1. Overview of the Dvorak Technique Visible and Infrared Technique Simplified Visible Technique given here (See Technical Report for full details) Uses patterns and measurements as seen on satellite imagery to assign a number (T number) representative of the cyclone’s strength. The T number scale runs from 0 to 8 in increments of 0.5.

2. Overview of the Dvorak Technique Cont’d In the following examples, only the Data T Number (DT) will be calculated, the final (official) T number assigned to a tropical cyclone includes further considerations. DT computations familiarize one to various tropical cyclone patterns.

3. Four Basic Patterns Curved Band Pattern Shear Pattern Central Dense Overcast (CDO) Pattern Eye Pattern

4. Patterns and associated T Numbers

5. Empirical relationship between T number and wind speed

6. Finding the Cloud System Center (CSC) First step in the Dvorak technique From Dvorak (1985): “The cloud system center is defined as the focal point of all the curved lines or bands of the cloud system. It can also be thought of as the point toward which the curved lines merge or spiral.” Several situations

7. Curved Band Pattern

8. DT number determined by curvature of band around 10  log spiral

9. Curved Band Pattern Cont’d 1.0 to 2.0 2.5 3.0 3.5 4.0 4.5 DT Number

10. Example: Tropical Storm Ivan 1115 UTC 23 September 1998

11. Example: Curved Band

12. Curved Band Pattern Tropical Storm Ivan curves 0.7 around log 10 spiral. This corresponds to DT=3

13. Shear Pattern

14. Shear Pattern DT Numbers 1° latitude = 60 nautical miles (nmi) = 111 km

15. Central Dense Overcast (CDO)

16. CDO No eye DT number determined by CF+BF=DT –CF=CENTRAL FEATURE –BF=BANDING FEATURE –DT=DATA T NUMBER

17. Example: Hurricane Georges 1545 UTC 21 September 1998

18. Example: CDO Central Feature (CF) Measure Diameter of CDO in degrees latitude For a well defined CDO –3/4 °CF=2 –1 1/4 °CF=3 –1 3/4 °CF=4 –>2 1/4 °CF=5 For an irregular CDO –1° to 1 1/2 °CF=2 –>1 1/2 °CF=3

19. Eye Pattern

20. DT number determined by CF+BF=DT –CF=CENTRAL FEATURE –BF=BANDING FEATURE –DT=DATA T NUMBER

21. Example: Hurricane Georges 1945 UTC 18 September 1998

22. Example: Eye - Central Feature (CF) CF=E-number+Eye Adjustment E-number a measure of the hurricane’s radius in degrees latitude –1/4°E-no.=3 –1/2° E-no.=4 –3/4°E-no.=5 –1°E-no.=6 –>1°E-no.=7

23. Eye Number

24. Eye - Central Feature Cont’d Eye adjustment 1. Poorly defined or ragged eyes: Subtract 0.5 for E-no.  4.5 and 1 for E-no.  5. 2. Large eyes: Limit T-no. to T6 for round, well defined eyes, and to T5 for large ragged eyes. 3. For MET  6, 0.5 or 1 may be added to DT for well defined eye in smooth CDO when DT < MET.

25. Eye Adjustment

26. Example: Eye - Banding Feature (BF) ( Same as with CDO)

27. Banding Feature (BF)

28. Data T Number CF + BF = DT CF = 6 - 1 = 5 BF = 0.5 DT = 5.5

29. Infrared (IR) Technique Can be used during night as well as during day At times more objective than visible technique

30. Example Digital IR: Hurricane Erika 1515 UTC 8 September 1997 Warmest eye pixel 16 °C Warmest pixel 30 nmi (55 km) from center - 71 °C Nomogram gives Eye no. =7

31. Summary The Dvorak technique uses patterns and measurements from satellite imagery to estimate the strength of a tropical cyclone. Four basic types –Curved band pattern –Shear pattern –CDO pattern –Eye pattern

32. ODT-Objective Dvorak Technique Original version – Dvorak (1984) – “analysis using digital IR data” Velden, Olander, Zehr (1998) – ODT Computation used for hurricane intensities remains essentially unchanged What is it? – Two IR temperature measurements, given a center location

33. Two IR temperature measurements 1) Surrounding temperature – Warmest pixel from those located on r=55 km circle 2) Eye temperature – Warmest pixel within the eye Table assigns intensity to nearest 0.1 T- No. Intensity increases as Surrounding T gets colder and as the Eye T gets warmer.

