DE2440312A1 - EARTHQUAKE MONITORING PROCEDURE - Google Patents

Description

PATENT LAWYERS

DR-ING. RICHARD GLAWE ■ DIPL-ING. KLAUS DELFS ■ DIPL-PHYS. DR. WALTER MOLL MUNICH HAMBURG MCWOH & Nq O 1 O

8MDNCHEN26 2HAMBURG52

PO Box 37 WAITZSTR. 12th LIEBHERRSTR. 20 TEL. (040) 89 22

TEL. (089) 22 65 48 TELEX 21 29 21 spec TELEX 52 25 05 spec

YOUR SIGN YOUR MESSAGE FROM OUR SIGN MONKS

A 68

REGARDS:

PETTY-RAY GEOPHYSICAL, INC. Houston, Texas / USA

Earthquake monitoring procedures

The invention relates generally to the study of seismology in geophysics, and more particularly to the Earthquake monitoring with improved interpretation and analysis of seismograms.

In the case of reflection earthquake monitoring, a Explosives, shock or vibration devices induce sound waves in the earth and the reflections that occur

509810/0312

ORIGINAL INSPECTED

These sound waves are picked up by geophones or detectors, which are arranged in flocks or nests (arrays or nests) on the surface of the earth. When sound in seismic form Energy is induced at a surface location, the energy is propagated down into the earth as wave fronts. In in any area where the speed of sound changes, part of the seismic energy is reflected upwards and recorded by the geophones. The outputs of the geophones are in the form of oscillograms or oscillographic traces recorded that have certain characteristic waveforms. These waveforms are called "seismic events" or simply referred to as "events" when they can be correlated from track to track.

Recently, the precise determination and interpretation of the various properties of seismic data and their lateral changes along the seismic profiles in determining the lithographic and printing conditions and in localizing Sands containing petroleum and natural gas have been shown to be advantageous. Among the most important of these properties is the total velocity information derived from the wave trains with common reflection points in the interface. Common reflection point earthquake monitoring is disclosed in U.S. Patent 2,732,906 described by Mayne. According to this method, seismic wave trains are common reflection points in the interface first for a normal move-out for the business

509810/0312

speed information corrected and then totaled (stacked) to create a single wave train.

Many methods for determining aggregate velocities from seismic data have been developed. Some of these methods are summarized in a "Search and Research" paper presented by Enders A. Robinson at the Mid-Western Meeting of the Society of Exploration Geophysicists on April 24, 1969. John C. Robinson also describes such a method in a paper ¹¹ HRVA A Velocity Analysis Technique Applied to Seismic Data "in the journal Geophysics, Vol. 34, Hr. 3 (June 1969), p. 330. Another method is described in a paper by Y / .H. Schneider and MM Backus "Dynamic Correlation Analysis", Geophysics, Volume 33, No. 1 (February 1968), page 105. Most of these earlier attempts to calculate seismic velocity resulted in a summation of the squared or absolute value through the wave trains in hyperbolic windows and left the geophysical analyst to identify the seismic reflection events simply on the basis of points of highest coherence (measured multiple times) in the vertical incident time-velocity plane. The article "Semblance and other Coherency Measures for Multi-Channel Data ", Geophysics, Vol. 36, Kr. 3 (June 1971), page 482, by NF Heidell and M. Turhan Saner describes this s procedure in detail. These methods did not take into account the shapes of the isocoherent velocity profiles in

509810/0312

the vertical time-velocity plane of incidence and yielded seldom precise time or sign information in the case of seismic ones Reflection events.

The present invention provides a method for earthquake monitoring with more accurate automatic speed analysis with means for distinguishing between apparent speed information based on primary reflections and those with random coincidence of seismic noise, and with facilities to create a novel form of representation that is of great use to the geophysical analyst about the velocity information with the usual earthquake information. These improved calculations of the Seismic velocities with measurements of their accuracies are in the depth point-vertical time-of-incidence plane shown together with the amplitude and sign information of the seismic reflection events. These data and this type of representation is particularly relevant in the case of manual or automatic generation and representation of one another constant velocity contours of the seismic data are useful.

