78
W. Alpers
In order to retrieve wind fields from the NRCS values measured by scatterometers (and SARs), one needs a transfer function, usually called Geophysical Model
Function (GMF) or wind scatterometer model functions, which relates NRCS values to wind vectors for a neutrally stable atmosphere at a height of 10 m above the
sea surface. Several such functions are in use for different scatterometers operating
at different radar frequencies. For C-band scatterometers, the most popular GMFs
are CMOD4 (Stoffelen and Anderson 1997), CMOD_IFR2 (Quilfen et al. 1998)
CMOD5 (Hersbach et al. 2007), and the CMOD5n (Hersbach 2008). The latter one
is presently used to retrieve neutral winds (not real 10 m winds) from ASCAT data.
All GMFs saturate at wind speeds around 25 m/s, independent of the radar
frequency. Thus it seems impossible to retrieve winds higher than 25 m/s from scatterometer data. However, recently an advanced Ku-band GMF has been developed
by Remote Sensing Systems (RSS) (Ricciardulli and Wentz 2011) that seem to allow retrieving also higher winds from Quikscat data. A disadvantage of Ku-band
scatterometers compared to C-band scatterometers is that Ku-band microwaves interact strongly with rain drops in the atmosphere (volume scattering and, attenuation)
causing biases in the wind retrieval when rain is present.
4.2.1 Tropical Cyclone
Examples of near surface wind fields retrieved from Quikscat data are depicted in
Fig. 4.1. These wind maps have been obtained from the website of Remote Sensing
Systems
4 . The four wind maps show the passage of the tropical cyclone Flavio from
the Indian Ocean towards to coast of Mozambique, East Africa. Flavio made landfall
on 22 February 2007 at the Mozambican coast near the city of Inhambane, where
it caused considerable damage. Note that the direction of rotation in the cyclone is
clockwise, which is typical for a southern-hemisphere cyclone. The scatterometer
data were acquired at 03:12 UTC on 18 February 2007 (upper left panel) at 0400 UTC
on 20 February (upper right panel), at 1548 UTC on 20 February (lower left panel),
and at 1544 UTC on 21 February (lower right panel). The data shown in panels a
and b were obtained during descending (morning) satellite passes of the satellite and
the one shown in Panels c and d during ascending (evening) passes. Such wind maps
are very useful for tracking the position of cyclones. While, in general, the wind
speeds and directions derived from Quickscat are quite reliable, this does not apply
for the inner regions of a tropical cyclone (or typhoon/hurricane), where the winds
can be quite high (above 25 m/s). As mentioned before, retrieving near-surface wind
fields with speeds above 25 m/s is questionable at all radar bands because above
this value the GMF saturates. At Ku band, the retrieval of typhoon/hurricane winds
from scatterometer data is further aggravated by the fact that typhoons/hurricanes are
associated with heavy rain. Raindrops in the atmosphere both scatter and attenuate
Ku-band microwaves and cause errors in the wind retrieval algorithm.
4 At http://www.ssmi.com/qscat/scatterometer_data_daily.html?rgn=indian_west&size=large.
W. Alpers
In order to retrieve wind fields from the NRCS values measured by scatterometers (and SARs), one needs a transfer function, usually called Geophysical Model
Function (GMF) or wind scatterometer model functions, which relates NRCS values to wind vectors for a neutrally stable atmosphere at a height of 10 m above the
sea surface. Several such functions are in use for different scatterometers operating
at different radar frequencies. For C-band scatterometers, the most popular GMFs
are CMOD4 (Stoffelen and Anderson 1997), CMOD_IFR2 (Quilfen et al. 1998)
CMOD5 (Hersbach et al. 2007), and the CMOD5n (Hersbach 2008). The latter one
is presently used to retrieve neutral winds (not real 10 m winds) from ASCAT data.
All GMFs saturate at wind speeds around 25 m/s, independent of the radar
frequency. Thus it seems impossible to retrieve winds higher than 25 m/s from scatterometer data. However, recently an advanced Ku-band GMF has been developed
by Remote Sensing Systems (RSS) (Ricciardulli and Wentz 2011) that seem to allow retrieving also higher winds from Quikscat data. A disadvantage of Ku-band
scatterometers compared to C-band scatterometers is that Ku-band microwaves interact strongly with rain drops in the atmosphere (volume scattering and, attenuation)
causing biases in the wind retrieval when rain is present.
4.2.1 Tropical Cyclone
Examples of near surface wind fields retrieved from Quikscat data are depicted in
Fig. 4.1. These wind maps have been obtained from the website of Remote Sensing
Systems
4 . The four wind maps show the passage of the tropical cyclone Flavio from
the Indian Ocean towards to coast of Mozambique, East Africa. Flavio made landfall
on 22 February 2007 at the Mozambican coast near the city of Inhambane, where
it caused considerable damage. Note that the direction of rotation in the cyclone is
clockwise, which is typical for a southern-hemisphere cyclone. The scatterometer
data were acquired at 03:12 UTC on 18 February 2007 (upper left panel) at 0400 UTC
on 20 February (upper right panel), at 1548 UTC on 20 February (lower left panel),
and at 1544 UTC on 21 February (lower right panel). The data shown in panels a
and b were obtained during descending (morning) satellite passes of the satellite and
the one shown in Panels c and d during ascending (evening) passes. Such wind maps
are very useful for tracking the position of cyclones. While, in general, the wind
speeds and directions derived from Quickscat are quite reliable, this does not apply
for the inner regions of a tropical cyclone (or typhoon/hurricane), where the winds
can be quite high (above 25 m/s). As mentioned before, retrieving near-surface wind
fields with speeds above 25 m/s is questionable at all radar bands because above
this value the GMF saturates. At Ku band, the retrieval of typhoon/hurricane winds
from scatterometer data is further aggravated by the fact that typhoons/hurricanes are
associated with heavy rain. Raindrops in the atmosphere both scatter and attenuate
Ku-band microwaves and cause errors in the wind retrieval algorithm.
4 At http://www.ssmi.com/qscat/scatterometer_data_daily.html?rgn=indian_west&size=large.