Earthquake estimation method, non-transitory computer readable medium, and earthquake estimation device

An earthquake estimation method comprising, by computer:

• generating an observation image showing a spatial distribution of seismic wave propagation based on an observation result of seismic waves at a plurality of observation points on a ground;
• and
• estimating a number of earthquakes with respect to the observation image by using an earthquake estimation model in which a parameter of an earthquake including at least a position of a hypocenter and a magnitude is associated with a simulated observation image showing a spatial distribution of seismic wave propagation on the ground obtained from a result of a numerical simulation of one or more of the earthquake, performed with the parameter, the earthquake estimation model performing machine learning by using, as teacher data, a set of the simulated observation image and the number of the earthquakes used, when generating the simulated observation image, to estimate the number of earthquakes.

A non-transitory computer readable medium storing an earthquake estimation program causing a computer to:

• generate an observation image showing a spatial distribution of seismic wave propagation based on an observation result of seismic waves at a plurality of observation points on a ground;
• and
• estimate a parameter of an earthquake with respect to the observation image by using an earthquake estimation model in which a parameter of an earthquake including at least a position of a hypocenter and a magnitude is associated with a simulated observation image showing a spatial distribution of seismic wave propagation on the ground obtained from a result of a numerical simulation of the earthquake, performed with the parameter.

A non-transitory computer readable medium storing an earthquake estimation program causing a computer to:

• generate an observation image showing a spatial distribution of seismic wave propagation based on an observation result of seismic waves at a plurality of observation points on a ground;
• and
• estimate a number of earthquakes with respect to the observation image by using an earthquake estimation model in which a parameter of an earthquake including at least a position of a hypocenter and a magnitude is associated with a simulated observation image showing a spatial distribution of seismic wave propagation on the ground obtained from a result of a numerical simulation of one or more of the earthquake, performed with the parameter, the earthquake estimation model performing machine learning by using, as teacher data, a set of the simulated observation image and the number of the earthquakes used, when generating the simulated observation image, to estimate the number of earthquakes.

An earthquake estimation device comprising:

• a data acquisition unit that generates an observation image showing a spatial distribution of seismic wave propagation based on an observation result of seismic waves at a plurality of observation points on a ground;
• and
• an earthquake estimation unit that estimates a parameter of an earthquake with respect to the observation image by using an earthquake estimation model in which the parameter of the earthquake including at least a position of a hypocenter and a magnitude is associated with a simulated observation image showing a spatial distribution of seismic wave propagation on the ground obtained from a result of a numerical simulation of the earthquake, performed with the parameter.

An earthquake estimation device comprising:

• a data acquisition unit that generates an observation image showing a spatial distribution of seismic wave propagation based on an observation result of seismic waves at a plurality of observation points on a ground;
• and
• an earthquake estimation unit that estimates a number of earthquakes with respect to the observation image by using an earthquake estimation model in which a parameter of an earthquake including at least a position of a hypocenter and a magnitude is associated with a simulated observation image showing a spatial distribution of seismic wave propagation on the ground obtained from a result of a numerical simulation of one or more of the earthquake, performed with the parameter, the earthquake estimation model performing machine learning by using, as teacher data, a set of the simulated observation image and the number of the earthquakes used, when generating the simulated observation image, to estimate the number of earthquakes.