The method according to the invention allows the speed scanning uncorrected wave train sets with a common reflection point at every point in time in the wave train set Help a number of hyperbolic windows that are about half that

509810/0312

- 4 -

are long (16 milliseconds) like the 'duration (20 - 40 milliseconds) of a typical data wave train in the wave train set. The difference in time between successive periods of time is set equal to half the window width in the typical Pail, i.e. 8 milliseconds, and the time is marked accordingly. The measurement of amplitude coherence used in the preferred embodiment of the invention is an algebraic summation of the amplitude of the wave trains at particular points within the hyperbolic Time window.

In a preferred embodiment of the present invention, the rate steps used in the rate scan for each value of vertical time of incidence are not uniform, but are such that they result in a single sample time change at the intersection of the hyperbolae on the furthest distance wave train of the wave train set , which was changed by the wave train attenuation at earlier recording times. Speed steps or speed quantization levels are calculated in terms of a formula with respect to a "speed control function" which is known for the region which has been examined _{for several values of the vertical time of incidence T Q.} These steps are plotted in relation to the speed control function in the speed sT _{Q sphere.}

5 0 9 8 1 0 / £ 3 1 2

Each measurement of identified amplitude coherence produced by this method can be converted into its corresponding Coordinate in the TQ speed function number plane being represented. It has been found that when representing implitude coherence values in this way, obtained by analyzing a noise-free artificial wave set with a common reflection point, isocoherent Outlines created by any man-made seismic event, roughly as concentric ellipses appear. The common major axis of these ellipses is inclined instead of parallel to the velocity axis. One two-dimensional least squares fit or a other corresponding two-dimensional interpolation technique is applied to the amplitude coherent surface that is results from the analysis of the actual seismic field data, and various error criteria, such as various mean Error Squares (EMS), allow the goodness of each obvious Determine speed information. The coordinates of the optimal quarter mean square error fit (least-squared-error-fit) or another equivalent two-dimensional interpolation technique result in the best possible Determination of the speed and time of this reflection event.

The sign and the amplitude of the detected seismic reflection events become depth point by depth point as functions of the vertical time of incidence T (interpolated to the

509810/0312

next millisecond) and the speed information (interpolated to the next tenth of the speed function). This information is eventually stored in of a kind yet to be explained in detail.

A branched search process (tree search) is carried out one after the other applied to data sets that have a predetermined number of reflection points for all time values. If there is coherence in time and sign under a predetermined number of depth points for a given seismic Horizon has been found and if the speeds of these events essentially coincide for at least four or more events, the information is for the user identified both the temporal coherence of the events and the calculated speed represents the earth to this horizon.

In the following, examples of the method according to the invention are explained in more detail with reference to the accompanying drawings. Show it:

Pig. 1 shows an idealized form of a multi-add set suitable for the processing according to the invention with common Reflection point,

Pig. 2 a representation of the areas enclosed by a series of curves in the vertical time of incidence

509810/031 2

Speed level used in the invention Determines speed quantization intervals,

Pig. 3a an idealized seismic reflection event, as it is in a wave train set with a common reflection point appears,

Fig. 3b hyperbolas for a vertical time of arrival and six speeds are determined,

3c shows an example of a hyperbolic window which is used for scanning the seismic wave train set,

Figures 4a to 4f show three groups of hyperbolic windows intersecting the seismic reflection event of Figure 3, and the effect of their respective Tq on the amplitude pattern,

Figures 5a through 5d are tables showing amplitude coherent patterns in a portion of the vertical time-of-incidence-velocity-function number space at a single depth point for a single seismic reflection event of positive amplitude represent,

Fig. 6 is a pictorial representation of a portion of the vertical time-of-arrival depth point plane with a structure search has been carried out according to the invention, and

509810/0312

7 shows part of an output representation according to the invention.

