The GEOSS under construction is to provide an integral and integrated monitoring on earth environment, Geo-hazard and global disasters. A generalized remote sensing (GRS) based on the integration of space-based, aero-based, near-surface based, in-situ based and underground-based monitoring is forming in the world (Wu and Liu, 2007). The international broad and sincere cooperation, inside the framework of GEOSS without discipline exclusion and data privacy, between seismologists, remote sensing scientists, meteorologists, geophysical scientists, geochemist and spatial information scientists in good faith is expected. Besides, a powerful spatial information system, especially for EQ early warning and short-coming prediction based on GEOSS, should be developed. It should has powerful functions such as massive information classification, smart theme mapping, easy map-layer overlay, fast features extraction, effective data fusion, intelligent data mine, powerful knowledge discovery, and easy access and share.

A possible technical procedure based on GRS for satellite TIR anomaly monitoring, analyzing and early warning of tectonic EQ inside the framework of GEOSS might be that: 1) the seismological geology background being the foundation of satellite TIR anomaly analysis; 2) the long-term GPS continuous monitoring and D-InSAR measurement being the guidance of tectonic stress detection and active fault identification; 3) the underground water temperature, near surface air temperature, radon & green gas, structural cloud anomaly dairy monitoring being the forerunner for preliminary identification of coming EQ; 4) the anomaly analysis of satellite TIR, NCEP temperature and ionosphere disturbance being the dominant for early warning of temporal-spatial-magnitude parameters of EQ; and 5) the additional celestial stress on active faults being a special leading disturbance for possible tectonic EQ.