35. Isidore’02 vs Lili’02 Lowest MSLP –Isidore’02934 hPa –Lili’02 938 hPa Maximum Surface Wind Speed –Lili’02125 kt –Isidore’02 110 kt

38. Tropical Cyclone Pressure Wind Relationship Pressure : Wind = MSLP : Vmax –MSLP = minimum sea-level pressure –Vmax = maximum surface wind (10-m, 1-min wind)

39. P-Wind-Deviation (MSLP Vmax) Definition: Difference between observed MSLP and observed maximum wind speed converted to MSLP with a Pressure-Wind Relationship…. P-Wind Dev. = MSLP min – MSLP Dvorak, f (Vmax max )

41. 95-F 95-O 98-G 01-E 01-M 02-I 02-L 95-M 96-F 99-B 99-L 01-I

42. Lili’02 Isidore’02

45. Lili’02 Isidore’02 Erin’01 Opal

46. Lili’02 Isidore’02

47. Isidore Lili

48. Tropical Cyclone Surface Wind Analysis -- based entirely on satellite data (along with center location and storm motion) --with sufficient resolution to depict intensity (maximum wind and its location) -- with sufficient coverage to depict size (area with winds exceeding gale force). -- frequent time intervals

49. Independent Surface Wind Analyses from Four Components IR-Inner core, time continuity, rapid changes AMSU-Inner and outer winds Scatterometer -Outer winds, weak TCs Satellite Winds -Outer winds, weak TCs

50. “Satellite only” Tropical Cyclone Surface Wind Analysis COMBINATION OF FOUR ANALYSES IR-Inner core, time continuity, rapid changes AMSU-Inner and outer winds Scatterometer -Outer winds, weak TCs Satellite Winds -Outer winds, weak TCs –Global coverage –Time continuity –Consistent analysis

51. Validation and developmental data sets Aircraft winds Dropsonde winds Conventional obs (surface, ship, buoy) HRD Wind Analyses Numerical Model Analyses

52. CONCLUSION: In order for a Combined Satellite Tropical Cyclone Surface Wind Analysis to be successful and reliable – i.e… capture the rapid changes and have consistency among analyses. …. it MUST incorporate the Objective Dvorak Technique (ODT)

54. 32 Intense Hurricanes (Cat 3+) in Atlantic during 10-years, 1994-2003 Characteristics at time of max. intensity R-34 kt, n. mi. (azimuthal average): –Ave. 142Max. 231Min. 69 R-50 kt, n. mi. (azimuthal average): –Ave. 82Max. 150Min. 23 RMW (Radius of Maximum Wind Speed), n. mi.: –Ave. 20 Max. 30Min. 5 MSLP (Minimum Sea-level Pressure), hPa: –Ave. 941Min. 905 Max. 968

55. Size differences – Atlantic Intense Hurricanes 1994-2003 Cindy, ’99 – Average R-34 kt : 231 n.mi. – R-50 kt : 144 n.mi. – Intensity:120 kt, 942 hPa Iris, ’01 – Average R-34 kt: 69 n.mi. R-50 kt: 23 n.mi. -- Intensity:125 kt, 948 hPa

56. Hurricane CINDY – 0715 UTC 28 Aug 99

57. Hurricane IRIS – 0015 UTC 9 Oct 01

58. IR Cloud Area vs TC Size Hurricane Bret and Hurricane Floyd, at 125-kt intensity Bret’s Average R-34 kt is 68 n. mi. Floyd’s Average R-34 kt is 183 n. mi.

61. 20 Atlantic Intense Hurricanes ’95-’99 LOWEST MINIMUM SEA-LEVEL PRESSURE 1. Mitch 905 hPa 2. Opal916 3. Floyd921 4. Felix 929 5. Gert 930 6. Edouard933 7. Lenny 933 8. Luis 935 9. Hortense935 10. Georges937 11. Cindy942 12. Bret944 13. Fran946 14. Erik946 15. Marilyn949 16. Bonnie 954 17. Roxanne956 18. Bertha 960 19. Lili960 20. Isidore960 MAXIMUM SURFACE WIND SPEED 1. Mitch 155 knots 2. Floyd135 3. Lenny135 4. Georges135 5. Opal 130 6. Gert 130 7. Edouard125 8. Bret125 9. Felix 120 10. Luis120 11. Hortense120 12. Cindy 120 13. Erika110 14. Fran105 15. Marilyn100 16. Bonnie 100 17. Roxanne100 18. Bertha 100 19. Lili100 20. Isidore100