Fig. 1 shows a group of seismic wave trains which are suitable for processing according to the invention. The seismogram consists of individual wave trains 12, which have been recorded with different distances between the source and detector are, but have common reflection points in the sub-area. The seismogram shown in Fig. 1 is a idealized artificial data representation and it should be noted that a normal field of recorded seismic Data degraded by random noise, multiple reflections, and surface disturbances is. Some of these forms of interference occur on Each wave train begins and is usually eliminated by taking each wave train from its first data point to the attenuation line 16 is attenuated. Processing of such data also typically includes retrieval the amplification, detangling, filtering out of the noise reduction and static correction, all of these measures in can be carried out in a known manner 'and as described in many issues of Geophysics, the Journal of the Society for Exploration Geophysicists, and are known to those skilled in the art.

It is known "that the wave trains of a wave train set with a common reflection point after increasing distances

509810/0312

are arranged between the source and detector, as shown in Fig. is shown, with seismic reflection events appearing as groups of wave parts that approximate along hyperbolas occur by the expression

ρ __ 2

T = Λ / ^Τ Γ> ⁺ (^ T ") seconds (1)

be determined.
It is

T = the running time from the shooting point (source) to the Reflector and to the receiver,

T ₀ = the vertical time of arrival, χ = the distance between the source and receiver,

ν = the total velocity associated with a given seismic event.

Such a hyperbolic arrangement of seismic wave lengths is shown in Fig. 5a. An alternative form of the Equation (1) is

ν = - Meters per second (2)

The total velocity associated with the seismic reflections varies as a function of vertical time of incidence, as shown by curve 22 of FIG. The curve 22 is unique for each examination area and is indicated by support points 23 with several values of the vertical input

509810/0312
- 10 -

if time is set, the support points 23 being known from previous seismic surveys in the area. The curve 22 changes for different wave train sets with a common reflection point in a "limited geographic area Area not drastic, but only shows small changes due to lateral changes in the lithology, surface pressure, local convolution, etc. The curve 22 therefore defines a "speed control function" for a "limited geographic area Area fixed, and each wave train set with a common reflection point is based on its speed deviations investigated from this "G-speed control function".

The procedure for analyzing the wave train theorem begins by generating the support points of a family of speed function number curves, where the running time associated with the interpolation point of each curve is a multiple of a selected time interval with respect to the corresponding transit time is that with the support point at the same vertical incidence time for the speed guide curve communicates.

In particular, speed steps with respect to curve 22 are defined for the vertical times of incidence, which determine the support points of curve 22. These steps are calculated using equation (1) and can be plotted with respect to the speed control function in the speed T ₀ plane. At each interpolation point 23 the initial value of T becomes corresponding to the speed control function

509810/0312

- 11 -

at the effectively furthest wave train (effective farthest trace) determined by solving equation (1) for T, the wave train being determined by the attenuation line 16 and the line 75. When a fixed step of time £. then added to this value of T and equation (2) is solved for ν, the support points 31 of the curve 30 are fixed _{at each value of T Q.} Then the support points 32 of the curve 29 are calculated by subtracting £ from the original value of T and then by solving equation (2) for the ν at each value of the vertical time of incidence which determine the support points 23 of the curve 22. This process of adding and subtracting the time from the original value of T in successive steps, ie + 2f,, + 3 ^, etc., is repeated until the support points of a desirable number of curves have been established. In a preferred embodiment of the invention, the value of T + is four milliseconds, and the number of curves for which the support points have been established is usually twenty-four. As shown in FIG. 2, the curves 20 and 21 are limit values for the speeds which are considered in the analysis of the wave train set 10 of FIG.

When all the support points of FIG. 2 have been established, the curves of FIG. 2 are created by connecting the support points formed with straight lines. It should be noted that above 1.4 seconds the curves 20 and 21 diverge with increasing time, since the time after this point the area 25 of the

509810/0312

- 12 -

End of attenuation exceeds. Before this end time of the damping zone, the curves can diverge towards earlier times. The reason for this is that at times such as 26 and 27 fewer wave trains make up the sentence (because of the damping), and therefore a moving part in which the now is effective far away wave train (effective far trace), a larger speed step at the furthest away x value 75. If the curves of the Pig. 2 in the area of the Constricting the attenuation end 25 too much, it is desirable ' that a new damping line 16 is established. After this If a new attenuation line has been established, the procedure described above would be repeated until the curves in the Area of the damping end 25 are set in a suitable manner. The geophysical reasons for choosing an attenuation force are well known to those skilled in the art.

