Method and equipment for weather image prediction

• US 5,974,360 A
• Filed: 12/12/1997
• Issued: 10/26/1999
• Est. Priority Date: 12/13/1996
• Status: Expired due to Fees

• Current Assignee: Nippon Telegraph and Telephone Corporation
• Original Assignee: Nippon Telegraph and Telephone Corporation
• Inventors: Otsuka, Kazuhiro, Suzuki, Satoshi, Horikoshi, Tsutomu, Sakaino, Hidetomo
• Primary Examiner(s): McElheny, Jr., Donald E.
• Application Number:
  US08/989,655
• Time in Patent Office: 683 Days
• Field of Search: 702/3, 702/4, 706/931, 342/26
• US Class Current: 702/3
• CPC Class Codes: G01S 13/951 ground based
• CPC Class Codes: Y02A 90/10 Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Abstract

An equipment for weather image prediction comprising: a field extracting portion, which calculates the velocity fields of the local echoes with respect to time and space from a plurality of time sequential frames of weather radar images for a plurality of frames therein, and which segments and extracts a field possessing similar velocities with respect to time and space as one precipitation field from said plurality of frames of weather radar images by employing said calculated velocity fields; a moving velocity estimating portion, which estimates the moving velocity of a precipitation field by obtaining the amount of movement, among the neighboring frames, of the precipitation field segmented and extracted by said field extracting portion; and a forecast image generating portion, which extrapolates a plurality of precipitation fields by employing a distance taken by multiplying the moving velocity of the precipitation field estimated by said moving velocity estimating portion by a desired forecasting time, and which generates a forecast radar image by synthesizing the resulting image.

Claims

1. An equipment for weather image prediction comprising:

a field extracting portion, which calculates the velocity fields of local echoes with respect to time and space from a plurality of time sequential frames of weather radar images for a plurality of frames therein, and which segments and extracts a field possessing similar velocities with respect to time and space as one precipitation field from said plurality of frames of weather radar images by employing said calculated velocity fields;

a moving velocity estimating portion, which estimates the moving velocity of a precipitation field by obtaining the amount of movement, among the neighboring frames, of the precipitation field segmented and extracted by said field extracting portion; and

a forecast image generating portion, which extrapolates a plurality of precipitation fields by employing a distance taken by multiplying the moving velocity of the precipitation field estimated by said moving velocity estimating portion by a desired forecasting time, and which generates a forecast radar image by synthesizing the resulting image.

2. An equipment for weather image prediction in accordance with claim 1 wherein said field extracting portion calculates said velocity fields of the echoes from the location where the cross-correlation coefficient of the partial image gray level of two weather radar images is greatest.

3. An equipment for weather image prediction in accordance with claim 1 wherein said field extracting portion obtains said precipitation field by way of comparing the respective velocity vectors of the adjacent blocks of said precipitation field with respect to space and time, and by way of segmenting the weather radar images so as to have fields possessing similar velocities belong to the same precipitation field.

4. An equipment for weather image prediction in accordance with claim 1 wherein said moving velocity estimating portion calculates the average value of the amount of displacement of the centroid of the corresponding precipitation field between frames, and estimates the moving velocity of that precipitation field by employing said average value.

5. An equipment for weather image prediction in accordance with claim 1 wherein said moving velocity estimating portion further comprises:

a precipitation field velocity candidate computing portion which obtains the velocity candidate of the precipitation field segmented and extracted by said field extracting portion, by employing the weather radar images from either two proximate frames or a plurality of frames;

a precipitation field characteristics determining portion which determines the characteristics of the precipitation field segmented and extracted by said field extracting portion; and

a precipitation field moving velocity estimating portion which estimates the moving velocity of the precipitation field by taking into consideration the characteristics of the precipitation field obtained by said precipitation field characteristics determining portion, and with respect to the velocity candidate of the precipitation field obtained by said precipitation field velocity candidate computing portion.

