Methods for generating data set

• US 20050192751A1
• Filed: 02/26/2004
• Published: 09/01/2005
• Est. Priority Date: 02/26/2004
• Status: Active Grant

• Current Assignee: Swiss Reinsurance Company
• Original Assignee: Swiss Reinsurance Company
• Inventors: Heck, Pamela, Lemcke, Gerry, Bresch, David N.
• Granted Patent: US 7,219,015 B2
• Time in Patent Office: Days
• US Class Current: 702/3
• CPC Class Codes: G01W 1/10 Devices for predicting weather conditions computers per se G06; display devices G09

Abstract

A computer method of generating a probabilistic dataset relating to a weather event includes inputting data representative of an historical weather event, and generating data representative of a plurality of alternative events. The data for the alternative events are generated, in certain embodiments, by a dependent sampling process. In particular embodiments, the dependent sampling process is a directed random walk process. Data for the alternative events may include data relating to the geographical positions or “tracks” of the event, data relating to atmospheric pressures, or both.

Claims

1. A computer method of generating a probabilistic data set relating to a weather event, comprising the steps of:

a) inputting data representative of an historical track of a weather event, said data including a plurality of points representative of geographical positions along the historical track; and

b) generating data representative of a plurality of alternative tracks based on said historical track, said data including a plurality of alternative points representative of geographical positions along said alternative tracks;

c) wherein said alternative points are generated from said points along the historical track by a dependent sampling process.

2. The method of claim 1, wherein said dependent sampling process is a directed random walk process.

3. The method of claim 1, wherein at least some of the plurality of alternative tracks have starting points that differ from a starting point of the historical track upon which said alternative tracks are based.

4. The method of claim 1, wherein the step of generating data representative of alternate tracks based on said historical track comprises the steps of:

1) generating a series of random tuples (x_r,y_r) for a historical point (x,y) of the historical track;

2) calculating a sum of random deviations (x′

,y′

) of the random tuples along the historical track; and

3) adding the sum of random deviations (x′

,y′

) to the historical point (x,y) of the historical track to produce alternative points along the alternative tracks.

5. The method of claim 1, wherein the step of inputting data representative of a track of an historical weather event comprises inputting a longitude and a latitude to define each of a plurality of points along said track.

6. The method of claim 1, wherein the step of inputting data includes inputting data representative of a plurality of historic tracks, and wherein the step of generating data includes generating a plurality of alternative tracks for more than one of said plurality of historical tracks.

7. The method of claim 1, wherein the step of inputting data representative of an historical track includes the step of inputting at least one of:

1) longitude and latitude of a plurality of points representative of the historical track;

2) an azimuth angle for at least some of the points along the historical track;

3) celerity for at least some of the points along the historical track;

4) a rate of change of azimuth angle for at least some of the points along the historical track; and

5) a rate of change of celerity for at least some of the points along the historical track.

8. The method of claim 1, further comprising the step of selecting a subset of the data representative of the alternative tracks for use in the probabilistic data set.

9. The method of claim 1, wherein the step of generating data representative of alternative tracks includes the step of limiting a variance of said alternative points from a respective historical point in accordance with one or more physical laws.

10. The method of claim 1, wherein the step of inputting data representative of a track of an historical weather event includes inputting data representative of intensity of the event.

11. The method of claim 10, wherein said data representative of intensity comprises atmospheric pressure data associated with at least some of the plurality of points along the historical track, said atmospheric pressure data defining an historical pressure profile of the historical track.

12. The method of claim 11, wherein said atmospheric pressure data includes an absolute pressure (P) and a derivative of absolute pressure with respect to time (dP/dT).

13. The method of claim 11, wherein said atmospheric pressure data includes a pressure distribution.

14. The method of claim 11, wherein the step of inputting data includes inputting data representative of a plurality of historical tracks, and further comprising the step of establishing a grid over a geographical area of interest, said area including at least a portion of the plurality of historical tracks.

15. The method of claim 14, further comprising the step of establishing a pressure climatology for selected cells in the grid, based upon the atmospheric pressure data associated with at least some of the plurality of points along the historical tracks located within said selected grid cells.

16. The method of claim 15, wherein the pressure climatology for at least one of the selected cells is a pressure distribution function.

17. The method of claim 15, wherein the pressure climatology for a selected cell in the grid is established from at least one of the atmospheric pressure data associated with the selected cell and the atmospheric pressure data associated with one or more cells adjacent the selected cell.

18. The method of claim 17, wherein the pressure climatology for a selected cell is established from a weighted averaging of pressure data associated with the selected cell and pressure data associated with one or more cells adjacent a selected cell.