The time difference between any two adjacent vertical arrival times that the support points 23 of the curve 22 Set in Fig. 2 is usually on the order of 100 milliseconds or more. In a preferred embodiment of the invention, there are many vertical incidence times between any two adjacent time values are provided that represent the support points 23, and at these times it is desirable to use the number curve associated with each speed function curve Knowing speed. These speed values are obtained by interpolating the speed at each vertical time of incidence from the rectilinear section

509810/0312

- 13 -

hold, the two adjacent support points for each speed function number curve connects.

All time calculations performed to determine the support points of the curves in FIG. 2 were made on the effective date most distant wave train carried out. If every x-value that corresponds to each wave train of wave train set 10 of FIG. 1, is substituted into equation (1), the values of! are the same number as the wave trains of the wave train theorem 10, generated for each speed function number curve of FIG. 2 at the value of the vertical time of incidence, at which the interpolation was carried out. These calculated Values of T determine a hyperbola for each velocity function number with the value of the vertical arrival time, and the hyperbolas thus determined for six speed function numbers are shown in FIG. 5b. the The interpolation and computation steps described are then used to identify hyperbolas at each to determine the vertical fall transit time in the wave train set 10.

Fig. 5c shows a hyperbola 70 thus determined at a given vertical time of incidence Iq. If 4 and 8 milliseconds to each value of T and each wave train in the wave train set for hyperbola 70 are added and subtracted, becomes a hyperbolic window 72 shown in FIG. 5c generated. This is done for each hyperbola 70 for each speed

509810/0312
- 14 -

244Q312

performance function number is performed at each value of T _Q. The process of adding and subtracting the time to determine the boundaries of the hyperbolic window 71 gives boundary lines 72 of the hyperbolic window 71 that are "parallel" to each other, which would not be the case if equation (1) were used, to determine the limit curves 72.

The method according to the invention therefore includes the sampling of the wave set with the hyperbolic windows 71 and the calculation of a unit of measurement of the sign, the amplitude and the coherence of the data contained therein. The width 72 of the hyperbolic windows 71 can be of any suitable value, but in a preferred embodiment of the invention it is about half the duration of a typical data wave. The unit of measure of the sign, the amplitude and the coherence of the speed is the algebraic sum over the wave train set of the amplitude of each wave train at a point in time within each hyperbolic window 71, i.e. T _Q , T ₀ + 4 milliseconds and T _Q + 8 milliseconds for each speed function number .

FIGS. 4a to 4f show the effect of a single seismic reflection event over the calculated units of the amplitude coherence. FIG. 4a shows two hyperbolic windows 41 and 42, each corresponding to _{the vertical time of arrival T Qw.} As explained above, the upper and lower limit curves of each of these windows are parallel to theirs

509810/0312

- 15 -

respective center line. The hyperbolic window 41 corresponds to the greatest deflected speed (a speed line on curve 20 of Figure 2) and window 42 corresponds to the lowest deflected speed (a speed line on curve 21 of Figure 2). For these windows and those windows which correspond to the speed function numbers between curves 20 and 21 for the vertical time of arrival equal to Tq, the algebraic sum of the data contained is calculated and their size can be plotted against the speed in FIG. 4b. It should be noted that the maximum of this curve, indicated by the reference numeral 45, corresponds to a speed lower than ν, the actual speed being indicated by the reflection event at the vertical time of incidence T _Qe .

When T- equals Tq ₆ , the actual vertical time of arrival of the event, the size of the algebraic sum of the data it contains reaches its maximum possible value. The maximum identified by the reference number 49 in FIG. 4d occurs at ν.