6. An equipment for weather image prediction in accordance with claim 5 wherein said precipitation field characteristics determining portion has said characteristics of the precipitation field be given by the degree of mobility or immobility, which is the characteristics of the movement of the precipitation field, which in turn, is the echo cells or the group of echo cells within the weather radar image.

7. An equipment for weather image prediction in accordance with claim 6 wherein said precipitation field characteristics determining portion determines the precipitation field to be mobile in the case where the velocity obtained by said precipitation field velocity candidate computing portion is equal to or exceeds a fixed value, and determines the precipitation field to be immobile in the case where said velocity does not reach the fixed value.

8. An equipment for weather image prediction in accordance with claim 6 wherein said precipitation field characteristics determining portion determines said degree of mobility or immobility by the similarities between the local velocities of the echoes within the weather radar image obtained by said field extracting portion and the global velocity of the precipitation field obtained by said precipitation field velocity candidate computing portion.

9. An equipment for weather image prediction in accordance with claim 8 wherein said precipitation field characteristics determining portion determines said similarity as a function of the angle formed by the velocity vector expressing said local velocities of the echoes and by the velocity vector expressing said global velocity of the precipitation field.

10. An equipment for weather image prediction in accordance with claim 8 wherein said precipitation field characteristics determining portion determines said similarity as a function of the magnitude of the difference of the velocity vector expressing said local velocities of the echoes and the velocity vector expressing said global velocity of the precipitation field.

11. An equipment for weather image prediction in accordance with claim 6 wherein said precipitation field moving velocity estimating portion estimates said moving velocity of the precipitation field as the velocity obtained by said precipitation field velocity candidate computing portion in the case where the precipitation field is determined to be "mobile" by said precipitation field characteristics determining portion, and estimates said moving velocity of the precipitation field as "0" in the case where the precipitation field is determined to be "immobile" by said precipitation field characteristics determining portion.

12. An equipment for weather image prediction in accordance with claim 6 wherein said precipitation field moving velocity estimating portion estimates the moving velocity of the precipitation field which is to be employed in the forecast by way of normalizing the velocity calculated by said precipitation field velocity candidate computing portion by the degree of mobility or immobility obtained by said precipitation field characteristics determining portion.

13. An equipment for weather image prediction in accordance with claim 1 wherein said moving velocity estimating portion records the moving velocity of the precipitation field segmented and extracted by said field extracting portion for a time in the past, and estimates the current moving velocity of the precipitation field by taking into consideration said moving velocity of the precipitation field for a time in the past.

14. An equipment for weather image prediction in accordance with claim 13 wherein said moving velocity estimating portion further comprises:

a precipitation field velocity candidate computing portion which obtains the velocity candidate of the precipitation field segmented and extracted by said field extracting portion, by employing the weather radar images from either two proximate frames or a plurality of frames;

a precipitation field velocity field recording portion which records the spatial distribution of the moving velocity of the precipitation field calculated by said precipitation field velocity candidate computing portion for a plurality of times;

a precipitation field moving velocity estimating portion which estimates the most current moving velocity of an arbitrary precipitation field by employing the spatial distribution and the temporal distribution of the precipitation fields from the present to an arbitrary time in the past recorded by said precipitation field velocity field recording portion.

15. An equipment for weather image prediction in accordance with claim 14 wherein said moving velocity estimating portion estimates said current moving velocity of the precipitation field by using the spatial distribution and the temporal distribution from the present to an arbitrary time in the past from the moving velocity of the precipitation field recorded by said precipitation field velocity field recording portion and by employing the value obtained by normalizing the moving velocity of the precipitation field of the time from the present to an arbitrary time in the past according to those times for the inside of the spatial area corresponding to the current precipitation field.

16. An equipment for weather image prediction in accordance with claim 5 wherein:

said precipitation field moving velocity estimating portion further comprises a precipitation field velocity field recording portion which records the spatial distribution of estimated moving velocity of the precipitation field obtained by said precipitation field moving velocity estimating portion for a plurality of times; and

said precipitation field moving velocity estimating portion estimates the moving velocity of the precipitation field by employing the velocity candidate of the precipitation field obtained by said precipitation field velocity candidate computing portion based on the characteristics of the precipitation field obtained by said precipitation field characteristics determining portion, and then performs estimating amendments of the moving velocity of an arbitrary precipitation field of the most current time by employing the spatial distribution and the temporal distribution of the precipitation field moving velocity from the present to an arbitrary time in the past recorded by said precipitation field velocity field recording portion.

