QSCAT Rain Effect on Curl and Divergence
- Annual Averages of Wind Stress Curl and Divergence
- Wind Stress Curl and Wind Stress Divergence Biases from Rain Effects on QSCAT Surface Wind Retrievals, Journal of Atmospheric and Oceanic Technology , 2004, vol 21(8), p.1216-1231.
Annual Averages of Wind Stress Curl and Divergence
In the following figures, several data sets are referred to:
- QSCAT: scatterometer derived surface winds (DIRTH).
- QSCAT+NCEP blended: QSCAT observations (rain15 included) and NCEP-Reanalysis blended in space and time;
global, 6-hourly maps on 0.5° grid. Currently available from July 1999 to December 2002. Version 3.0 available in May 2003.
- OSU-NCEP-FNL: NCEP FNL is the final assimilation and forecast run in each
operational cycle, available 6-hourly on 1° grid. Global data were interpolated to QSCAT wind vector
cells (WVC) for each orbit. These data can be analyzed with or without rain-flagged WVC.
- NCEP-Reanalysis: Weather-center analyses, based on the forecast model and Global Data
Assimilation System (GDAS) version fixed for the reanalysis project; 6-hourly data on a T62 Gaussian
grid (approximately 2°).
- (no rain): rain-flagged WVC are not included.
- (rain,sp>15): rain-flagged WVC are only excluded when wind speed < 15 m/s.
- (rain): all rain-flagged WVC are included.
Fig. 1: Curl in year 2000 for QSCAT and QSCAT+NCEP blended.
Fig. 1a: Curl in year 2000 for QSCAT, QSCAT+NCEP blended, and NCEP Reanalysis.
All valid points are used to compute the zonal averages. Marginal seas and points along land are included.
Fig. 1b: Curl in year 2000 for QSCAT, QSCAT+NCEP blended, and NCEP Reanalysis, Atlantic only.
Marginal seas and points within ca. 2° along land are excluded.
Fig. 1c: Curl in year 2000 for QSCAT, QSCAT+NCEP blended. and NCEP Reanalysis, Pacific only.
Marginal seas and points within ca. 2° along land are excluded.
Fig. 2: Curl in year 2000 for OSU-NCEP-FNL, with and without rain-flgged WVC.
Fig. 3: Curl in 2000: blended (rain15) vs. QSCAT (no rain), and OSU-NCEP (rain) vs. (no rain).
The differences in the lower figure are dotted - solid from the upper figure, i.e. QSCAT (no rain) - blended (rain15), and OSU-NCEP (no rain) - (rain).
Fig. 4: Divergence in 2000: blended (rain15) vs. QSCAT (no rain), and OSU-NCEP (rain) vs. (no rain).
Fig. 5: Divergence in years 2000, 2001, and 2002 for QSCAT+NCEP blended.
Fig. 6: Curl in years 2000, 2001, and 2002 for QSCAT (rain,sp>=15).
Fig. 7: Curl in years 2000, 2001, and 2002 for OSU-NCEP-FNL (rain).
Fig. 8: Divergence in year 2000 for QSCAT and QSCAT+NCEP blended.
Fig. 9: Divergence in year 2000 for OSU-NCEP-FNL, with and without rain-flgged WVC.
Fig. 10: Global Map of Divergence in year 2000 for OSU-NCEP-FNL (rain - no rain).
Fig. 11: Global Map of Curl in year 2000 for OSU-NCEP-FNL (rain - no rain).
Wind Stress Curl and Wind Stress Divergence Biases from Rain Effects on QSCAT Surface Wind Retrievals
Ralph F. Milliff, Jan Morzel (Colorado Research Associates, a division of NorthWest Research Associates),
Dudley B. Chelton, and Michael H. Freilich (College of Ocean and Atmosphere Sciences, Oregon State University)
Journal of Atmospheric and Oceanic Technology , accepted March 2004
Abstract
Surface vector wind datasets from scatterometers provide essential
high resolution surface forcing information for analyses and models
of global atmosphere-ocean processes affecting weather and climate.
The importance of realistic amplitude, high-wavenumber, surface
wind forcing from scatterometer data has been demonstrated in a
variety of ocean modelling applications.
However, the radar backscatter signal from which surface vector
wind estimates are retrieved is attenuated and/or contaminated in heavy rain.
The QuikSCAT (QSCAT) dataset flags rain contaminated wind vector cells where
retrievals are either highly uncertain or not available.
Zonal and annual averages of wind stress curl and divergence
for 2000, 2001 and 2002 are derived and compared across three surface wind
datasets; QSCAT-only, reanalysis winds from the National Centers for
Environmental Predictions (NCEP-Reanalysis), and blended QSCAT+NCEP.