If T _{Q is} greater than T _Q , the maximum value of the algebraic sum of the contained data occurs at a speed greater than v. This is illustrated in Figures 4e and 4f.

509810/031 2

- 16 -

2 4 / f O 31 2

Each algebraic summation is stored in a "first" table in the T _o speed function number level.
If this summation for all values of the vertical transit time
Tq is complete, _{a search algorithm is applied to each T Q} to determine the maximum absolute value of any measurement of the amplitude coherence contained therein. This value is recorded in a second table and its sign is generated again, if all 17ths of the vertical run-of-time of the first table have been searched in the manner described above, the second table is examined to find out what it contains To determine maxima and minima. A maximum occurs when a value in the second table is more positive than the values immediately before it. lie or immediately follow it, and a minimum is present if a value of the
second table is more negative than the values immediately
are in front of him or immediately follow him. All values in the
second table, which are not maxima or minima, are then set equal to zero. The points of the maxima and minima are retained and are used _{to generate T Q} values and signs for a set of third tables, which are taken from the first table.

Pig. FIG. 5a shows the representation of these numerical measured values of the amplitude coherence in their correct coordinates in FIG
vertical time-of-arrival-speed-plane. Fig. 5a is generated by examining the second table until the
first maxima or minima meet. If this is found

50981 0/0312

- 17 -

the values _{corresponding to this T 0} are entered in column 3 of FIG. 5a. Columns 1 and 2 and columns 4 and 5 contain measurements of the polarity amplitude for the two values of the vertical time of incidence immediately before and immediately after the value of the vertical time of incidence for which there was a maximum or minimum, respectively. This process is repeated for each maximum or minimum of the second table, producing a number of Mg. 5a for each seismic wave train set 10. In a typical operation, the number of vertical times of incidence in a wave train set would be about 500 and about 100 maxima or minima would be found using the method described above.

Typical numerical values shown in FIG. 5a resulted from an analysis carried out with a single, artificial, seismic reflection event similar to that in FIG. 3a. The T _{0 of} this event was equal to 1.816 seconds and its ν ν; ar was equal to 2575 m / sec (8450 feet per second). To simplify this representation, values of the amplitude coherence measurements were not shown that were smaller than 35 percent of the maximum coherence 53

Fig. 5b shows several amplitude-coherent contours 51 for this seismic reflection event in the vertical time-of-arrival velocity plane It can be seen that these contours form approximately concentric ellipses, the

509810/031 2
- 18 -

common main aciise is inclined away from the numerical axis of the velocity function.

Investigations carried out within the scope of the invention of the directed amplitude coherence surfaces such as those in the -Pig. 5a and 5b led to the development of an a priori knowledge of such areas which are representative of actual seismic reflection events. This Arpriori knowledge has been used to develop various validity tests for amplitude coherence surfaces generated by seismic reflectors in actual seismic field data. Figure 5c is useful in explaining these tests. The maximum measured value of coherence that has been calculated for this artificial seismic reflection event is indicated by the reference number 53. In order for them to be accepted as characteristic of a valid event from the start, three of the four values of the amplitude coherence which immediately surround the value 53 and which _{have been calculated for the same value T Q must} exceed 50 percent of the maximum value 53. In addition, a maximum amplitude coherence value between zero and four speed steps above the speed value which corresponds to the maximum value 53 must appear in the column of the amplitude coherence values, which is identified by the reference number 54. Its amplitude coherence must be less than the value 53> but must exceed 50 percent of the maximum value 53. Finally, the column of the amplitude coherence values identified by the reference number 52 must

509810/0312

- 19 -

244Q312 .οίο.

