Method for estimating geothermal gradient and device for estimating geothermal gradient

• US 10,641,662 B2
• Filed: 08/24/2016
• Issued: 05/05/2020
• Est. Priority Date: 12/24/2015
• Status: Active Grant

• Current Assignee: Korea Electric Power Corporation (Government of South Korea)
• Original Assignee: Korea Electric Power Corporation (Government of South Korea)
• Inventors: Yun, Kwan Hee, Shim, Jung Tak, Jung, Eun Ju
• Primary Examiner(s): Patel, Paresh
• Application Number:
  US16/065,298
• Publication Number:
  US 20190078944A1
• Time in Patent Office: 1,350 Days
• Field of Search: None
• CPC Class Codes: E21B 47/07 Temperature
• CPC Class Codes: G01K 13/10 for measuring temperature within piled or stacked materials by special arrangements for conducting heat from the object to the sensitive heat element G01K1/16
• CPC Class Codes: G01V 1/00 Seismology; Seismic or acoustic prospecting or detecting
• CPC Class Codes: G01V 1/28 Processing seismic data, e.g. for interpretation or for event detection G01V1/48 takes precedence
• CPC Class Codes: G01V 1/30 Analysis G01V1/50 takes precedence
• CPC Class Codes: G01V 9/005 by thermal methods, e.g. after generation of heat by chemical reactions
• CPC Class Codes: G06F 17/18 for evaluating statistical data , e.g. average values, frequency distributions, probability functions, regression analysis forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: G06N 3/08 Learning methods

Abstract

The present invention relates to a method for estimating a geothermal gradient and an apparatus for estimating a geothermal gradient, and provides a method for estimating a geothermal gradient, the method comprising the steps of: calculating, at each of a plurality of points, a seismic wave crustal attenuation characteristic parameter of a seismic wave measured on the basis of the measured seismic wave information, with respect to the plurality of points; deriving an artificial neural network relational formula on the basis of the seismic wave crustal attenuation characteristic parameter; and calculating a geothermal gradient of an area of interest excluding the plurality of points on the basis of the artificial neural network relational formula, such that the geothermal gradient is estimated at a low cost by using a highly reliable method.

Claims

1. A method for estimating a geothermal gradient, the method comprising calculating a seismic wave crustal attenuation characteristic parameter of a measured seismic wave at each of a plurality of points on the basis of seismic wave information measured for the plurality of points;

deriving an artificial neural network relational formula on the basis of the seismic wave crustal attenuation characteristic parameter; and

calculating a geothermal gradient of an area of interest except for the positions on the basis of the artificial neural network relational formula.

2. The method of claim 1, wherein the calculation of a seismic wave crustal attenuation characteristic parameter calculates Q₀ and a frequency dependence value η

at a geometrical location and an in-crust depth of an area where the seismic wave information was measured, through a spectrum and a Q tomography inversion method based on the measured seismic wave information when anelastic attenuation of a seismic wave is modeled into Q(f)=Q₀f^η

.

3. The method of claim 2, wherein a dependency difference η

_diff (η

_diff=η

−

η

′

) that is a difference between an estimation value η

′

of the frequency dependency value calculated from a function F_η

(Q₀) between the Q₀ and frequency dependency value η

calculated by the Q tomography inversion method and the frequency independency value η

calculated by the Q tomography inversion method is additionally calculated in the calculation of a seismic wave crustal attenuation characteristic parameter.

4. The method of claim 1, further comprising calculating at least any one of a topographic characteristic parameter and a crustal heterogeneity parameter about the area of interest as the earth science information, and

the deriving of an artificial neural network relational formula derives the artificial neural network relational formula by further including at least any one of the calculated topographic characteristic parameter or crustal heterogeneity parameter.

5. The method of claim 4, wherein the topographic characteristic parameter includes at least any one of a height, a relative height, a slope, and roughness table for the earth surface of the area of interest.

6. The method of claim 4, wherein the crustal heterogeneity parameter includes at least any one of a mantle depth and a crustal property of the area of interest.

7. The method of claim 4, wherein the deriving of au artificial neural network relational formula drives the artificial neural network relational formula by using the seismic wave crustal attenuation characteristic parameter, the topographic characteristic parameter, and the crustal heterogeneity parameter about the plurality of points as input of the artificial neural network relational formula and by using geothermal gradients measured for the plurality of points as output, and

the calculation of a geothermal gradient calculates a geothermal gradient of the area of interest by calculating a seismic wave crustal attenuation characteristic parameter, a topographic characteristic parameter, and a crustal heterogeneity parameter at the area of interest and substituting the parameters into the derived artificial neural network relational formula.

8. A method for estimating a geothermal gradient, the method comprising:

configuring an artificial neural network relational formula by using a seismic wave crustal attenuation characteristic parameter, a topographic characteristic parameter, and a crustal heterogeneity parameter calculated about a plurality of points as input of the artificial neural network and by using geothermal gradients measured for the plurality of points as output of the artificial neural network;

calculating artificial neural network relational formulae, in which a relationship between the measured geothermal gradients and a geothermal gradient value calculated on the basis of an artificial neural network satisfies a predetermined condition, as much as a predetermined number;

calculating a plurality of geothermal gradient estimation values by inputting a seismic wave crustal attenuation characteristic parameter, a topographic characteristic parameter, and a crustal heterogeneity parameter at an area of interest except for the plurality of joints, into the calculated artificial neural network relational formulae; and

calculating a final geothermal gradient estimation value by calculating a statistic value for the plurality of geothermal gradient estimation values.

9. The method of claim 8, further comprising calculating a linear regression formula between the measured geothermal gradient values and geothermal gradient values calculated by inputting the parameters at the plurality of points into the calculated artificial neural network relational formulae.

10. The method of claim 8, wherein the predetermined condition means that an error between the measured geothermal gradient and a geothermal gradient calculated on the basis of an artificial neural network is a first reference value or less.

11. The method of claim 9, wherein the predetermined condition means that a correlation coefficient for a geothermal gradient value calculated by an artificial neural network relational formula having a minimum error of artificial neural network relational formulae satisfying the condition that the error is the first reference value or less.

12. The method of claim 8, wherein the predetermined condition means that a correlation coefficient for a geothermal gradient value calculated by an artificial neural network relational formula having a minimum error of a plurality of artificial neural network relational formulae calculated by artificial neural network repetitive learning is a second reference value or less.

13. A device for estimating a geothermal gradient, the device comprising:

a parameter extraction unit extracting a parameter showing a characteristic of input information;

an artificial neural network relational formula calculation unit calculating an artificial neural network relational formula having the parameter calculated by the parameter extraction unit as input; and

a geothermal gradient estimation unit calculating a geothermal gradient estimation value by inputting parameters at an area of interest extracted by the parameter extraction unit into an artificial neural network relational formula calculated by the artificial neural network relational formula calculation unit,

wherein the parameter extracted by the parameter extraction unit includes a seismic wave crustal attenuation characteristic parameter, a topographic characteristic parameter, and a crustal heterogeneity parameter, and

the artificial neural network relational formula calculation unit calculates the artificial neural network relational formula on the basis of parameters extracted at a plurality of points by the parameter extraction unit and geothermal gradients measured at the plurality of points.