Presentation is loading. Please wait.

Presentation is loading. Please wait.

Transition of a Revised Dynamical Model Track Prediction Evaluation Expert System (DYMES) Mark Boothe Department of Meteorology Naval Postgraduate School.

Similar presentations


Presentation on theme: "Transition of a Revised Dynamical Model Track Prediction Evaluation Expert System (DYMES) Mark Boothe Department of Meteorology Naval Postgraduate School."— Presentation transcript:

1 Transition of a Revised Dynamical Model Track Prediction Evaluation Expert System (DYMES) Mark Boothe Department of Meteorology Naval Postgraduate School LT David Roberts Tropical Prediction Center National Hurricane Center Funding provided by: USWRP Acknowledgments: Dan Martinez and Sean Wells Computer Sciences Corporation

2 Environment Structure TC Structure Synoptic Pattern Synoptic Region Intensity Size Transitional Mechanisms Environmental Effects TC-Environment Transformations Determines the direction & speed of the steering flow that moves the TC TC-independent processes that change environment structure (steering) TC-dependent processes that change environmental structure (steering) and sometimes TC structure Inputs to TC-Environment Transformations Results of Transitional Mechanisms Determines best steering level Key factor in TC-Environment Transformations TC Motion Meteorology Knowledge Base Framework

3 Environment Structure TC Structure Synoptic Pattern Synoptic Region Intensity Size Transitional Mechanisms Environmental Effects TC-Environment Transformations Inputs to TC-Environment Transformations Results of Transitional Mechanisms TC Motion Meteorology Knowledge Base Framework Equatorial Westerlies (EW) Tropical Easterlies (TE) Poleward Flow (PF) / Equatorward Flow (EF) Midlatitude Westerlies (MW) / Easterlies (ME) Exposed Low-level (XL) Tropical Depression (TD) Tropical Storm (TS) Hurricane (H) Midget (M) Small (S) Average(A) Large (L) Advection by Environment (ADV) Upper-level Low Evolutions: Upper-level Low Formation (ULF) Upper-level Low Dissipation (ULD) Midlatitude System Evolutions: Cyclogenesis (MCG) Cyclolysis (MCL) Anticyclogenesis (MAG) Anticyclolysis (MAL) Beta Effect-Related: Beta-Effect Propagation (BEP) Ridge Modification by TC (RMT) Reverse Trough Formation (RTF) Cyclone Interactions: Midlatitude-Related: Direct (DCI) Response to Vertical Semi-direct (SCI) Shear (RVS) Indirect (ICI) Baroclinic Cyclone Interaction (BCI) Standard (S) Upper-level Low (U) Poleward (P) Midlatitude (M)

4

5 Atlantic Frequent Error Mechanisms NGPSGFDNUKMOGFSGFDL E-DCIXX E-RMTXXX E-RVSX I-BCIXXX E-TrofX

6 Excessive Midlatitude Trough ------------- Insufficient Midlatitude Trough

7 Consensus Forecasts GUNAAviation (GFS) GFDL UK Met GUNsNOGAPS NCONGFDN

8 S

9 Small Spread Large Error Small Spread Small Error Large Spread Large Error Large Spread Small Error 150 7 26 12 195 Cases NCON NCON OK NCON can be improved

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63 Potential Improvement for Large Spread/Large Error Cases (12 Cases in 2003) Average 72-h Forecast Track Error (n mi) NCON342 SCON 170 OFCL 354

64 Model errors in bins Still a fair number of large errors for each model

65 Prior to 2003 NHC-Requested Code Modifications Utilize NOGAPS four times a day Use the ATCF’s simpler interpolated track Add 12 and 36 h positions to NCON & SCON Display model fields at 6-hourly interval

66 Flexible Fields Display Interval Quick Jump To Start Of Model Runs

67 Plan: Modifications Prior to 2004 Slight changes to error mechanisms Display height fields Display difference fields –Comparison of current models –Intercomparison for trend analysis

68 Number of Large (> 250 n mi) 72-hour Forecasts Atlantic 2003 23Aviation (GFS) 20GFDL 36UK Met 37 NOGAPS 27GFDN

69 Excessive Midlatitude Trough ------------- Insufficient Midlatitude Trough

70 Plan: Modifications Prior to 2004 Slight changes to error mechanisms Display height fields Display difference fields –Comparison of current models –Intercomparison for trend analysis

71 HghtWind HghtWind Difference

72 HghtWind HghtWind Difference

73 HghtWind HghtWind Difference

74

75

76

77

78

79 AVNO GFDN UKMO NGPS GFDL

80

81 E-DCI degrading UKMO ?

82 Excessive Direct Cyclone Interaction (E-DCI)

83 E-RMT degrading AVNO ?

84 RMT

85

86

87

88

89

90

91

92 Verifying position Value added: 48 n mi

93

94

95

96 Rejected at least one dynamical model (SCON = NCON) and 72h forecast verified 27 times Improved 11 Degraded 15 Equal 1 Average Errors (n mi): NCON: 211 SCON: 222

97 AVNO GFDL GFDN UKMO NGPS

98

99 E-DCI degrading GFDN ?

100 Excessive Direct Cyclone Interaction (E-DCI)

101 E-BCI degrading GFDL ?

102 BCI

103

104

105

106

107

108 AVNO GFDL GFDN UKMO NGPS Degraded By I-BCI ? Degraded By E-BCI ?

109 (AVNO) (GFDL)

110

111

112

113 Verifying position

114

115

116 Atlantic 2002 Ran DYMES 191 times 63% 12% 25% Assigned: Small Spread 120 Compensating Errors 23 One error Mechanism 48

117 PLANS Add to the model traits knowledge base by evaluating rest of large model errors (72 h FTE greater than 250 n mi) of 2003 Provide on-line training module by mid May Intensive real-time test during last half of July


Download ppt "Transition of a Revised Dynamical Model Track Prediction Evaluation Expert System (DYMES) Mark Boothe Department of Meteorology Naval Postgraduate School."

Similar presentations


Ads by Google