also contain a maximum that is smaller than the maximum value 53? but has at least 50 percent of this. This maximum in column 52 must occur between zero and four speed steps below the speed value which corresponds to the maximum 53. Each column corresponding to each "determined T ₀ for which there is a maximum or minimum is subjected to the structural tests described above, and those who pass these structural tests are tentatively cataloged as valid seismic reflection events. Each of the amplitude coherence surfaces can then be adjusted to the mean squares on a two-dimensional square surface or any other two-dimensional interpolation scheme that takes into account the inclination of the major axis of the ellipses, using the three columns indicated by 52, 53 and 54 and possibly an additional column, if necessary, to create the pattern If the mean least squares method is used, it is necessary that the quadratic fit function be an elliptical paraboloid, and a description of such surfaces appears in "An Elementary Treatise on Solid Geometry" by C. Smith at Macmillan and Co., ltd., 1940, pages 52-53, and details of the general weighted least squares fit method are given in "Mathematical Analysis of Observation", BM Shchigolev, London Life Books, Ltd., 1965, pages 247-250, described.

509810/0312

- 20 -

Pig. 5d shows the remaining error values deviating from Hull, as they result from an adaptation according to the mean least squares method to the suitable amplitude coherence data of FIG. 5b. The mean square deviation (RMS value) of these remaining values is only about 2 percent of the maximum value 53, and the EMS values that are less than 5 percent are considered normal. If this RMS value exceeds 5 percent, it can be assumed from the velocity maximum in that it has a lower quality. The coordinates of the maximum of the area obtained by the fit with the least squares method allow a good determination of the T _Q and the velocity of each seismic reflection.

If one follows the logic and the interpolation for the automatic singling out of an event, as it is in the discussion of the 5, a fourth table, consisting of columns and rows, then becomes in the depth-point vertical time-of-arrival plane educated. The columns of the table represent depth points for which a seismic wave train set 10 is obtained has been, and the entries in the columns of the table are either zero or amplitude, speed and timing measurements of those events that passed the tests enumerated in the discussion of FIG. 5 above. The place each entry in the column corresponds to the value of the index Tq of the vertical arrival time in the seismic wave train set 10 for which it was received, i.e. if column

509810/0312

- 21 -

- dead 3 in Pig. 5o was the 55th T _Q in seismic wave train set 10 and if event 53 passed the criteria enumerated above, the amplitude, velocity, and time calculations for event 53 would be entered in the 55th row of the column corresponding to its depth point is equivalent to.

An examination of this table is then initiated to identify valid seismic reflection horizons, where In a preferred embodiment of the invention, six or more points are analyzed in each examination process. In other words, in the first inspection process the depth points (columns) 1 to 6 of the table are examined, in the second process the reflection points (columns) 7 to 12 etc.

The procedure begins with an investigation row by row of the six reflection points until a non-zero value in one of the columns is found. Such a value is represented by the point number 60 in Fig. 6, and point 60 then becomes a node for the tree search process. A wedge 61 is formed by the node 60, and this Wedge extends both to the right and to the left from the node 60. The boundaries 62 of the wedge 61 have Slopes with a specific number of milliseconds per depth point. An investigation will then begin in the area which is formed by the wedge 61 to determine whether other Points are not equal to Hull and have the same sign.

509810/0312

- 22 -

If the search method (right «or left) has more than two consecutive columns are found in which no Values not equal to Hull with the same sign as the node value • are then the ongoing investigation for this node 60 canceled. If a point with opposite Sign is found, the, time code, at which the point with the opposite gate sign has been found, a boundary point for the investigation. If, on the other hand, a point 64 is found within the wedge 61 which is not equal to Hull and has the same sign, the investigation is continued in the same sighting (left or right) using a new wedge, the is formed at point 64.

If there are at least four such points within the formed Wedges have been found that will measure the amplitude The corresponding speeds are recovered, and if so these speeds are all within +1.5 speed function ion numbers are of each other, the weighted mean of these speeds is then calculated. This mean speed is stored along with the corresponding exact time and amplitude of the event.