Missing QSCAT surface wind retrievals due to rain contamination lead to
statistically significant discrepancies of up to 50% in the
implied Sverdrup transports in sub-tropical and sub-polar gyre regions
of the Northern and Southern Hemispheres. Dataset to dataset
wind stress divergence amplitude differences due to
rain contamination are also large in the mid latitude storm track
regions. Discrepancies occur in the tropics due to rain
contamination effects on QSCAT data, and due to high-wavenumber
deficiencies in the NCEP-Reanalysis winds. In addition, NCEP operational
forecast model surface wind analyses (NCEP-FNL)
have been tri-linearly interpolated to the QSCAT wind vector cell locations
and sample times. The NCEP-FNL winds are not affected by rain,
so it is possible to compare NCEP-FNL interpolated surface
wind fields and related quantities calculated
with and without wind vectors at the rain-flagged
wind vector cell locations. When all locations are included,
wind stress curl amplitudes are found to be skewed
toward cyclonic curl in both hemispheres. Vectors at rain-flagged
locations in both hemispheres are also
skewed toward large-amplitude, cyclonic curls. This is because
mid-latitude synoptic systems are the meteorological sources of
large-amplitude cyclonic curls, as well as the places where the rain flag
bias in wind stress curl is largest.
Blended QSCAT+NCEP surface winds ameliorate the rain-flag induced biases in
zonal averages of wind stress curl and wind stress divergence, while
retaining high-wavenumber properties of the scatterometer winds.
Evidence to support rain flag algorithm refinement for high wind speeds
is presented.
The publication is available
on-line .
Surface Wind Data Sets
We will compare and contrast wind-stress curl computed
from 5 different surface wind datasets introduced in
this section. For the purposes of this paper, the surface wind
datasets are referred to as:
- QSCAT-only :
derived from the QuikSCAT spaceborne scatterometer mission
launched in July 1999. Rain-flagged wind vector cells (WVC) are not included.
- NCEP-Reanalysis :
from the continuation of the NCEP-reanalysis winds from the 40 year
(1957-1996) NCEP/NCAR reanalysis described in Kalnay et. al (1996). Zonal and meridional
components of the 10m wind are analyzed using the forecast model
and Global Data Assimilation System (GDAS) versions fixed for the
reanalysis project. The 10m wind component data are available for the
globe at 00, 06, 12, and 18 UTC each day, on Gaussian grids
corresponding to a T62 spectral truncation (about 2° resolution).
A growing archive is maintained at: http://dss.ucar.edu/datasets/ds090.0/.
The NCEP-reanalysis winds do not contain the influence of scatterometer
retrievals.
- blended QSCAT+NCEP :
The Chin et al. (1998) blending method has been adapted to the
QSCAT $10\,m$. This method uses monthly and regional estimates of the high-wavenumber
power-law relation as measured by QSCAT to create a blended scatterometer-weather
center analysis 10m wind product. In regions where satellite observations
had not occurred within 12hr, the NCEP-reanalysis winds were extended
according to the regional, monthly, power-law relation from QSCAT.
Otherwise, the QSCAT wind retrievals are reported directly.
Global, four-times daily blended QSCAT and NCEP-reanalysis winds are available
in the archive at: http://dss.ucar.edu/datasets/ds744.4/ on a 0.5° grid.
Updates to this archive occur roughly on 6 month intervals.
- OSU-NCEP-FNL-norain :
The final datasets for comparison in this work are produced by the
Ocean Vector Wind research group at Oregon State University (OSU),
College of Ocean and Atmosphere Sciences. The OSU group has interpolated
the NCEP FNL fields in space and time to the WVC locations for each
QSCAT swath. The NCEP FNL run is the final assimilation and forecast
run in each operational cycle. The FNL output of interest includes
6hr forecasts of the 10m winds available at 00, 06, 12, and
18UTC, after late arriving conventional and satellite data have
been assimilated in the Global Data Assimilation System (GDAS) that is
operational at NCEP. The FNL winds are converted from the spectral
representation of the forecast model (changing to T254 in late 2002)
to a 1° grid for dissemination. The OSU group tri-linearly interpolates
the NCEP-FNL to the QSCAT WVC for each orbit in the QSCAT data record.
AS in the the QSCAT-only dataset, all rain-contaminated WVC have been
excluded.
- OSU-NCEP-FNL-rain :
As above, but also included are all WVC which were rain-flagged.
Fig. 1: Power spectral density (PSD) vs. wavenumber spectra or the zonal component of the surface wind
Comparison of spectral energy density of zonal velocity (U). Spectra are based on 30° long North-South segments. The blended product and the NCEP-Reanalysis are available on a 0.5° x 0.5° global grid. In the area of interest, they were sampled along longitudes 180°, 190°, 200°, 210°, and 220° E, from 20° to 50° N. The satellite data were sub-sampled on 30° long segments along the satellite track at 25km resolution (QSCAT and collocated NCEP-FNL). Only segments with at most two consecutive points missing (i.e. a gap of ca. 75km) were included. All rain-flagged wind vector cells (WVC) were eliminated. In the case of NCEP-FNL, the spectra were also computed when all rain-flagged WVC were included. The annual averages are based on the following number of spectra for each data set: 7300 (blended and NCEP-Reanalysis), 3500 (QSCAT and collocated NCEP-FNL), and 7400 (NCEP-FNL with rain WVC).
