
Surface winds, divergence, and vorticity in
stratocumulus regions using QuikSCAT and reanalysis
winds
B. D. McNoldy, P. E. Ciesielski, W. H. Schubert, and R. H. Johnson
Department of Atmospheric Science, Colorado State University, Fort Collins, Colorado, USA
Received 19 February 2004; revised 16 March 2004; accepted 25 March 2004; published 22 April 2004.
[1] High spatial resolution QuikSCAT data are used to
examine surface winds, divergence, and vorticity over
oceanic regions during the boreal summer. These analyses
are compared to those from earlier observational studies,
as well as ECMWF and NCEP reanalysis products.
QuikSCAT analyses generally confirm the results from
early studies and the reanalyses, but add important details
to our view of the surface circulation in Atlantic and East
Pacific regions.
INDEX TERMS: 3309 Meteorology and
Atmospheric Dynamics: Climatology (1620); 3319 Meteorology
and Atmospheric Dynamics: General circ ulation; 3 394
Meteorology and Atmospheric Dynamics: Instruments and
techniques. Citation: McNoldy, B. D., P. E. Ciesielski, W. H.
Schubert, and R. H. Johnson (2004), Surface winds, divergence,
and vorticity in stratocumulus regions using QuikSCAT and
reanalysis winds, Geophys. Res. Lett., 31, L08105, doi:10.1029/
2004GL019768.
1. Introduction
[2] Persistent low-level stratocumulus clouds occupy
large portions of the eastern Pacific and eastern Atlantic
Oceans. These areas of stratocumulus convection are most
extensive during the boreal summer when upward motion
in the Inter-Tropical Convergence Zone (ITCZ) and
downward motion in the sub tropical highs are at a
maximum. Past attempts to analyze wind fields o ver
these vast oceanic, stratocumulus regimes h ave been
hampered by a paucity of observations. Using climato-
logical data from millions of ship reports du ring the
period 1885–1933 [McDonald, 1938] and radiosonde
data collected during research cruises in 1949–1952,
Neiburger et al. [1961] described the structure of the
lower troposphere off the coast of California. Since the
results of Neiburger et al. are base d on McDonald’s
analysis, hereafter their combined work will be referred
to as the N-M analysis. More recently, Dai and Deser
[1999] presented global maps of mean seasonal surface
winds and divergence based on t he Co mprehens ive
Ocean-Atmosphere Data Set (COADS) during 1976–
1997. To our knowledge, these are the only studies that
document the boreal summer, large-scale structure of
surface winds and divergence in the eastern ocean stra-
tocumulus regimes.
[
3] In t his study, winds derived from the SeaWinds
scatterometer (aboard QuikSCAT) are used to examine the
boreal summer surface winds and the derived fields of
divergence and vorticity over these regions. In light of the
widespread use of the N-M analysis for theoretical and
modeling work, comparisons of their surface wind/diver-
gence climatology with recent SeaWinds observations will
be examined. In addition, the QuikSCAT analyses will be
compared to those generated from coarser resolution NCEP
and ECMWF reanalysis products to assess the accuracy of
these model-derived fields.
2. Data
[4] SeaWinds is a microwave scatterometer on the
QuikSC AT satellite, launched in June 1999. QuikSCAT
winds, which cover 93% of the global ocean daily under
clear and cloudy conditions at 0.25 resolution, are cali-
brated to a reference height of 10 m. Because recent studies
have demonstrated the value of SeaWinds as an accurate
global oceanic wind sensor [Ebuchi et al., 2002; Pickett et
al., 2003; Bourassa et al., 2003], the QuikSCAT data set is
used here to judge the accuracy of the N-M analyses as well
as those from reanalysis products. For this latter compari-
son, we use 10-m reanalysis winds which are available
globally every 6 h.
[
5] The ECMWF data from the ERA-40 reanalysis
project, which used 3D variational techniques applied at
T159 (140 km) resolution, are available on a 2.5 grid.
On this same grid, the NCEP data are from their rean-
alyses done at T62 (360 km). While QuikSCAT data are
not used in either of these reanalysis systems, ERS1 and
ERS2 scatterometer winds, which have lower spatial
resolution and coverage than QuikSCAT winds, were
assimilated into ECMWF analyse s from the early 1990s
until January 2001. The QuikSCAT winds used in this
study are a Jet Propulsion Laboratory ( JPL) Level 3
product. The analyses and comparisons described herein
are based on a four-year mean of surface wind, diver-
gence, and vorticity fields computed for July (1999 –
2002).
3. Discussion of Surface Wind Analyses
[6] QuikSCAT analysis of streamlines and isotachs of the
surface wind speed (i.e., the magnitude of the vector mean
velocity) averaged for July 1999–2002 are shown in
Figure 1. Overall, these winds compare quite favorably
with the reanalysis winds (not shown). Table 1 lists the
spatial correlation coefficients, rms difference, and mean
biases between the July-mean QuikSCAT and reanalysis
winds. To compute these statistical parameters, reanalysis
fields were interpolated onto the 0.25 QuikSCAT grid. The
GEOPHYSICAL RESEARCH LETTERS, VOL. 31, L08105, doi:10.1029/2004GL019768, 2004
Copyright 2004 by the American Geophysical Union.
0094-8276/04/2004GL019768$05.00
L08105 1of5