62. Intensification Rate -- from “Best Track” data, greatest MSLP decrease in 24 hr (- hPa / day) 1. Mitch54 2. Opal49 3. Edouard47 4. Felix 40 5. Floyd38 6. Bret35 7. Luis34 8. Lenny34 9. Hortense32 10. Georges31 – from Aircraft Center Fix MSLP, greatest decrease in 12 hr (- hPa / day) – limited to Cat 4 or greater hurricanes with aircraft observations 1. Opal80 2. Mitch54 3. Felix50 4. Floyd44 5. Hortense44 6. Lenny42 7. Bret42 NOTE: The following Cat 4’s did not have aircraft observations during initial intensification: Edouard, Luis, Georges, Cindy, Gert

63. Onset of Rapid Intensification Onset = Beginning of maximum intensification rate period Average Intensity at Onset (8 Atlantic cases) = 87 kt, 970 hPa What are IR image characteristics at Onset vs –Non – Intensifying Cat 2 at 85-90 kt –Pre – Cat 3 at 85-90 kt ?

66. IR Image Characteristics associated with Onset of Rapid Intensification Unusually cold IR temp.’s near the center Symmetric cloud patterns Outflow cirrus spreading all directions Trend of cold IR cloud becoming more concentrated toward the center Ragged eye becomes well-defined

67. Rapid vs Non-Rapid “Rapid” cases prior to onset have at least two and usually 5 of the characteristics “Non-Rapid” typically have one or less of the characteristics How can this be quantified and developed into an objective technique?

68. IR cloud asymmetries Measurements of IR temperature defined area centroid locations and their distance and bearing from surface center Related to : –Environmental vertical wind shear profile –TC motion –TC intensity and structure

71. OBTAINING WIND SPEED AND DIRECTION FROM THE OCEAN SURFACE USE SCATTEROMETRY (Theory) The scatterometer sensor is an active microwave imager that sends and receives microwave energy off the ocean surface Microwave energy is sensitive to the “roughness” of the ocean surface that is generated by the surface wind field. This roughness is manifested in small capillary size waves (or ripples) known as Brag Waves Due to the assymetric nature of these Brag Waves in relationship to the wind speed and direction, it is possible to derive a wind field from an inversion technique by viewing the same area of ocean from several angles

72. SCATTEROMER DAILY COVERAGE ERS-2 QUIKSCAT (ASCN) QUIKSCAT (DSCN)

73. BACKGROUND MICROWAVE RADIATION OVER THE OCEAN (Incident Angle Dependence) Fore and Aft look by the H-pol and V-pol sensor allow for up to 4 solutions No H-pol solution w/i 7 wind vector cells of edge Along subtrack, only get 180 deg opposite solutions EFFECTS ON POLARIZATION QuikSCAT (13.4 GHz) 52 o 46 o (13.4 GHz)

74. WIND RETRIEVAL and AMBIGUITY SOLUTIONS ERS-2 uses CMOD4 wind retrieval method to estimate wind speeds from Normalized Radar Cross-Section (Sigma-0) of backscatter microwave radiation over the oceans –3 Antenna at 3 different angles (can not see at nadir) –ECMWF used as initialization in ambiguity (direction) process QuikSCAT uses NSCAT2 (QuikSCAT1) wind retrieval method –Circular Scan at 2 zenith angles (46 o H-pol and 52 o V-pol) –Fore and Aft views allow up to 4 solutions –Solutions are “Ranked” based on Most Likelihood Estimator (MLE) –AVN used as initialization in ambiguity (direction) process –A multi-process “Buddy System” using a medium filter evaluates neighboring Wind Vector Cells (WVC) to make the final “Selection” –Each WVC is assigned a Rain Flag based on a likelihood determination

75. Model Function: Directional Ambiguities Model Function: Directional Ambiguities From Dr. M. Freilich, Oregon State University

76. QUIKSCAT PROBLEM AREAS ( Usually in Low Skill Areas) Edge of Swath (~ 7 wvc) and Along Sub-Track(3-4 wvc) Sensitivity to Heavy Rain –Surface Roughness (Especially in Low Wind areas!) –Rain Scattering Sensitivity to Errors in NWP Model in Low Skill Locations “Practical” Wind Regime between 05 AND 45 m/s –Problems in both LIGHT winds and very HEAVY winds Resolution (25 km) of Footprint will Limit Wind Retrieval in Tight Gradient Regions Ambiguity Selection Process and How Rain Flags are Treated (no direct measurement of rain on QuikSCAT) can Affect the Final Solution –Watch out for Rain Blocks caused by ‘rain contagion’