Fig. 2: Zonal Average of Wind Stress Curl
Zonal and annual average wind stress curl for (a) 2000, (b) 2001,
and (c) 2002.
The blue lines are the zonal annual average
wind stress curl computed from the QSCAT-only datset (all rain-flagged data excluded); the green lines
represent QSCAT-rain15, which includes rain-flagged data when wind
speed is >= 15m/s; black lines are for the NCEP-reanalysis case; and red lines are for the
blended QSCAT+NCEP. Approximate confidence intervals are computed from a bootstrap method, and
are plotted at latitudes 50° S, 30° S, 20° S, 0°, 20° N, 30° N, and 50° N.
Panel (d) shows the number of wind stress curls averaged as a function of latitude and dataset. The NCEP-reanalysis and blended QSCAT+NCEP datasets occur 4-times daily
on regular 0.5° grids and thus contribute more wind stress
curls to each zonal annual average (red and black dashed lines), than
does the QSCAT-only dataset (blue line).
Fig. 3: Wind Stress Curl Histograms
Wind stress curl histograms for the calendar year 2000,
in the latitude band 41$deg - 43$deg N for the QSCAT-only (thick line),
OSU-NCEP-FNL with rain WVC included (thin line with bubble), and
OSU-NCEP-FNL with rain WVC excluded (thin line). In panel (a)
the histograms are depicted using a logarithmic ordinate. In panel
(b) the OSU-NCEP-FNL datasets are normalized by the different total
numbers of wind stress curls in each distribution; OSU-NCEP-FNL with
rain WVC included contains 1,279,775 curls, and OSU-NCEP-FNL excluding
the rain-flagged WVC contains 1,217,488 curls. Each wind stress
curl bin is multiplied by the curl magnitude value for the bin
center. Panel (c) depicts the difference of the standardized
histograms in panel (d); "rain included" minus "rain excluded".
Bins are shaded in the difference histogram for which the difference is
positive. A bias toward positive wind stress curl
in the rain-flagged WVC for this latitude band is evident
Fig. 4: Wind Stress Curl Histogram Differences
Standardized wind stress curl difference histograms for
the latitude bands (a) 50° - 53° N, (b) 34° - 37° N,
and (c) 10° - 13° N; computed as in Fig. 3 panel (c).
Fig. 5: Maps of Wind Stress Curl Differences, OSU-NCEP-FNL rain - no rain, year 2000
Global distribution of the annual average wind stress curl
difference; OSU-NCEP-FNL rain included minus OSU-NCEP-FNL rain excluded
for the calendar year 2000. The rain-flagged WVC in the Northern (Southern)
Hemisphere are dominantly positive (negative). Large magnitude differences
occur in storm track regions in both hemispheres and in regions of tropical
convergence.
Wind stress curl can also be computed for OSU-NCEP-FNL by excluding rain-flagged data only when wind speeds are less than
15m/s. This product is called "rain15". The difference of rain15 minus no rain is shown in the map below.
The next figure shows the difference of "rain" minus "rain15", i.e. the difference of the two maps above.
When excluding all rain-flagged data, there is a big bias in wind stress curl. Including rain-flagged data at strong wind speeds ameliorates this bias.
Fig. 6: OSU-NCEP-FNL surface winds, colocated to QSCAT observation swaths
A typical 5hr sample of a synoptic event occurring
in the North Pacific on 31 January 2000.
Shown are three consecutive ascending satellite swaths, passing
over the North Pacific from East to West.
Panel (a) depicts the
surface wind vectors from the OSU-NCEP-FNL datasets. The WVC locations
that are rain-flagged in the QSCAT-only data are indicated with red wind
vectors. Panel (b) shows the wind stress curl analysis for the
same swaths using the "rain included" OSU-NCEP-FNL winds (i.e. black and
red wind vectors in panel a). Panel (c) shows the wind stress
curl analysis derived from the QSCAT-only (no-rain) data for the same event.
Fig. 7: Zonal Average of Wind Stress Divergence
Zonal and annual average wind stress divergence for (a) 2000, (b) 2001,
and (c) 2002. Color scheme is the same as in figure 2. Number of wind stress divergences
averaged as a function of latitude and dataset are the same as for wind stress curl, and are shown for year 2000 in Fig.2 (d).
Fig. 8: Map of Wind Stress Divergence Differences, OSU-NCEP-FNL rain - no rain, year 2000
Global distribution of the annual average wind stress divergence difference; OSU-NCEP-FNL rain included minus
OSU-NCEP-FNL rain excluded for the calendar year 2000.
last modified on March 12, 2003
Go Back
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)