Layout 6 Magnitude and energy of earthquakes Beno Gutenberg and Charles Francis Richter Annali di Geofisica, Vol. 9, n. 1, 1956. ANNALS OF GEOPHYSICS, VOL. 53, N. 1, FEBRUARY 2010 ABSTRACT Discrepancies arise among magnitudes as derived from local earthquake data (ML), body waves (MB) and surface waves (MS). The relation of ML to the others is as yet not definitive; but MS – mB = a (MS – b). The latest revision gives a = 0.37, b = 6.76. Pending further research it is recommended that ML continue to be used as heretofore, but MS (and ultimately ML) should be referred to mB as a general standard, called the unified magnitude and denoted by m. Tentatively log E = 5.8 + 2.4 m (E in ergs). Revised tables and charts for determining m are given. This paper is in continuation of previous investigations [Gutenberg and Richter 1942, 1956]. The earthquake magnitude has statistical and other uses independent of the relation between magnitude and energy. Indeed, it is possible that there is no complete one-to-one correlation between magnitude and energy for large and complex tectonic events. Even so, a mean or representative relation is a legitimate object of inquiry. In attempting to refine the magnitude-energy relation it was found [Gutenberg and Richter 1956] that three imperfectly consistent magnitude scales had been in use: ML determined from records of local earthquakes according to the original definition [Richter 1935]; MS from the amplitudes of surface waves for shallow teleseisms, [Gutenberg and Richter 1936; Gutenberg 1945a]; mB from the amplitude/period ratio of body waves for teleseisms, shallow and deep-focus [Gutenberg 1945b, 1945c]. The two latter were originally adjusted to coincide near M = 7, but were later found to diverge linearly so that MS – mB = a (MS – b) (1) For a number of years reductions were carried out with a = 1/4, b = 7, converting values of mB into the corresponding MS. The result of this reduction may be designated MB. The final value given for M was a weighted mean between MB and MS. This may be taken as defining M without subscript. The adjustment between MS and MB can now be performed with considerable accuracy, using the relation (1) with revised parameters a = 0.37, b = 6.76. This is equivalent to mB = 0.63 MS + 2.5 = MS – 0.37 (MS – 6.76) (2) The revision is based on a large body of data. Magnitudes have been derived by the senior author, from surface waves and from body waves separately, for a selection of better recorded large shallow earthquakes as listed by Gutenberg and Richter [1954]. Those for which there was suspicion of depth in excess of the normal (believed to be about 25 km) were rejected. Values of mB were plotted against those of MS, and (1) derived from the plot. The values a = 0.37, b = 6.76 are comparable with those found by Bath [1955] as follows: At Pasadena, a weighted mean is taken between mB as found directly from body waves, and mS, the corresponding value derived from MS by applying the relation (1), or still better from tables and charts set up to give mS directly from surface wave data. This weighted mean is designated the unified magnitude denoted by m. In Figure 1 residuals mB – mS on the basis a = 0.37, b = 6.76 are plotted against m, using amplitude and period data from all available station bulletins, a) for all shocks in Table 13 of Gutenberg and Richter [1954] for which there was no indication of depth exceeding 30 km, excluding all uncertain or doubtful magnitudes; b) using all similar data for Table 14 of Gutenberg and Richter [1954] for 1936-1939 and 1950-1952 (inclusive). There is little indication of systematic deviation from the axis of zero residuals. The slight apparent excess of positive residuals may be due to the use of a few shocks with depths somewhat greater than supposed, which should result in a 7 Station Body wave used a b Uppsala Kiruna PZ PH SH PZ PH SH 0.45 0.46 0.23 0.59 0.50 0.30 6.3 6.4 5.6 6.2 6.5 6.1 decrease of mS. Comparable data for magnitudes below 7 are rare. Ten of eleven shocks in the California region, with magnitudes near 6, give mB – mS from + 0.1 to – 0.2; the eleventh gives – 0.4. Figure 2 is a nomogram prepared by Mr. J.M. Nordquist for the direct determination of mS from surface wave amplitudes. The corresponding values of MS and of log E from equation (6), are also indicated. The adjustment of ML to m or M cannot yet be determined so closely as that of m to M but can be stated with an error not likely to exceed 0.5 magnitude unit for those shocks (magnitudes 3 to 6) most often rated in terms of ML. Representative results are given in Table 1. Values in parenthesis are outside the observable range. Until 