Talking Points for Medium Range Teletraining

Steve Flood

NCEP HPC



1. Title Slide



2. Pattern Recognition in Forecasting



3. The Big Picture: Given by the Longwave Pattern



4. Tools Used in Assessing Patterns



5. Main Objectives

a) Describe how D charts are generated, why they are practical, and illustrate their use.

b) Develop a sense of the reliability of D charts as predictors.

c) See particular model errors based on particular patterns

d) Understand how a changing mean longwave pattern relates to changes in the character of model errors.



6. What is a "D" Chart?

a) An average of 10 500 Millibar Height Plots.

b) The Types of "D" Charts

c) Composition of a "D+0" Chart



7. Composition of Other Commonly used "D" Charts



8. Why are "D" charts Practical?



9. Example of a "D-3" Chart



10. Accessing "D" Charts From the Internet- Other Useful Web Sites



11. How representative are "D" Charts?



12. Interpreting a D+3, D+6, or D+8 chart using a D-3 Chart



13. General Application of D+3 means



14. Complex Application of D+3 means



15. Case Study: May 14-29, 1999



16. D-3 centered 5/14/99



17. D-3 centered 5/19/99



18. D-3 centered 5/24/99



19. D-3 centered 4/29/99



20. Longwave Pattern Becoming More Progressive



21. May 18: Errors with a Flattening Pattern



22. Errors with a More Progressive than Forecast Pattern



23. Continued Errors with a Progressive Pattern



24. Error Consistency in a Pattern



25. May 23: Gulf of Alaska Deepening



26. May 24: Downstream Amplification



27. Time Lag in Model Response



28. Question: What Model Errors, Time Lag do we see in Rapidly Amplifying Patterns?



29. More Model Errors in a Deamplifying Longwave Pattern



30. Pattern Change in Eastern Pacific



31. More Progressive Pattern Errors



32. Model Errors are Regime Dependent



33. Other Important Considerations



34. Importance of 12 hour difference in Run Time between the MRF and ECMWF

35. How do I adjust MOS in a regime of High Uncertainty?



36. How do I adjust MOS in a regime of High Certainty?



37. Review of Main Objectives given in Slide 5



38. Conclusion: Questions, Insights, Answers



39. Title Page- US West Coast Sequence Review



40. D-3 centered 5/29/99



41. D-3 centered 6/3/99



42. D-3 centered 6/8/99



43. D-3 centered 6/13/99



44. Model Verification May 29 1999



45. Model Verification May 31 1999



46. Model Verification June 2 1999- Eastern Pacific Ridge the Key



47. Underestimated digging along the W Coast.



48. Models verify well in CA, but MRF too strong with Eastern Pacific Ridge



49. Model Errors in Extreme Deepening in Eastern Pacific



50. Models miss new System in Western Gulf of Alaska



51. Question: Model Sensitivity to errors in the shape of the Eastern Pacific Ridge



52. How long do closed upper lows sit over CA? Relate May 23-25 model verification to changes in the observed D-3 means May 19th-24th.



53. D-3 Centered 5/14/99



54. D-3 centered 5/19/99



55. Model verification in Gulf of Alaska Low Deepening



56. Model Verification in Downstream troughing over North Central States



57. Model Time Lag Responding to Longwave Pattern.



58. D-3 centered 5/14/99



59. D-3 centered 5/19/99



60. D-3 centered 5/24/99



61. D-3 centered 5/29/99



62. Challenge in Cyclonic Patterns in the Gulf of Alaska



63. Models Underestimate Deepening in the Eastern Pacific



64. Model miss new system in the Western Gulf of Alaska



65. Model Error Consistency June 12-13 in Gulf of Alaska compared to the Western CONUS.



66. Conclusion of Western CONUS reviews