Sinkage prediction device and sinkage prediction method

A sinkage prediction device comprising:

• a variation data acquisition unit analyzing time-series observation data measured by using an earth observation technology and acquiring horizontal direction variation data which is data indicating a horizontal direction variation on a ground surface;
• a sinkage prediction unit determining, based on the horizontal direction variation data acquired by the variation data acquisition unit, whether a variation in a time axis or a variation in a space axis on the ground surface is in a predetermined state and predicting a location on the ground surface indicating the variation determined to be in the predetermined state as a potential sinkage location where there is a possibility of occurrence of sinkage,
• wherein the variation data acquisition unit includes:
• an LoS variation data acquisition unit analyzing time-series observation data measured by using the earth observation technology and acquiring north-bound LoS variation data which is data indicating the variation in a line-of-sight direction when the ground surface is viewed from a satellite in which a satellite traveling direction is in a north-bound trajectory and south-bound LoS variation data which is data indicating the variation in the line-of-sight direction when the ground surface is viewed from a satellite in which a satellite traveling direction is in a south-bound trajectory;
• an LoS variation data detection unit performing matching between the north-bound LoS variation data and the south-bound LoS variation data acquired by the LoS variation data acquisition unit and detecting matched variation data for each predetermined unit time and for each predetermined unit area;
• and
• a variation direction decomposition unit decomposing the north-bound LoS variation data and the south-bound LoS variation data detected by the LoS variation data detection unit as matched data for each predetermined unit time and for each predetermined unit area into horizontal direction variation data which is data indicating horizontal direction variation and vertical direction variation data which is data indicating vertical direction variation, and
• wherein the sinkage prediction unit determines, based on the north-bound LoS variation data, the south-bound LoS variation data, the horizontal direction variation data, and the vertical direction variation data, whether a variation in a time axis or a variation in a space axis on the ground surface is in the predetermined state for each of the north-bound LoS direction, the south-bound LoS direction, the horizontal direction, and the vertical direction and predicts the location on the ground surface indicating the variation determined to be in the predetermined state for at least one direction of the horizontal direction, the north-bound LoS direction, the south-bound LoS direction, and the vertical direction as the potential sinkage location.

A sink prediction method, comprising:

• a first step of a variation data acquisition unit of a sinkage prediction device analyzing time-series observation data measured by using an earth observation technology and acquiring horizontal direction variation data which is data indicating a horizontal direction variation on a ground surface;
• and
• a second step of a sinkage prediction unit of the sinkage prediction device determining, based on the horizontal direction variation data acquired by the variation data acquisition unit, whether a variation in a space axis on the ground surface is in a predetermined state and predicting a location on the ground surface indicating the variation determined to be in the predetermined state as a potential sinkage location where there is a possibility of occurrence of sinkage;
• wherein the variation data acquisition unit acquires a variation direction and an amount of variation on the ground surface as the horizontal direction variation data for each predetermined unit time and for each predetermined unit area in the first step;
• wherein the processes of the sinkage prediction unit in the second step includes:
• a singular point extraction step determining whether, for each of the predetermined unit areas, the variation trend on the ground surface between a most recent final time point and a one time point in the past and the variation trend on the ground surface in the past before the one time point in the past are in a state of being different and extracting, as a singular point, the predetermined unit area determined to be in the state in which the variation trends that are specified before and after the one time point in the past as a boundary are different;
• a grouping area specification step determining, as the predetermined state, whether it is in a state in which the predetermined unit area where the variation directions face each other at intervals within a first predetermined distance in the same predetermined unit time within the predetermined search target range from the singular point extracted in the singular point extraction step exists and specifying the grouping area in a predetermined range including the facing predetermined unit area satisfying the state;
• and
• a prediction step predicting the grouping area specified in the grouping area specification step as the potential sinkage location.