1954, the writers generally reported magnitudes for large deep shocks, and for large shallow teleseisms as determined from body waves, effectively in terms of MS, first determining mB and then correcting to MS by applying equation (1) or an earlier approximation to it. The correction was usually applied only to shocks of magnitude 7 or over. It now develops [Gutenberg and Richter 1956] that many outstanding difficulties disappear if the linear relation (2) is consistently extended to magnitudes below 7. Whereas shocks of the largest magnitude record with surface waves relatively large compared with the body waves, shocks of magnitude below 7 show relatively small surface waves when recorded at teleseismic distances. Many long-period instruments do not record such shocks clearly; this makes assignment of magnitude from the data of distant stations difficult. Short-period instruments in such cases may show a measurable P; the absence of recorded surface waves is then sometimes misinterpreted as evidence for deep focus. When equation (2) is used, and data for both body waves and surface waves are available, two different determinations are in effect available for either m or M. The equation gives mB = MS for a value near 6 3/4. When the magnitude does not greatly differ from this figure, problems of adjustment are minor, and reduce to judgement as to the relative reliability of the two groups of data. Although at present many more stations report amplitudes for surface waves than for body waves, some ten years' experience indicates that mB provides the better data in practice as shown by fewer systematic errors and more consistent results, as well as being theoretically preferable. In using station bulletins to determine MS, the maxima of surface waves can be used for magnitude only when the period is near 20 seconds. If the period is not specified, there is risk that the reported maximum amplitude may refer to much longer or shorter waves, which seriously falsifies MS. With some exceptions, magnitudes currently being reported in station bulletins are either ML determined from nearby stations, or MS. There is less general determination of mB, and the relation in equation (1) or (2) is often overlooked. Occasionally MS is even given for deep shocks as found directly from surface waves; if the hypocenter is deeper than about 30 km, calculation on this basis gives too low a value. Routine station bulletins issued from Pasadena continue to list magnitudes M which are either ML or MS; but beginning with 1954 the annual list of large shocks also tabulates m, which is an intermediate step toward a definitive magnitude-energy relation. The practical definition of the unified magnitude m consists in a system of tables and charts for calculating magnitude from the quotient amplitude/period for the maximum waves of the principal wave groups P, PP, and S. This quotient is used in the form q = log u/T or q = log w/T (3) where u and w are respectively the horizontal and vertical components of the ground displacements in microns and T the period in seconds. Each table or chart gives for all distances and focal depths a quantity Qsuch that for corresponding distance and depth m = q + Q + s (4) MAGNITUDE AND ENERGY OF EARTHQUAKES 8 Figure 1. Residuals mB – mS on the basis a = 0.37, b = 6.76 are plotted against m, using amplitude and period data from all available station bulletins. Table 1. Values of M, m and log E for given values of ML, using M= 1.27 (ML– 1) – 0.016 ML², m= 0.63 M+ 2.5, log E= 5.8 + 2.4 m(E= energy in ergs). 9 where s is a ground correction characteristic of the station used. Charts and tables of this type were first given by Gutenberg [1945b, 1945c], where Q was designated A. Those accompanying the present paper (Table 2, Figures 3, 4 and 5) represent no change in fundamental concept, but only a revision. The statistical processes by which the tables and charts published in 1945 were derived have now been repeated by the senior author using a much larger body of data, and, it is hoped, with greater precision. One effect has been to remove a persistent discrepancy between magnitudes determined from horizontal and vertical components; this discrepancy was discovered independently by Bath [1955]. This procedure places the unified magnitude m on a self- consistent and independent basis as satisfactory for teleseisms as that of ML for local earthquakes, and with the great advantage of being applicable directly to seismograms recorded on instruments of all types and at all stations. If desired, a formal definition for m may be phrased as follows: m – 7.0 = q n (5) at a distance of 90˚ for normal shallow focal