Rawinsonde Kite Profiles

1
Rawinsonde Kite Profiles

• LT Dave Kuehn
• 18 Sep 2002

2
Comparing Kite Profiles to Bulk Method (Fairall
1996)

• Why
• Evap duct
• Operational significance
• How
• Brief overview on methods
• Matlab program
• What
• Results
• Kite data vs. Bulk Method

3
Ducting
300 200 100 0
Height (m)
M
4
Evaporation Duct
15 10 5 0
Z
Height (m)
Evap Duct
M
80 90 100 RH ()
5
Propagation Loss for Radar Within Evaporation Duct
Duct Ht 65 ft, radar _at_ 55 ft
Greatly Increased Detection Ranges Possible
Duct
6
Method and Manipulation

• Program assumes height based on press
• Negative heights not possible
• Modify based on in-situ height and time
  recordings (purple line)

7
Method and Manipulation

• Select bad data (purple)
• Ship influence
• Readings prior to launch
• Kite dipping
• Green is ship data
• Difference in press and RH

8
Method and Manipulation

• Select interval average based on similar airmass
  characteristics
• Pot Temp and RH and M profile data
• (red marks)

9

• Average over selected interval
• Select average air temp, sea temp and RH

• Fairalls Bulk Method derived Pot Temp, RH and M
  profile (black line)
• Mean kite data over interval (lavender circles)

10
Results17 July 20021400 - 1415
11
Results17 July 20021500 - 1520
12
Results19 July 20021945 - 1955

• Evap duct present later in
• afternoon on 19 July.
• Coincides closely with Rh
• gradient.
• Evap first noted at 1910 on 19 July

13
Results 19 July 1845 - 2010
Evap duct
No duct
Evap Duct 1910
Evap duct
14
Conclusions

• Subjective Error
• Surface height, kite height, sonde response,
  contaminated averages, air temp vs sea temp
• Bulk Method did not represent atmosphere well,
  high atmospheric variability (esp RH)
• Bulk Method always shows duct height 6-9 meters.
  Kite did not.
• When duct was measured (starting 1910 on 19 July)
  Bulk Method generally showed a lower duct height
  (7 meters) than kite data (11 meters)
• Evap ducts of lt1-2m extremely difficult to detect