Fig. 7 shows part of a representation of the result of this method. The abscissa of the representation is the distance along the seismic profile, country, the ordinate is the vertical Einfallslauf.zeit, Each "Plus" Tand "minus ¹¹ ISIN

509810/0312

- 23 -

draws the sign of a seismic effect ion event, that using the above in explaining the Pig. 5c "are identified and stored and which has been screened out by the pole application process. The larger the "plus" or "minus" sign, the greater the amplitude of the event. The decimal numbers with four (or five) digits represent the calculated mean speed values for the next 15.2 m / sec (50 feet per second) over six adjacent depth points and appear at intervals of six depth points. The star-shaped sign below each four-digit number indicates the coordinates of the maximum speed and the size of this symbol indicates the relative quality of the speed determination. This quality can, for example, be inversely proportional to the BMS value of residual error, as in Pig. 5d is shown.

Apparatus for carrying out the method according to the invention is readily available and its selection is for the Skilled in the art obviously. In order to carry out the method of the invention, it is probably best to use a universal use a digital computer and an automated display device. Methods according to the invention are in several embodiments with IBM Computer Model Nos. 360-44 and 360-65, and is a cathode ray tube has been used to enable the results to be presented.

509810/0312

- 24 -

Claims (8)

Claims 1. Process for earthquake monitoring, characterized by the following process steps; - (a) Definition of a number of wave trains with a set of seismic wave trains with a common reflection point, each V / ellenzugsatz as a puncture increasing vertical inflow is arranged and the wave trains from a number of different reflection points be obtained (b) Formation of a number of number curves of a speed function in the velocity-vertical time-of-arrival space, (c) Formation of a number of narrow time windows at the first desired vertical arrival time within one first wave train theorem, with each time window having a one-to-one relationship (one-to-one correspondence) has to a predetermined number of the speed function, (d) Sampling the first wave train set at a first vertical one Incidence time with the number of windows formed there, 509810/0 312 - 25 - (e) algebraic summation of the amplitude of all wave trains at points within each window when sampling the first wave train set at the first vertical time of arrival, (f) Arrangement of the algebraic sum of the first vertical time of arrival as a function of the number of velocities f unction, (g) repeating steps (c) through (f) for all desired vertical arrival times in the first wave train set, (h) Analysis of the vertical time-of-arrival number of the velocity function space formed by steps (c) through (g) to determine the sign, the amplitude and the coherence of the sums, (i) Identify a first pattern of polarity, amplitude, and coherence in the vertical time-of-arrival number the velocity function space and initial assumption if it passes predetermined test criteria, (j) Perform two-dimensional interpolation in the first Pattern, if initially accepted, to determine the maximum value of the pattern in terms of its amplitude, to determine the number of the speed function and the time, 509810/0312 - 26 - (k) repeating steps (i) and (j) for all patterns, which have been found in the vertical time-of-arrival number of the velocity function space, (1) Arrangement of the found amplitude, the number of the speed function and the vertical arrival time of the maximum values in the vertical arrival time depths p unkt space, (m) Repeat steps (a) ″ to (1) for each wave set from every depth point, (n) Investigation of the vertical time-of-arrival-depth point space according to a number of points that have a temporal and spatial consistency, the same sign and speeds have, which are substantially the same, and (o) Generation of a representation in seismic recording coordinates, with the respective Yqrzezug and the averaged Speed of the points found in step (n). 2. The method of claim 1, wherein step (c) of forming a number of narrow time slots "each vertical Incidence time is characterized by 509810/0312 - 27 - (a) Formation of support points of a curve of a speed control function in your speed vertical insert all-time-Eaum due to the knowledge of the investigation area, (Td) Connection of the support points by straight sections, (c) Calculation of the transit time to the most distant wave train of the wave train theorem at each interpolation point, (d) Generation of the interpolation points of a group of the number curves of the speed function, the interpolation point The running time assigned to each curve is a multiple of a selected time interval with respect to the corresponding one Is the running time that was calculated in step (c) for support points of the curve of the speed control function, (e) Connecting the support points established in step (d) by straight-line sections, with a group of numerical curves a speed function in the speed vertical Time-of-arrival space is formed, (f) linear interpolation of a speed from a Numerical curve of a first speed function, the speed at a first desired vertical Time of arrival in the wave train theorem of this curve assigned, 509810/0312 - 28 - (g) Calculation of the vertical arrival time for each wave train the 7 / ellenzzugsatzes for the number curve of the first Speed function, (h) Formation of the width of the time window by adding and subtracting equal, short time segments to the for vertical arrival times calculated for each wave train, and (i) Repeat steps (f), (g) and (h) for all number curves the speed function. 