depth, where q = log w/T refers to PZ, and the station constant s is taken as zero, representing average station ground conditions. Since the relation of ML to m is not yet on a definitive basis, the authors suggest that the «Richter scale» as defined in 1935 be retained for determining magnitudes of local shocks. For teleseisms, the use of the unified scale m is preferred and strongly recommended. For magnitudes from about 5½ to 7, the departure between the two scales is within the usual limits of error under the now existing conditions of recording and reporting amplitudes. Gutenberg and Richter [1954] have not assigned magnitudes below 6 to shocks outside the California area (within which ML is reported); such shocks are merely designated by the letter d. Above magnitude 7½ the scales diverge significantly; but then determinations from the data of numerous stations scatter increasingly, and it is advisable to distinguish carefully between determinations from body waves and from surface waves. It is urgent that magnitudes determined from seismograms at single stations should not be published unaccompanied by the amplitude and period readings on which they are based. It is hoped that before many years have passed it will be possible to express the entire range of observed magnitudes in terms of the unified magnitude m. Since the provisional use of m is especially intended for investigations relating to energy, m is being published together with the energy calculated from it by the relation log E = 5.8 + 2.4 m (6) to be established on a later page. For most types of publication the writers think it preferable to follow a suggestion by Dr. L.B. Slichter, giving the MAGNITUDE AND ENERGY OF EARTHQUAKES Figure 2. Nomogram for the direct determination of mS from surface wave amplitudes. The corresponding values of MS and of log E from equa- tion (6) are also indicated. Table 2. Values of Q for shallow shocks. value of log E together with equation (6), and so avoiding confusion due to use of numerically different magnitude scales. Most calculations of the magnitude-energy relation depend directly or indirectly on the equation for a wave group from a point source [Gutenberg and Richter 1956] E = 2π³h²vρ (A/T)²t (7) where E is energy, h is linear distance from the source, v is velocity, ρ is density, A and T are amplitude and period of sinusoidal waves, and t is the duration of the wave group (which hence contains n = t /T waves). This applies at the epicenter when h is hypocentral depth, and includes a factor which takes account of the effect of the free surface. Taking v = 3.4 km/sec for transverse waves, applying a factor 3/2 to allow for half as much energy in longitudinal waves, and using h = 16 km, ρ = 2.7 gm/cc, this reduces to log E = 12.34 + 2 qo + log to (8) where q = log A/T and the subscript zero refers to the epicenter. A fundamental empirical equation is qo = − 0.6 + 0.8 ML − 0.01 ML² (9) This is a revised result drawn from the plot of qo as a function of M for California shocks (Figure 3) [Gutenberg and Richter 1956]. Two further important equations are derived from plotted data (Figures 6): log to = − 1 + 0.4 qo (10) and qo = m − 2.3 (11) For the latter result most of the data cover a relatively small range of m. Combining these log to = 0.4 m − 1.9 (12) On the other hand, if in (10) we substitute for qo its expression in terms of ML from (9), we obtain log to = − 1.24 + 0.32 ML − 0.04 ML² (13) which differs only slightly from the corresponding equation setup empirically in Gutenberg and Richter [1956], showing that the derivation of (9) and (10) has been consistent. Combining (9) and (11) m = 1.7 + 0.8 ML − 0.01 ML² (14) This is drawn on Figure 7; it is not inconsistent with the plotted data. MAGNITUDE AND ENERGY OF EARTHQUAKES 10 Figures 3, 4, 5. From top: Revised values of Q for SH; Revised values for Q for PPZ; Revised values for Q for PZ. 11 MAGNITUDE AND ENERGY OF EARTHQUAKES If instead of (9) we had used the corresponding equation in Gutenberg and Richter [1956], which has a larger coefficient of the quadratic term, the resulting equation replacing (14) would lead to calculated values of MB − ML which for large m are systematically too small to suit the observations. This is the chief reason for revising the empirical relation between qo and ML to the form (9). If we apply the relation (2) to (14) we find MS = 1.27 (ML − 1) − 0.016 ML² (15) Equation (6) results from substituting in (8) the expressions for qo and log to from (11) and (12). It has also been verified approximately by the following calculation. For a train of n (= t/T) sinusoidal body waves, emerging to the surface of the earth