Global Monitoring ofGlobal Monitoring ofTropical Cyclones with aTropical Cyclones with a

DualDual--Frequency ScatterometerFrequency Scatterometer

Dr. Richard D. KnabbDeputy Director, Central Pacific Hurricane Center, Honolulu

Dr. Michael J. BrennanSenior Hurricane Specialist, National Hurricane Center, Miami

Workshop on Scatterometry and ClimateArlington, Virginia 19 August 2009

Hurricanes and Climate:Hurricanes and Climate:Why So Important?Why So Important?

• Long-term changes in hurricane frequency, intensities, and durations, depending on their existence and magnitude, could have significant societal ramifications worldwide due to direct impacts

• Tropical cyclones play a key role in the global energy and moisture budget, so changes in their intensities and durations could be significant

• Tropical cyclone intensity change is our #1 forecast improvement need, and requires accurate climatology

Hurricanes and Climate:Hurricanes and Climate:ChallengesChallenges

• The ongoing historical data record is insufficient for identifying long-term changes in global hurricane intensities and durations

• Difficult to improve forecasts on tropical cyclone • Difficult to improve forecasts on tropical cyclone intensity and structure, when those parameters are so often inaccurately observed

Ship logs and land observations

Transmitted ship observations

Radiosonde network (weather balloons)

Military Aircraft Reconnaissance

Coastal Radar Network (Conventional) (Doppler)

Aircraft Reconnaissance (Research)

Tropical Cyclone Observational Platforms since 1900

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Aircraft Reconnaissance (Research)

Polar-Orbiting Satellites (Conventional) (Microwave)

Geostationary Satellites

Aircraft launched Dropsondes (Omega) (GPS)

Ocean Data Buoys

Scatterometer Data (QuikSCAT)

Stepped Frequency Microwave Radiometer

Geostationary SatellitesGeostationary Satellites

• Best data source to monitor tropical cyclones on a nearly continuous basis

• Best data source to monitor tropical cyclones on a nearly continuous basiscontinuous basis

• In most of the world’s tropical cyclone basins, most of the time, it’s the only data source

continuous basis• In most of the

world’s tropical cyclone basins, most of the time, it’s the only data source

September 15, 2004September 15, 2004

The Dvorak TechniqueThe Dvorak TechniqueThe Dvorak TechniqueThe Dvorak Technique

•• A statistical method A statistical method for for estimatingestimating the the intensity of TCs from intensity of TCs from satellite imagerysatellite imagery

•• Uses infrared and Uses infrared and visible imageryvisible imagery

•• A statistical method A statistical method for for estimatingestimating the the intensity of TCs from intensity of TCs from satellite imagerysatellite imagery

•• Uses infrared and Uses infrared and visible imageryvisible imagery

Dvorak Technique Cloud PatternsDvorak Technique Cloud PatternsDvorak Technique Cloud PatternsDvorak Technique Cloud Patterns

visible imageryvisible imagery•• Uses “measurement” Uses “measurement”

of the TC cloud pattern of the TC cloud pattern and a set of rulesand a set of rules

•• Comes in manual and Comes in manual and automated versionsautomated versions

•• Related technique for Related technique for subtropical cyclonessubtropical cyclones

visible imageryvisible imagery•• Uses “measurement” Uses “measurement”

of the TC cloud pattern of the TC cloud pattern and a set of rulesand a set of rules

•• Comes in manual and Comes in manual and automated versionsautomated versions

•• Related technique for Related technique for subtropical cyclonessubtropical cyclones

No Threat to LandMostly Satellite Data

10

20

30

40

50

60

70

80

90

100

110

120

130

140

8/24 8/26 8/28 8/30 9/1 9/3 9/5

BEST TRACKSat (TAFB)Sat (SAB)Obj T-NumAC (sfc)AC (flt>sfc)AC (DVK P>W)QuikSCATSurfaceDrop (sfc)Drop (LLM xtrp)Drop (MBL xtrp)

Win

d S

peed

(kt

)

Date (Month/Day)

Hurricane Gustav25 August - 4 September 2008

Threat to LandMore Data

Dvorak Technique has limitations

U.S. Aircraft ReconnaissanceU.S. Aircraft ReconnaissanceU.S. Aircraft ReconnaissanceU.S. Aircraft Reconnaissance• Available when a TC threatens

land, but only in western Atlantic, far eastern Pacific, and central Pacific basins

• Location of circulation center, and minimum pressure

• Flight-level winds (along line)• GPS dropsonde profiles (single • GPS dropsonde profiles (single

spot vertical profiles)• Stepped-Frequency Microwave

Radiometer (SFMR; surface wind speed retrievals along line)

• Airborne doppler radar

105 kt

90 kt95 kt

10

Recon Flight Level (10,000 ft) Winds

Hurricane Georges 9/20/98 20-23Z

105 kt

90 kt

TC intensity and structure TC intensity and structure determination heavily dependent determination heavily dependent

upon available dataupon available data

Hurricane Humberto 2001

Flight level Flight level/drops Flight level/SFMR

QuikSCAT Limitations in TCs• At most two passes a day

at low and mid-latitudes

• Gaps between swaths approach 1000 km in deep Tropics

• Sensitive to rain � problem in TCs and elsewhere

• Can’t measure maximum wind in most hurricaneswind in most hurricanes– Resolution, instrument

design, rain effects

• Directional uncertainty limits ability to identify or locate TC centers

• Subjective analysis of “ambiguities” by forecaster required

• The unavailability of data near shore

DFS vs. QuikSCAT and XOWVMDFS vs. QuikSCAT and XOWVMSimulated Retrievals based on Katrina (2005)Simulated Retrievals based on Katrina (2005)

XOVWMQuikSCAT DFS

TS

H1

H2

H3

H4

H5

50

•DFS captures true wind signal where

QuikSCAT high winds are tied to rain

•DFS accurately depicts hurricane force

wind radii and retrieves winds into

category 2 range, but not into cat 3 range

•DFS cannot identify small scale wind

maxima seen by XOVWM

“Truth”

DFS vs. QuikSCAT and XOWVMDFS vs. QuikSCAT and XOWVMKatrina Simulated RetrievalsKatrina Simulated Retrievals

XOVWM

QuikSCAT

DFS

•DFS captures true wind signal

well into hurricane-force range

•DFS shows underestimation of

winds ≥ 80-85 kt (not seen in

XOVWM)

•Significant improvement over

QuikSCAT

“True” Wind from WRF simulations

XOVWM

Re

trie

ve

d W

ind

Sp

ee

d

Summary: Data LimitationsSummary: Data Limitations

• Geostationary satellite imagery indispensible for monitoring TCs worldwide, but Dvorak Technique for intensity estimation has limitations

• Aircraft data availability extremely limited, and even when available, suffer from severe sampling limitationslimitations

• QuikSCAT provides broad swath of surface wind data over many TCs, but has almost no utility for hurricane intensity, and requires manual interpretation to identify a TC’s closed circulation

Summary: Benefits of DFSSummary: Benefits of DFS• DFS would reliably measure a broad swath of surface

winds in TCs, including intensities and structures of many hurricanes (although not most major hurricanes), and circulation centers of developing TCs

• One DFS would have same sampling limitations as QuikSCAT due to polar-orbiting platformQuikSCAT due to polar-orbiting platform

• DFS would be a critical new tool in the box for developing more accurate climatologies of TC intensities, structures, and durations worldwide, especially when aircraft reconnaissance not available

• Addition of DFS would offer huge benefits for monitoring hurricane and climate relationships, and for improving forecasts of TC intensity and structure