17. An equipment for weather image prediction in accordance with claim 16 wherein said precipitation field moving velocity estimating portion performs said estimating amendments of the moving velocity of the precipitation field by employing the spatial distribution and the temporal distribution from the present to an arbitrary time in the past from the moving velocity of the precipitation field recorded by said precipitation field velocity field recording portion and by employing the value obtained by normalizing the moving velocity of the precipitation field of the time from the present to an arbitrary time in the past according to those times for the inside of the spatial field corresponding to the current precipitation field.

18. An equipment for weather image prediction in accordance with claim 1 wherein said moving velocity estimating portion further comprises:

a precipitation field velocity candidate computing portion which obtains the velocity candidate of the precipitation field segmented and extracted by said field extracting portion, by employing the weather radar images from either two proximate frames or a plurality of frames; and

a precipitation field moving velocity estimating portion which estimates the moving velocity of the precipitation field by employing either one of the local velocities of the echoes within the weather radar image obtained by said field extracting portion, or the global velocity of the precipitation field obtained by said precipitation field velocity candidate computing portion, depending on the forecasting time requested at said forecast image generating portion.

19. An equipment for weather image prediction in accordance with claim 1 wherein said precipitation field moving velocity estimating portion estimates the moving velocity of the precipitation field by employing the local velocities of the echoes within the weather radar images obtained by said field extracting portion in the case where the forecasting time requested at said forecast image generating portion is equal to or less than a predetermined value, and estimates the moving velocity of the precipitation field by employing the global velocity of the precipitation field obtained by said precipitation field velocity candidate computing portion in any other case.

20. An equipment for weather image prediction in accordance with claim 1 wherein said forecast image generating portion uses the weather radar images contained within each of the individual precipitation fields segmented and extracted by said field extracting portion as a forecast base image which is to be the base image for the forecast, and extrapolates a plurality of precipitation fields by way of performing parallel translation of each of the forecast base images by a distance obtained by multiplying said moving velocity with said requested forecasting time.

21. An equipment for weather image prediction in accordance with claim 20 wherein said forecast image generating portion selects the greatest value of the gray level of said forecast base image from the overlapped plurality of precipitation fields resulting from said parallel translation for each of said plurality of precipitation fields.

22. An method for weather image prediction comprising:

a field extracting step, which calculates the velocity fields of local echoes with respect to time and space from a plurality of time sequential frames of weather radar images for a plurality of frames therein, and which segments and extracts a field possessing similar velocities with respect to time and space as one precipitation field from said plurality of frames of weather radar images by employing said calculated velocity fields;

a moving velocity estimating step, which estimates the moving velocity of a precipitation field by obtaining the amount of movement, among the neighboring frames, of the precipitation field segmented and extracted by said field extracting step; and

a forecast image generating step, which extrapolates a plurality of precipitation fields by employing a distance taken by multiplying the moving velocity of the precipitation field estimated by said moving velocity estimating step by a desired forecasting time, and which generates a forecast radar image by synthesizing the resulting image.

23. A method for weather image prediction in accordance with claim 22 wherein said field extracting step calculates said velocity fields of the echoes from the location where the cross-correlation coefficient of the partial image gray level of two weather radar images is greatest.

24. A method for weather image prediction in accordance with claim 22 wherein said field extracting method obtains said precipitation field by way of comparing the respective velocity vectors of the adjacent blocks of said precipitation field with respect to space and time, and by way of segmenting the weather radar images so as to have fields possessing similar velocities belong to the same precipitation field.

