Weather model predictor

• US 7,792,664 B1
• Filed: 07/28/2006
• Issued: 09/07/2010
• Est. Priority Date: 07/28/2006
• Status: Active Grant

• Current Assignee: DTN, LLC (TBG AG)
• Original Assignee: WSI Incorporated
• Inventors: Neilley, Peter P., Crawford, Todd
• Primary Examiner(s): Rodriguez; Paul L
• Assistant Examiner(s): Kim; Eunhee
• Application Number:
  US11/495,300
• Time in Patent Office: 1,502 Days
• Field of Search: 703/6, 702/3, 706/62
• US Class Current: 703/6
• CPC Class Codes: G01W 1/10 Devices for predicting weather conditions computers per se G06; display devices G09

Abstract

A computer-implemented method of predicting an output of a model includes receiving output data from a first run of a target predictive model. One or more additional sets of output data is received from runs of one or more alternate predictive model. The alternate predictive models predict parameters that relate to the output data of the first predictive model. A future output of a second run of the target predictive model is predicted based on the output data from the first run of the target predictive model and one or more sets of additional output data from runs of the alternate prediction model.

Claims

1. A computer-implemented method of predicting by a first computer an output of a target predictive model, the method comprising:

(a) receiving, by the first computer from a target model data source maintained by a second computer different than the first computer, a complete set of output data produced as a result of a first run of the target predictive model;

(b) receiving, by the first computer from at least one alternate model data source maintained by a third computer different than the first computer, at least one additional complete set of output data, each additional complete set of output data produced as a result of a run of at least one alternate predictive model, wherein the at least one alternate predictive model identifies parameters related to the complete set of output data of the target predictive model; and

(c) predicting, by the first computer, a future output expected to be produced as a result of a second run of the target predictive model utilizing a statistical analysis of the complete set of output data from the first run of the target predictive model and the at least one additional complete set of output data from the at least one alternate predictive model, wherein the predicting of the future output expected to be produced as a result of the second run of the target predictive model occurs without the first computer performing runs of the target predictive model and the at least one alternate predictive model, and without observable weather data being used by the first computer in the predicting of the future output.

2. The method of claim 1, further comprising:

(d) displaying the predicted future output from step (c) on a display, wherein the display includes a visual indicator of the difference between the predicted future output and the output data of the first run.

3. The method of claim 1, wherein the target and alternate predictive models include weather models.

4. The method of claim 1, wherein the predicted future output includes a prediction of the direction of change of a subset of the output data produced as a result of the first run of the target predictive model.

5. The method of claim 1, wherein utilizing the statistical analysis includes output from a statistical model based on predictive patterns and statistical relationships between the output data from the target and alternate predictive models derived from output data produced as a result of previous runs of the target and alternate predictive models.

6. The method of claim 5, wherein the statistical model tracks an accuracy for each of the at least one alternate predictive models in predicting the output of a future run of the target predictive model by comparing previous outputs of the at least one alternate predictive model with previous outputs of the target predictive model.

7. The method of claim 6, wherein the accuracy of each of the at least one alternate predictive models in predicting the output of the future run of the target model is accounted for in the statistical model using variable factors.

8. The method of claim 5, wherein variable factors are used in the statistical model to minimize the difference between the predicted future output and actual output data produced as a result of the second run of the target predictive model.

9. The method of claim 8, wherein the variable factors include weighting coefficients.

10. The method of claim 1, wherein the future output expected to be produced as a result of the second run of the target predictive model is predicted by the first computer before the start of the second run of the target predictive model.

11. The method of claim 1, further comprising:

(d) evaluating a change in at least one of a future price and a future demand for a financial instrument based on the future output expected to be produced as a result of the second operational run of the target predictive model predicted in step (c); and

(e) initiating a business transaction based on the evaluated change in the at least one of the future price and the future demand for the financial instrument, wherein the business transaction is initiated prior to the completion of the second operational run of the target predictive model.

12. A computer-implemented method of predicting by a first computer an output of a target weather model, the method comprising:

(a) receiving, by the first computer from a first data source maintained by a second computer different than the first computer, a set of current output data produced as a result of a first operational run of the target weather model;

(b) receiving, by the first computer from a second data source maintained by a third computer different than the first computer, at least one additional set of output data, each produced as a result of a run of at least one alternate weather model, wherein the at least one alternate weather model identifies parameters related to the set of current output data of the target weather model; and

(c) predicting, by the first computer an output expected to be produced at a future time as a result of a second operational run of the target weather model based on the set of current output data and the at least one set of additional output data, wherein the output expected to be produced at the future time is obtained by the first computer prior to the start of the second operational run of the target weather model and occurs without the first computer performing runs of the target weather mode and the at least one alternate weather model, and without observable weather data being used by the first computer in predicting the output.