77. EDGE PROBLEMS Along the whole edge… OR small portion… FNMOC DISPLAY

78. PROBLEMS WITH RAIN CONTAMINATION and AMBIGUITY SELECTION PROCESS

79. RAIN SELECTION: MUDH vs RSS Which one correct?---Answer is somewhere in- between... TC Ando (04S) RSS FNMOC-NOGAPS FNMOC-NRT SAME Coverage MUCH LESS Coverage

80. Typical Rain Patterns Rain effects: Cross swath vectors Higher wind speeds Some intense rain not flagged RSS slide

81. TYPES OF RAIN PATTERNS and AMBIGUITY SELECTION ISSUES AND HOW TO SOLVE

82. RAIN EFFECTS (Rain Blocks--Perpendicular to Swath) In a ‘rain block’, one or two winds solutions may be bad due to excessive scattering of the signal. The ‘default’ direction appears to be perpendicular to the swath direction. Problem exists because of the buddy system approach to neighboring wind vector cells causing a ‘rain contagion’ effect. Rain Block Region Bad directions. Do not use! Speed may be ok.

83. RAIN EFFECTS (Solution: A fix position is possible. Use the good winds, outside of the rain-block) TY Jelawat, 05Aug 0942Z, TS Chanchu, 28 Jul 0624Z (90kt)

84. AMBIGUITY SELECTION (With Rain effects--TS 28W) FNMOC-NOGAPS FNMOC-NRT Compare the models and look for non-flagged winds This version treats non-flagged data separately from the ‘rain block’

85. AMBIGUITY SELECTION (With Rain effects--TY Soulik 31W) FNMOC-NOGAPSFNMOC-NRT

86. WHERE’S IS TROPICAL CYCLONE 21S (HUDAH)? Located within the trough--no Circulation! ? MAX WIND 95 KTS Try to fix in trough equator-ward of the strongest winds

87. QUIKSCAT-- MODEL INITIALIZATION TC 24S (Paul), 20 April ? ? c c MAX WIND 55 KTS (Light winds?) -----Low Skill AVN 19/12Z tau 24 20/2356Z 10S 20S In this case, poor model initialization combined with a lower skill nadir position, picks proper wind speed, but NO circulation center

89. SEA SURFACE TEMPERATURE FROM 12-18 AUGUST 2001 FROM 12-18 AUGUST 2001

90. OCEANIC HEAT CONTENT CLIMATOLOGY FOR AUGUST CLIMATOLOGY FOR AUGUST

91. Microwave Images DMSP SSMI – 85H Ghz NOAA AMSU-B – 89 GhZ

96. MODIS (Moderate Resolution Imaging Spectroradiometer) NASA Aqua and Terra 38-channel Imager, with Truecolor 250-m resolution

97. MODIS (Moderate Resolution Imaging Spectroradiometer) -- NASA EOS Terra and Aqua 38-channel imager– Truecolor images at 250-m resolution

103. NOAA N43RF departed 11:15am AST (1515 UTC), Sept 12, 2003, for 8-hour Ocean Winds mission into Hurricane Isabel

104. Hurricane Isabel near horizon through cockpit windows about half way between St Croix and the eye

105. Hurricane ISABEL’s Eye – 12 Sep 03

107. INTRODUCTION Definition -- SUBTROPICAL CYCLONE manifestation of “cut-off low” at the surface –definition -- cut-off low -- cold low displaced equatorward of westerly flow –occur in cool season –also called subtropical lows (Kona storms) –located in N. Pacific (15N to 35N -- 140W to 175E) and N. Atlantic (15N to 35N -- 30W to 60W) –primarily during November to March

108. Vertical Profile of Temperature Tropical Cyclone -- warm core through the troposphere, except in the lowest 1-2 km –warm core is strongest in the upper troposphere (~250 hPa), and over a broad area but very warm within the eye Easterly Waves and Monsoon Trough -- warm core in the upper troposphere, but cold core at lower levels (below ~600 hPa)

109. Vertical Temperature Profiles Subtropical Cyclone -- strong cold core in the upper/middle troposphere (~250-400 hPa) and weak cold core at lower levels –warm core above the strong cold core in the upper troposphere (~150-200 hPa), with a double tropopause structure ---------- OR ---------- in the stratosphere, with a tropopause lowering