3. The method according to claim 1 or 2, characterized in that the width of the time window formed is about half the width of a waveform of typical seismic data. 4. The method according to any one of claims 1 to 3, 'wherein step (h) of claim 1 for analyzing the vertical Incidence time number of the velocity function space g e character e i chne t is through (a) Finding the maximum absolute value of the algebraic sum over all numbers of the velocity function for the first vertical arrival time, (b) storing this maximum value in a table and retrieving its sign, 509810/0312 - 29 - (c) Perform steps (a) and (b) for all verticals Times of incidence in the wave train theorem, (d) Analyze the table of the steps (a), (b) and (c) stored signed values, the maximum points being formed when a sum of steps (a) and (b) is more positive than the sums immediately preceding and immediately following it, and wherein minimum points are formed when a sum of steps (a) and (b) is more negative than that sums immediately preceding or immediately following it, and (e) Replace the sum values in the table with the value zero, which are not maximum or minimum points. 5. The method according to any one of claims 1 to 4, wherein the identification of the first pattern of the sign, the amplitude and the coherence in step (i) of claim 1 is characterized by (a) Selection for analysis of the index of the first vertical time of incidence for which there is a maximum or minimum, and four vertical arrival times, two being the vertical arrival time for which a maximum or Minimum is present, immediately precede it and two immediately follow it, 509810/0312 - 30 - (b) Perform pattern checks of the algebraic Sums that are in the vertical time-of-flight number of the Speed function space for the ones selected in step (a) vertical times of incidence exist, and (c) Classifying the maximum or minimum as a valid point to proceed with, if any Pattern has been found that matches predetermined criteria. 6. The method according to claim 5, characterized in that the predetermined criteria are found when the following conditions are met: (a) The polarity and amplitude of the measurements for three of the four numbers of the velocity function that make up the maximum or minimum point at the same vertical time of incidence, each exceed 50 percent of the maximum or minimum, (b) the amplitude measurement of the maximum for the vertical time of incidence, which immediately follows the vertical arrival time for which there was a maximum or minimum has, lies between zero and four numbers of the speed function higher than the number of the speed function for which the maximum or minimum exists and is at least 50 percent of the maximum or minimum, 509810/0312 - 31 - but less than the maximum or minimum, and (c) the amplitude measurement of the maximum for the vertical time of arrival, that of the vertical time of arrival immediately for which there is a maximum or minimum is between zero and four numbers of the speed function below the number of the G-speed function for which there is a minimum or maximum, and is smaller, but at least 50 percent, than the maximum or minimum. 7. The method according to any one of claims 1 to 6, wherein the investigation of the vertical time-of-arrival depth point space in step (n) of claim 1 is characterized by (a) Selecting the first set of more than five reflection points in the vertical time-of-arrival depth point space, (b) Investigation in the direction of the vertical time of incidence of the room until a value not equal to zero is found, (c) Formation of an opening to the right and to the left of this value not equal to NuIl, the opening being a wedge equals, the origin of which is not equal to zero at the value, . and wherein the opening has limit curves, the slopes of which are predetermined 509810/0312 - 32 - (d) Look to the right and left in the opening to another non-zero point with the same sign within to find two depth points from the originally found point, (e) Repeat steps (c) and (d) for each point found that is not equal to zero and has the same sign Has, (f) Classification of the points as indicative of a lower seismic horizon if more than three are found whose corresponding numbers of the speed function within a predetermined Limit match, (g) Calculation of the mean speed of the points found in step (f), (h) repeating step (b) until a new starting point has been found or until all vertical arrival times have been exhausted, and (i) Carrying out steps (a) to (h) for all groups of reflection points that have the same number as those in Step (a) selected number. 8. Seismic record made according to the procedure according to one of claims 1 to 7 509810/0312 - 33 -