at arc distance Θ from a surface source with horizontal ground displacement u, the total energy calculated as radiated from the source is E = 8π³R²ρvt (u/T)²/U²L (16) where U²/fi² = tan i di / sin Θ d Θ (17) Here E = energy, R = radius of the earth, ρ = density, v = velocity, t = duration of wave train, T = period, i = angle of incidence, fi is a factor expressing the effect of the free surface as a function of i (otherwise it depends only on Poisson's ratio; see Gutenberg 1944), and L is a factor to allow for absorption, scattering, internal friction, effects of discontinuities, etc. There are several simplifying assumptions: the earth is taken as spherically symmetrical, the effect of hypocentral depth is neglected (it is easily corrected for), energy flux is calculated by the ray method as used in geometrical optics, and the use of (16) to calculate total energy implies radiation equally in all directions from the source. In what follows it is assumed that 1/3 of the original energy is radiated as longitudinal waves, and a factor 3 is accordingly applied. Analogous equations to (17) and (18) apply to the vertical component of ground displacement, replacing u by w and U² by a similar factor W². We next take ρ = 3 gm/cc, v = 6.3 km/sec (applying to longitudinal waves), R = 6370 km; we also take q = log u/T or q = log w/T, where u and w are expressed in microns. Taking the logarithm of (17) with proper attention to the units used, we arrive at log E = 18.8 + log t + 2q − log U − log L (19) We now assume that t = to; this has been confirmed roughly by Dr. C. Lomnitz from seismograms recorded at Pasadena. Applying (12) with t in place of to, and putting q = m – Q, log E = 16.9 + 2.4 m − 2Q − 2 log U − log L (20) Comparing this with (6) we should have 2Q + 2 log U + log L = 11.1 (21) and a similar equation for the vertical component. Here, 2Q may be taken from Table 2, and log U can be calculated from (17). Working this out for the vertical component of P waves, the following values are found: Θ 20˚ 42˚ 100˚ Log L −2.0 −1.3 −1.5 The calculation cannot be extended reliably to distances less than 20˚. The contribution of absorption to log L should be about 0.4 near 100˚ and 0.3 at moderate distances. Loss by refraction at the Mohorovicic, Conrad, and other discontinuities in the crust may account for a few tenths in log L. This leaves about one unit in log L unaccounted for. If all assumptions are correct, energy flux is reduced to roughly one tenth within the first 20˚ of distance; this must occur within the upper 200 km of the mantle. If this is correct, we should expect a smaller constant term in the energy- magnitude relation corresponding to (6) for shocks at greater depths. This would agree with the relatively low energy Figure 6. Two important equations are derived from plotted data: log to = – 1 + 0.4 to and qo = m – 2.3. Figure 7. Combining equations (9) and (11), m = 1.7 + 0.8 ML – 0.01 ML ². MAGNITUDE AND ENERGY OF EARTHQUAKES 12 calculated by Sagisaka [1954] for a shock at a depth of 360 km. However, the constant term 5.8 in (6) and the coefficient 0.4 in (12) are not accurately fixed, and log E calculated from (6) may be in error by as much as one unit. Acknowledgements. Contribution n. 750, Division of the Geological Sciences, California Institute of Technology, Pasadena, California. References Bath, M. (1955). The Problem of Earthquake Magnitude Determination (unpublished). Gutenberg, B. (1944). Energy Ratio of Reflected and Refracted Seismic Waves, Bull. Seismol. Soc. Amer., 34, 85-102. Gutenberg, B. (1945a). Amplitudes of Surface Waves and Magnitude of Shallow Earthquakes, Bull. Seismol. Soc. Amer., 35, 3-12. Gutenberg, B. (1945b). Amplitudes of P, PP, and S and Magnitude of Shallow Earthquakes, Bull. Seismol. Soc. Amer., 35, 57-69. Gutenberg, B. (1945c). Magnitude Determination for Deep-focus Earthquakes, Bull. Seismol. Soc. Amer., 35, 117-130. Gutenberg, B., and C.F. Richter (1936). On Seismic Waves (third paper), Gerlands Beitrage zur Geophysik, 47, 73-131. Gutenberg, B., and C.F. Richter (1942). Earthquake Magnitude, Intensity, Energy and Acceleration, Bull. Seismol. Soc. Amer., 32, 163-191. Gutenberg, B., and C.F. Richter (1954). Seismicity of the Earth, second ed., Princeton Press. Gutenberg, B., and C.F. Richter (1956). Earthquake Magnitude, Intensity, Energy and Acceleration (second paper), Bull. Seismol Soc. Amer., 46, 2, 105-145. Richter, C.F. (1935). An Instrumental Magnitude Scale, Bull. Seismol. Soc. Amer., 25, 1-32. Sagisaka, K. (1954). On the Energy of Earthquakes, Geophys. Mag. Tokyo, 26, 53-82. © 2010 by the Istituto Nazionale di Geofisica e Vulcanologia. All rights reserved.