110. Vertical Temperature Profiles Tropical Upper Troposphere Trough -- TUTT -- is comprised of upper level cyclones (TUTT cells) -- same as subtropical cyclones, except that the temperature anomalies are much weaker –shallow cyclones, confined to upper levels –lower troposphere winds and pressures are typically not influenced by TUTT cells above

111. Tangential Wind Profiles Tropical Cyclone -- strong, deep --maximum cyclonic wind at low-levels (1-2 km) Easterly Waves and Monsoon Trough -- weak, deep -- maximum cyclonic wind at mid- levels (~ 600 hPa)

112. Tangential Wind Profiles Subtropical Cyclone -- strong, deep -- maximum cyclonic wind at upper levels (~250 hPa) Tropical Upper Troposphere Trough -- weak, shallow -- maximum cyclonic wind at upper levels (~250 hPa)

113. Transition from Subtropical A few Tropical Cyclones originate from subtropical cyclones A few Subtropical Cyclones maintain subtropical characteristics but intensify to produce Storm Force winds (> 34 kt) (Subtropical Storm) During transition, a “hybrid” tropical storm has characteristics of both types

114. Terminology Clarifications The term “hybrid” is used to refer to a tropical cyclone that has originated not solely from “latent heat release” with a typical pre-existing tropical disturbance. This may involve baroclinic processes, as with extratropical cyclones, subtropical lows, etc. Other terms have been used for “hybrid” cyclone-- –half breed cyclone –neutercane –semitropical cyclone –intermediate cyclone

115. Terminology Clarifications Other tropical weather systems have been identified which may at times resemble or evolve into a “hybrid” cyclone –Monsoon Depressions –West African Cyclones –Arabian Sea Cyclones –Tropical Cyclones in Extratropical Transition

116. Extratropical Transition Extratropical: A term used in advisories and tropical summaries to indicate that a cyclone has lost its "tropical" characteristics. The term implies both poleward displacement of the cyclone and the conversion of the cyclone's primary energy source from the release of latent heat of condensation to baroclinic (the temperature contrast between warm and cold air masses) processes. It is important to note that cyclones can become extratropical and still retain wind of hurricane or tropical storm force.

117. The Hebert-Poteat Subtropical Cyclone Technique NATIONAL HURRICANE CENTER JACK BEVEN WHERE AMERICA’S CLIMATE AND WEATHER SERVICES BEGIN

118. What is the Hebert-Poteat Technique? A pattern-matching method of estimating the intensity of subtropical cyclones A compliment to the Dvorak technique

119. Similarities Between Hebert- Poteat and Dvorak Both techniques use convective overcast Both techniques use the distance of the Cloud System Center (CSC) from the overcast ST cloud features are selected so that ST- Numbers correspond to T-Numbers if the cyclone becomes tropical Both techniques assume modeled development of the cyclone, with the T or ST numbers normally changing by  1.0 per day

120. Differences Between Hebert-Poteat and Dvorak HP considers environment in determining cyclone type HP permits a classification of ST1.5 or ST2.5 on the first day HP cannot have the CSC under a Central Dense Overcast (CDO) HP uses curvature of convective features for all ST classifications in the absence of bands HP Designates a wind speed range for each ST category Translational speed excess above 20 kt added to the ST cloud feature wind estimate HP uses one rule regarding intensity changes

121. Hebert-Poteat Output

122. Hebert-Poteat Criteria ST 1.5 Low-level circulation center located 1/2 o to 2 o of latitude from the edge of poorly organized convection (not necessarily dense) For cold lows, connection may not be connected to other systems, and a small area (<3 o latitude) of deep layer convection exists near the center

123. Hebert-Poteat Criteria ST 2.5 Low-level circulation center located 1/2 o to 2 o from increased deep-layer convection (not necessarily dense) with greater curvature of broad cloud lines or bands than on the previous day Outer convective band 5 o -10 o east of the center, and possibly another convective band 2 o -4 o west-north of the center

124. Hebert-Poteat Criteria ST 3.0 Same criteria as for ST 2.5 except greater curvature of cloud lines or bands, and better organized convection Evidence of banding within 1 o of the circulation center

125. Hebert-Poteat Criteria ST 3.5 Deep-layer convection (frequently dense overcast) in band(s) 1 o -3 o from the center (no CDO) Outer convective band 5 o -10 o to the east weaker than for ST 3.0, but new band may form 5 o -10 o west of the center For systems moving rapidly eastward, there may be only a dense overcast (  3 o latitude) about 2 o -4 o east of the center

126. Hebert-Poteat Patterns