Transcript Atmospheric profile and precipitation properties derived from radar

Atmospheric profile and precipitation properties
derived from radar and radiosondes during
RICO
Louise Nuijens
With thanks to:
Bjorn Stevens (UCLA)
Margreet van Zanten &
Pier Siebesma (KNMI)
September 18 - 2006
In this talk
Analysis of S-Pol radar data
How frequent and with what intensity do trade
wind cumuli produce rain?
 from a large scale perspective
Analysis of sounding data
Can we relate this rainfall to the vertical
(dynamic) structure of the atmosphere?
Can we provide statistics for modeling studies?
Radar analysis: S-Pol
 S-Band horizontal (surveillance) scans
 Time resolution:
~ 70 scans a day
100 km
 a scan each 20 minutes
20 km
 Filtering of ground / sea clutter, birds and Bragg
scatter (10 dBZ threshold)
 Z-R relation:

Raw
Z  248  R
1 mm/h  24 dBZ > 10 dBZ
1.75
Uijlenhoet, 2004
Filtered
Data selection based on range attenuation
z ~ 500 m
20 km
-
60 km
The radar
observes rain
below cloud
base (~ 600 m)
Area rainfall during RICO:
Time series and probability of rainfall
Time series of area rainfall
Mean rainrate 0.026 mm/h ~ 18 W/m2
many small rain peaks
“trade cu period”
Mean echo coverage: 1.7 %
Rainfall probability
How much do the small peaks contribute to
the total rainfall?
50 %
0.125 mm/h
Echo structure and rain rates
Coverage = 2.55 %
Recho = 2.12 mm/h
Coverage = 1.41 %
Recho = 1.52 mm/h
Echo cells (3 km)
Rainband (> 20 km)
Dec 22,
05 UTC
Dec 29,
01 UTC
Can we relate the relative amount of rainfall
to the atmospheric profiles ?
Sounding analysis
Three type of soundings:
1) Spanish Point on Barbuda
2) R/V Seward Johnson
3) C-130 dropsondes
NE radar domain

q
SST
Sounding analysis
Do soundings relate to small rain events?
•
Relate each sounding to a 6 hour average area
rainrate
•
Group by following
rainrates in mm/h :
Dec 13, 14, 15
-2 18
9, Wm
13, 14,
1. 0 – 0.002
~Jan1.4
2. 0.002 – 0.02 ~ 15 Wm-2
3. > 0.02
NE
Vertical profiles
Specific humidity
0 - 0.002
0.002 - 0.02
>0.02
1.5 g/kg
Vertical profiles
Temperature
Vertical profiles
Relative humidity
15-20 %
Rainfall versus cloud depth
Estimating cloud depth for each sounding
Using a very simple parcel ascent scheme for a
diluting air parcel:
•
•
•
Air parcel released from max v in the lower subcloud
layer
Giving small release excesses l and qt
With entrainment rate inversely proportional to height of
parcel
Cloud base = LCL, cloud top  LZB (v, parcel = v)
Rainfall versus cloud depth
Some last thoughts…
•
How do the LES simulations and profiles
compare to these findings?
•
Difficult to obtain a close correlation between
parameters derived from soundings and the area
rainfall
– water vapor deficiency?
– theta gradient in cloud layer?
– …?
Random / scattered
Clusters?
Deeper convection
Comparison to other estimates
S-Pol rain rate coincides with relative rain rate from surface
station. Suggests that rain reaches the surface.