19. The method of claim 17, wherein each cell in the grid is assigned a land/sea value, and wherein pressure data associated with an adjacent cell is used to establish the pressure climatology of a selected cell only if the adjacent cell and the selected cell have the same land/sea value.

20. The method of claim 15, comprising the additional step of generating one or more alternative pressure profiles for one or more of the historical tracks using the pressure climatology for the selected cells in the grid.

21. The method of claim 15, comprising the additional step of generating one or more pressure profiles for one or more of the alternative tracks.

22. The method of claim 21, further comprising the additional step of generating one or more alternative pressure profiles for one or more of the alternative tracks using the pressure climatology for the selected cells of the grid.

23. The method of claim 22, wherein at least one of the alternative pressure profiles for the historical tracks, the pressure profiles for the alternative tracks, and the alternative pressure profiles for the alternative tracks are modified based, at least in part, on the historical pressure profile along the historical track of the associated weather event.

24. The method of claim 1, wherein the step of inputting data includes inputting data representative of a plurality of historical tracks and inputting data representative of atmospheric pressure associated with at least some of the plurality of points along the historical tracks, said atmospheric data defining historical pressure profiles of the historical tracks, and wherein the step of generating data includes generating a plurality of alternative tracks for more than one of said plurality of historical tracks, and further comprising at least one of the following steps:

a) generating one or more alternative pressure profiles for one or more of the historical tracks;

b) generating one or more pressure profiles for one or more of the alternative tracks; and

c) generating one or more alternative pressure profiles for one or more of the alternative tracks.

25. The method of claim 23, further comprising the step of extracting a subset of data from the data representative of the historical tracks, the alternative tracks, and the pressure profiles, based on climatological conditions for a selected time period.

26. A computer method for generating a probabilistic data set relating to a weather event, comprising the steps of:

(a) inputting data representative of an historical weather event, said data including a plurality of points representative of geographical positions along a track of the historical weather event; and

(b) inputting data representative of atmospheric pressure associated with at least some of the plurality of points along the historical track, said atmospheric data defining a pressure profile of the historical weather event; and

(c) generating a plurality of alternative pressure profiles associated with the track of the historical weather event.

27. The method of claim 26, wherein the step of generating a plurality of alternative pressure profiles comprises the steps of:

(d) identifying a point of occurrence of an absolute pressure minimum along the track of the historical weather event;

(e) selecting an alternative pressure value at the point identified in step (d);

(f) adjusting pressure values at a plurality of other points along the track of the historical weather event, in accordance with the selected alternative pressure value, to create an alternative pressure profile; and

(g) repeating steps (d), (e), and (f) to produce a plurality of alternative pressure profiles associated with the track of the historical weather event.

28. The method of claim 26, further comprising the steps of generating data representative of a plurality of alternative tracks based on said track of the historical weather event, said data including a plurality of alternative points representative of geographical positions along said alternative tracks.

29. The method of claim 28, further comprising the step of generating a plurality of alternative pressure profiles associated with at least some of said plurality of alternative tracks.

30. The method of claim 29, wherein the step of generating said plurality of alternative pressure profiles associated with said plurality of alternative tracks comprises the steps of:

(d) identifying a point of occurrence of an absolute pressure minimum along one of said alternative tracks;

(e) selecting an alternative pressure value at the point identified in step (d);

(f) adjusting pressure values at a plurality of other points along said alternative track, in accordance with the selected alternative pressure value, to create an alternative pressure profile; and

(g) repeating steps (d), (e), and (f) to produce a plurality of alternative pressure profiles associated with said plurality of alternative tracks.

31. The method of claim 29, wherein the step of generating a plurality of alternative pressure profiles associated with at least some of the plurality of alternative tracks comprises the steps of:

(d) identifying a pressure value at a first position along the historical track, and setting a pressure value at a corresponding position along an alternative track equal to the identified pressure value;

(e) determining a percentile of pressure change along the historical track between said first location and a second location;

(f) varying the percentile by a selected amount;

(g) determining a pressure value at a second location of the alternative track based upon the varied percentile;

(h) repeating steps (d), (e), (f), and (g) for additional points along the alternative track to create an alternative pressure profile associated with the alternative track; and

(i) repeating steps (d), (e), (f), (g), and (h) to create pressure profiles for other ones of the plurality of alternative tracks.

32. The method of claim 31, wherein the percentile is varied by approximately plus/minus fifteen percent.

33. The method of claim 31, comprising the step of creating additional alternative pressure profiles by at least one of:

(a) selecting a different position along the historical track as a starting point; and

(b) varying the percentile by a different amount.