13. The method of claim 12, further comprising:

(d) transmitting the predicted output from step (c) to a user prior to the start of the second operational run of the target weather model.

14. The method of claim 12, further comprising:

(d) displaying the predicted output from step (c) on a display, wherein the display includes a visual indicator of the difference between the predicted output and the set of current output data of the first operational run of the target weather model.

15. The method of claim 12, wherein the current set of output data includes predicted future weather conditions.

16. The method of claim 12, wherein the at least one additional set of output data includes predicted future weather conditions.

17. The method of claim 12, wherein the at least one additional set of output data is produced after the first operational run of the target weather model.

18. The method of claim 12, wherein the at least one additional set of output data is produced prior to the first operational run of the target weather model.

19. The method of claim 12, wherein the predicted output is obtained using a statistical model based on predictive patterns and statistical relationships between the output data from the target and alternate predictive models derived from output data produced as a result of previous runs of the target and alternate weather models.

20. The method of claim 19, wherein the statistical model tracks an accuracy for each of the at least one alternate weather models in predicting the output of a future operational run of the target weather model by comparing previous outputs of the at least one alternate weather model with previous outputs of the target weather model.

21. The method of claim 20, wherein the accuracy of each of the at least one alternate weather models in predicting the output of the future operational run of the target weather model is accounted for in the statistical model using variable factors.

22. The method of claim 19, wherein variable factors are used in the statistical model to minimize the difference between the predicted output and the set of current output data produced as a result of the first operational run of the target weather model.

23. The method of claim 22, wherein the variable factors include extended regional weather patterns.

24. A system comprising a first computer for predicting an output of a target weather model, the system comprising:

a receiver configured to receive, from a first data source maintained by a second computer different than the first computer, a set of current output data produced as a result of a first operational run of the target weather model and to receive, from a second data source maintained by a third computer different than the first computer, at least one additional set of output data produced as a result of a run of at least one alternate weather model, wherein the at least one alternate weather model identifies parameters related to the set of current output data of the target weather model;

a memory configured to store the set of current output data and at least one additional set of output data; and

a processor configured to predict an expected direction of change of the set of current output data based on the at least one set of additional output data, wherein the expected direction of change is produced prior to the start of a second operational run of the target weather model, the processor predicting the expected direction of change independently of observable weather data, the target weather model, and the at least one alternate weather model.

25. The system of claim 24, further comprising:

a transmitter for transmitting the expected direction of change to a user prior to the start of the second operational run of the target weather model.

26. The system of claim 24, further comprising:

a display for displaying the expected direction of change.

27. The system of claim 24, wherein the expected direction of change is obtained using a statistical model based on output data from previous runs of the target and alternate weather models.

28. An article of manufacture for predicting an output of a target weather model, the article of manufacture comprising a computer-readable medium holding computer-executable instructions for performing a method comprising:

(a) receiving, by a first computer from a first data source maintained by a second computer different than the first computer, a set of current output data produced as a result of a first operational run of the target weather model;

(b) receiving by the first computer from a second data source maintained by a third computer different than the first computer, at least one additional set of output data, each produced as a result of a run of at least one alternate weather model, wherein the at least one alternate weather model identifies parameters related to the set of current output data of the target weather model; and

(c) predicting, by the first computer an output expected to be produced at a future time as a result of a second operational run of the target weather model based on the set or current output data and the at least one set of additional output data, wherein the output expected to be produced at the future time is obtained by the first computer prior to the start of the second operational run of the target weather model and occurs without the first computer performing runs of the target weather model, and the at least one alternate weather model, and without observable weather data being used by the first computer in predicting the output.

29. The article of manufacture of claim 28, the computer-executable instructions performing a method further comprising:

(d) transmitting the predicted output from step (c) to a user prior to the start of the second operational run of the target weather model.

30. The article of manufacture of claim 28, the computer-executable instructions performing a method further comprising:

(d) displaying the predicted output from step (c) on a display, wherein the display includes a visual indicator of the difference between the predicted output and the current set of output data of the first operational run of the target weather model.

31. The article of manufacture of claim 28, wherein the at least one additional set of output data is produced after the first operational run of the target weather model.

32. The article of manufacture of claim 28, wherein the at least one additional set of output data is produced prior to the first operational run of the target weather model.

33. The article of manufacture of claim 28, wherein the predicted output is obtained using a statistical model based on predictive patterns and statistical relationships between the output data from the target and alternate predictive models derived from output data produced as a result of previous runs of the target and alternate weather models.

34. The article of manufacture of claim 33, wherein variable factors are used in the statistical model to minimize the difference between the predicted output and the set of current output data produced as a result of the first operational run of the target weather model.