25. A method for weather image prediction in accordance with claim 22 wherein said moving velocity estimating step obtains the velocity candidate of the precipitation field segmented and extracted by said field extracting step, by employing the weather radar images from either two proximate frames or a plurality of frames, determines the degree of mobility or immobility, which is the characteristics of the movement of the precipitation field, which in turn, is the echo cells or the group of echo cells within the weather radar image segmented and extracted by said field extracting step, and estimates the moving velocity of the precipitation field by taking into consideration said determined degree of mobility or immobility of the precipitation field, and with respect to said obtained velocity candidate of the precipitation field.

26. A method for weather image prediction in accordance with claim 22 wherein said moving velocity estimating step records a past moving velocity of the precipitation field segmented and extracted by said field extracting step, and estimates the current moving velocity of the precipitation field by taking into consideration said past moving velocity of the precipitation field.

27. A method for weather image prediction in accordance with claim 22 wherein said moving velocity estimating step obtains the global velocity candidate of the precipitation field segmented and extracted by said field extracting step, by employing the weather radar images from either two proximate frames or a plurality of frames, estimates the moving velocity of the precipitation field by employing the local velocities of the echoes within the weather radar images obtained by said field extracting step in the case where the forecasting time requested at said forecast image generating step is equal to or less than a predetermined value, and estimates the moving velocity of the precipitation field by employing said global velocity candidate of the precipitation field in any other case.

28. A computer readable recording medium which records a program for weather image prediction for execution by a computer comprising:

a field extracting step, which calculates the velocity fields of local echoes with respect to time and space from a plurality of time sequential frames of weather radar images for a plurality of frames therein, and which segments and extracts a field possessing similar velocities with respect to time and space as one precipitation field from said plurality of frames of weather radar images by employing said calculated velocity fields;

a moving velocity estimating step, which estimates the moving velocity of a precipitation field by obtaining the amount of movement, among the neighboring frames, of the precipitation field segmented and extracted by said field extracting step; and

a forecast image generating step, which extrapolates a plurality of precipitation fields by employing a distance taken by multiplying the moving velocity of the precipitation field estimated by said moving velocity estimating step by a desired forecasting time, and which generates a forecast radar image by synthesizing the resulting image.

29. A recording medium which records a program for weather image prediction in accordance with claim 28 wherein said field extracting step calculates said velocity fields of the echoes from the location where the cross-correlation coefficient of the partial image gray level of two weather radar images is greatest.

30. A recording medium which records a program for weather image prediction in accordance with claim 28 wherein said field extracting step obtains said precipitation field by way of comparing the respective velocity vectors of the adjacent blocks of said precipitation field with respect to space and time, and by way of segmenting the weather radar images so as to have fields possessing similar velocities belong to the same precipitation field.

31. A recording medium which records a program for weather image prediction in accordance with claim 28 wherein said moving velocity estimating step obtains the velocity candidate of the precipitation field segmented and extracted by said field extracting step, by employing the weather radar images from either two proximate frames or a plurality of frames, determines the degree of mobility or immobility, which is the characteristics of the movement of the precipitation field, which in turn, is the echo cells or the group of echo cells within the weather radar image segmented and extracted by said field extracting step, and estimates the moving velocity of the precipitation field by taking into consideration said determined degree of mobility or immobility of the precipitation field, and with respect to said obtained velocity candidate of the precipitation field.

32. A recording medium which records a program for weather image prediction in accordance with claim 28 wherein said moving velocity estimating step records a past moving velocity of the precipitation field segmented and extracted by said field extracting step, and estimates the current moving velocity of the precipitation field by taking into consideration said past moving velocity of the precipitation field.

33. A recording medium which records a program for weather image prediction in accordance with claim 28 wherein said moving velocity estimating step obtains the global velocity candidate of the precipitation field segmented and extracted by said field extracting step, by employing the weather radar images from either two proximate frames or a plurality of frames, estimates the moving velocity of the precipitation field by employing the local velocities of the echoes within the weather radar images obtained by said field extracting step in the case where the forecasting time requested at said forecast image generating step is equal to or less than a predetermined value, and estimates the moving velocity of the precipitation field by employing said global velocity candidate of the precipitation field in any other case.