References

1 - K. Arun, C. Swapnamita, 2005 Thermal remote sensing technique in the study of pre-earthquake thermal anomalies. J. Ind. Geophys. Union, 9 3 197 207
2 - B. Brady, G. Rowell, 1986 Laboratory investigation for the electrodynamics of rock fracture. Nature, 321 29 488 492
3 - C. Cui, J. Zhang, Q. Xiao, 1999 Monitoring the thermal IR anomaly of Zhangbei earthquake precursor by satellite remote sensing technique, Proc. 20th Asia Remote Sensing Congress, 1179 1184 , Hongkong
4 - M. Deng, J. Qian, J. Yin, et al. 2001 Research on the application of infrared remote sensing in the stability monitoring and unstability prediction of large concrete engineering. Chinese J Rock Mech. Engi., 20 2 147 50 .
5 - F. Freund, 2002 Charge generation and propagation in igneous rocks. J. Geodynamics, 33(4-5): 543-570.
6 - N. Geng, C. Cui, MD et. Deng, al, 1992 Remote sensing detection on rock fracturing experiment and the beginning of Remote Sensing Rock Mechanics. ACTA Seismologic SINICA, 14(supp) : 645 52 .
7 - V. Gorny, A. Salman, A. Tronin, et al. 1998 The earth outgoing IR radiation as an indicator of seismic activity, Proc. Acad. Sci. USSR, 30 1 67 69 .
8 - J. Hudson, J. Harrison, 1997 Engineering rock mechanics, Elsevier Science Inc., New York, 85 112
9 - D. Liu, K. Peng, W. Liu, et al. 1999 Thermal omens before earthquake. Acta Seismologica Sinica, 12 6 710 715
10 - S. Liu, L. Wu, J. Li, et al. 2007 Features and mechanism of the satellite thermal infrared anomaly before Henchun earthquake in Taiwan Region. Science & Technology Review, 25 6 32 37
11 - S. Liu, L. Wu, Y. Wu, et al. 2002 Quantitative study on the thermal infrared radiation of dark mineral rock in condition of uni-axial loading. Chinese J. Rock Mech. & Engi., 21 11 1585 89
12 - M. Luong, 1990 Infrared thermovision of damage processes in concrete and rock. Engi. Fracture Mechanics, 35 (1~3): 127-35.
13 - G. Martelli, P. Smith, A. Woodward, 1989 Light, radiofrequency emission and ionization effects associated with rock fracture. Geophy. J. Int., 98 2 397 401 .
14 - O. Molchanov, E. Fedorov, A. Schekotov, et al. 2004 Lithosphere-atmosphere-ionosphere coupling as governing mechanism for preseismic short-term events in atmosphere and ionosphere. Natural Hazards and Earth System Sciences, 4 757 767
15 - D. Mounatin, J. Webber, 1978 Stress pattern analysis by thermal emission (SPATE), Proc. Soc. Photo-Opt. Inst. Engi, 164:189.
16 - A. Nicolas, J. Bouchez, J. Blaise, et al. 1977 Geological aspects of deformation in continental shear zones. Tectonophysics, 42 1 55 73
17 - M. Ohtake, T. Matumoto, G. Latham, 1981 Evaluation of the forecast of the 1978 Oaxaca Southern Mexico earthquake based on a precursory seismic quiescence. In: Earthquake Prediction Maurice Ewing Series American Geophysics Union 4), 53 62
18 - D. Ouzounov, F. Freund, 2004 Mid-infrared emission prior to strong earthquakes analyzed by remote sensing data. Adv. Space Res. 33 268 273 .
19 - S. Pulinets, 1998 Seismic activity as a source of the ionospheric variability. Adv. Space Res., 22 6 903 907
20 - Z. Qiang, X. Xu, C. Dian, 1990 Thermal infrared anomaly-precursor of impeding earthquakes. Chinese Science Bulletin, 35 17 1324 1327
21 - D. Renata, 1977 Electromagnetic phenomena associated with earthquakes. Geophsical survey, 3 2 157 174 .
22 - R. Sibson, et al. 1980 Power dissipation and stress levels on faults in the upper crust. J. Geophysical research, 85(B11): 6239-6247.
23 - W. Thomson, 1853 Trans. R. Soc. Edinburgh, 20 83 261 .
24 - A. Tronin, 1996 Satellite thermal survey-a new tool for the studies of seismoactive regions. J Remote Sensing, 17 8 1439 1455
26 - X. Wang, 2003 A research on the inorganic carbon dioxide from rock by thermal simulation experiments. Advance in Earth Science, 18 8 515 20
27 - L. Wu, S. Liu, 2007a Generalized remote sensing for solid Earth hazards under condition of GEOSS. Science & Technology Review, 25 6 5 11
28 - L. Wu, J. Wang, 1998 Infrared radiation features of coal and rocks under loading. Int. J. Rock Mech. & Min. Sci, 35 7 969 976
29 - L. Wu, C. Cui, N. Geng, et al. 2000 Remote Sensing Rock Mechanics (RSRM) and associated experimental studies. Int J Rock Mech Min Sci, 37 6 879 88 .
30 - L. Wu, J. Li, S. Liu, 2009 Infrared anomaly analysis based on reference fields for earthquake remote sensing. Seismology Geology, 2009(in press)
31 - L. Wu, J. Li, X. Xu, et al. 2007b Theoretical analysis to impending tectonic earthquake warning based on satellite infrared anomaly, Proc. 2007 IEEE Int. Geosciences &Remote Sensing Symposium, 3723 3727
32 - L. Wu, S. Liu, W. Shi, et al. 2003 Experimental study on infrared anomaly of tectonic earthquake, Proc. SPIE, Remote Sensing For Environmental Monitoring, GIS Applications & Geology Ⅲ, 5239 376 387 .
33 - L. Wu, S. Liu, Y. Wu, 2006c The experiment evidences for tectonic earthquake forecasting based on anomaly analysis on satellite infrared image, Proc. 2006 IEEE Int. Geosciences &Remote Sensing Symposium, 2158 2216
34 - L. Wu, S. Liu, Y. Chen, et al. 2008 Satellite thermal infrared and cloud abnormities before Wenchuan earthquake. Science & Technology Review, 26 10 28 29
35 - L. Wu, S. Liu, Y. Wu, et al. 2002 Changes in IR with rock deformation. Int. J. Rock Mech. & Min. Sci., 39 6 825 31
36 - L. Wu, S. Liu, Y. Wu, et al. 2004a Remote Sensing Rock Mechanics (I):laws of thermal infrared radiation from disjointed jointed faults fracturing and its meanings for tectonic earthquake omens. Chinese J Rock Mech. & Engi., 23 1 24 30
37 - L. Wu, S. Liu, Y. Wu, et al. 2004b Remote Sensing Rock Mechanics (Ⅱ):laws of thermal infrared radiation from bi-sheared faults friction sliding and its meanings for tectonic earthquake omens. Chinese J Rock Mech. Engi., 23 2 192 198
38 - L. Wu, S. Liu, Y. Wu, et al. 2004c Remote Sensing Rock Mechanics (Ⅳ): laws of thermal infrared radiation from compressively-sheared fracturing rock and its meanings for earthquake omens. Chinese J Rock Mech. & Engi., 23 4 539 544
39 - L. Wu, S. Liu, Y. Wu, et al. 2006a Precursors for rock fracturing and failure-part I: IRR image abnormalities. Int J Rock Mech Mining Sci, 43 3 473 482
40 - L. Wu, S. Liu, Y. Wu, et al. 2006b Precursors for rock fracturing and failure-part II: IRR T-curve abnormalities. Int J Rock Mech Mining Sci, 43 3 483 493
41 - L. Wu, S. Liu, X. Xu, et al. 2004d Remote Sensing Rock Mechanics (Ⅲ):laws of thermal infrared radiation and acoustic emission from friction sliding intersected faults and its meanings for tectonic earthquake omens. Chinese J Rock Mech. & Engi., 23 3 401 407
42 - X. Xu, X. Xu, Y. Wang, 2000 Satellite infrared anomaly before Nantou Ms=7.6 earthquake in Taiwan, China. ACTA Seismologica SINICA, 22 6 656 659
43 - I. Yamada, K. Masuda, H. Murakami, 1989 Electromagnetic and acoustic emission associated with rock fracture. Phys. Earth Planet. Inter., 57(1-2):157-168.
44 - Y. Zhang, W. Shen, 2002 Satellite thermal infrared anomaly before the Xinjiang Qianhai boder Ms8.1 earthquake. Northwestern Seismological J., 34 1 1 4