The relationship between Wind Speed and Precipitation in the Pacific ITCZ

1
The relationship between Wind Speed and
Precipitation in the Pacific ITCZ

• Larissa Back and Chris Bretherton
• University of Washington
• Dept. of Atmospheric Sciences

2
Motivation use observations to

• Further our understanding of why deep convection
  occurs where it does over the tropical oceans
• Find basic relationships that help constrain
  convective parameterization problem
• Compare statistical relationship to
  parameterizations
• Predict convection
• In this case, looking at relation between
  evaporation and convection as suggested by
  EPIC2001 data

3
Passive microwave data analysis method for
surface wind speed, water vapor path,
precipitation

• SSM/I (Special Sensor Microwave Imager) and TMI
  (TRMM Microwave Imager)
• (Wentz 97 algorithm used for both)
• 1998 2001 daily averages, ¼ degree resolution
  averaged up to 2.5 x 2.5 degree grid, some
  figures 5 x 5 degree grid points
• 2-6 satellite passes daily in tropics
• Grid-point day refers to one day spatially
  averaged data at given location
• Retrieval algorithm - retrieval errors
  uncorrelated (Wentz 97)

4
Precipitation increases exponentially with column
relative humidityfrom Bretherton, Peters, Back,
04, J. Climate

• Water Vapor Path (WVP) / Saturation WVP (from
  ECMWF)

5
Does satellite data (SSM/I-TMI) show wind speed-
precipitation correlation at EPIC location? (as
in Raymond et. al, 2003)

• each x is mean of 60 gridpoint-days (5 x 5
  degrees)
• binned by windspeed
• ( evaporation )
• Correlation coefficient 0.32

6
Is there a similar correlation in other parts of
the ITCZ?
No, but

• Also, trade-cumulus boundary layers are example
  of high winds with little precipitation

7
Is there a similar correlation in other parts of
the ITCZ?
-Precipitation increases exponentially with
humidity
Water Vapor Path (WVP) / Saturation WVP (WVP if
atmosphere were fully saturated) From Bretherton,
Peters, Back, 04 J. Climate
8
Is there a similar correlation in other parts of
the ITCZ?
Use only grid-point days with column rel. hum. gt
0.75
9
Latitude 10 N
160 E
175 E
170 W
155 W
Precipitation mm/day
140 W
95 W
125 W
110 W
Windspeed m/s
10
Are They Separate Effects?

• Windspeed Precipitation

• Relative Humidity Precipitation

11
Col. rel. hum. Bins 120W-85W combined
Microwave Precipitation (mm/day)
SSM/I-TMI wind speed (m/s)
QuikSCAT Vector mean wind speed (m/s)
12
Convergence Feedback (95 W)

• Substantially greater increase in precipitation
  than in evaporation

dP/dE gt 1
To first order Precipitation moisture
convergence (not causal though)
Precipitation (mm/day)
Increased evaporation triggers convection which
is associated with moisture convergence
Evaporation (NCEP)
13
Wind speed- Precipitation Correlation in
Reanalyses ERA40 NCEP SSM/I-TMI
NCEP has a stronger, more consistent correlation
than ERA40 (even w/assimilation)
14
Column integrated horizontal moisture advection
as a function of wind speed
horizontal dry advection increases sometimes
balance evaporation (this is also the case in the
boundary layer)
15
Summary/Conclusions

• More wind more rainfall
• More humidity associated with greater increase in
  precipitation with wind speed
• Geographic variability in convergence feedback
• Wind- precipitation correlation in NCEP similar
  to data. Not consistent in ERA40
• Dry advection may be comparable to evaporation in
  parts of the ITCZ, complicating interpretation of
  this relation
• Full manuscript (accepted by J. Clim.) available
  at http//www.atmos.washington.edu/larissa