35. The article of manufacture of claim 34, wherein the variable factors include extended regional weather patterns.

36. The article of manufacture of claim 33, wherein the statistical model tracks an accuracy for each of the at least one alternate weather models in predicting the output of a future operational run of the target weather model by comparing previous outputs of the at least one alternate weather model with previous outputs of the target weather model.

37. The article of manufacture of claim 36, wherein the accuracy of each of the at least one alternate weather models in predicting the output of the future operational run of the target weather model is accounted for in the statistical model using variable factors.

38. A computer-implemented method of providing information related to a target predictive model, the method comprising:

(a) obtaining, by a first computer from a first data source maintained by a second computer different than the first computer, first output data produced as a result of a first run of the target predictive model;

(b) predicting, by the first computer, second output data expected to be produced as a result of a second run of the target predictive model prior to completion of the second run of the target predictive model; and

(c) providing, by the first computer prior to the start of the second run of the target predictive model, a comparative indication of the second output data expected to be produced as a result of the second run of the target predictive model with respect to the first output data produced as a result of the first run of the target predictive model, wherein the predicting of the second output data expected to be produced as a result of the second run of the target predictive model occurs without the first computer performing runs of the target predictive model and without observable weather data being used by the first computer in the predicting of the future output.

39. The method of claim 38, wherein the second output data includes a prediction of an expected direction of change of the first output data.

40. The method of claim 38, wherein the predicting in step (b) includes using the first output data from the first run of the target predictive model and at least one set of additional output data from runs of at least one alternate predictive model, the target model and the at least one alternate predictive model being weather models.

41. The method of claim 40, wherein the second output data is obtained using a statistical model based on predictive patterns and statistical relationships between the output data from the target and alternate predictive models derived from output data produced as a result of previous runs of the target predictive model and the at least one alternate predictive model, wherein the statistical model tracks an accuracy for each of the at least one alternate predictive models in predicting output of a future run of the target predictive model by comparing previous outputs of the at least one alternate predictive model with previous outputs of the target predictive model.

42. The method of claim 38, wherein the comparative indication includes a visual representation of the direction of change of the second output data with respect to the first output data.

43. The method of claim 38, wherein the comparative indication represents a prediction of an actual direction of change corresponding to actual output data obtained from the completion of the second run of the target predictive model.

44. The method of claim 38, wherein the comparative indication is used in evaluating a change in at least one of a future price for a financial instrument and a future demand for a commodity.

45. The method of claim 38, wherein the comparative indication is used to facilitate a business transaction prior to the completion of the second run of the target predictive model for a least one of a financial instrument and a commodity.

46. The method of claim 38, wherein the target predictive model is the Global Forecast System model, and the second run of the target predictive model is a 1200 UTC morning operational run of the Global Forecast System model.

47. The method of claim 38, wherein the comparative indication is provided to a subscriber in an email message.

48. A system comprising a first computer for providing information related to a target predictive model, the system comprising:

means for obtaining, from a first data source maintained by a second computer different than the first computer, first output data produced as a result of a first run of a target predictive model;

means for predicting second output data expected to be produced as a result of a second run of the target predictive model prior to completion of the second run of the target predictive model; and

means for providing, prior to the start of the second run of the target predictive model, a comparative indication of the second output data expected to be produced as a result of the second run of the target predictive model with respect to the first output data produced as a result of the first run of the target predictive model, wherein the predicting of the second output data expected to be produced as a result of the second run of the target predictive model occurs without the first computer performing runs of the target predictive model and without observable weather data being used by the first computer in the predicting of the future output.

49. The system of claim 48, wherein the means for predicting includes using the first output data from the first run of the target predictive model and at least one set of additional output data from runs of at least one alternate predictive model, the target and alternate predictive models being weather models.

50. The system of claim 49, wherein the means for predicting includes using a statistical model based on predictive patterns and statistical relationships between the output data from the target and alternate predictive models derived from output data produced as a result of previous runs of the target predictive model and the at least one alternate predictive model, wherein the statistical model tracks an accuracy for each of the at least one alternate predictive models in predicting output of a future run of the target predictive model by comparing previous outputs of the at least one alternate predictive model with previous outputs of the target predictive model.

51. The system of claim 48, wherein the comparative indication includes a visual representation of an expected direction of change of the second output data with respect to the first output data.

52. The system of claim 48, wherein the comparative indication represents a prediction of an actual direction of change corresponding to actual output data obtained from the completion of the second run of the target predictive model.

53. The system of claim 48, wherein the target predictive model is the Global Forecast System model, and the second run of the target predictive model is a 1200 UTC morning operational run of the Global Forecast System model.