Assessing Hurricane Intensity Using Satellites - Satellite-based Applications on Climate Change - page 154

Microwave Imager (TMI), and Advanced Microwave Scanning Radiometer-EOS
(AMSR-E). As described above, channels from the AMSU-B sounder can also be
used for imagery applications. The microwave imager instruments have much higher
spatial resolution than the sounder data and provide detailed information about cloud
and rain structure below the cloud top. Figure 10.6 shows an 85-GHz microwave
image from SSMI for Hurricane Celia in the eastern North Pacific and the
corresponding color-enhanced GOES IR (channel 4) image. The microwave imagery
shows more clearly the cloud organization below the cirrus canopy in the IR image.
Despite the usefulness of microwave imagery for qualitative interpretation of
tropical cyclone structure, a quantitative algorithm for intensity estimation with
sufficient accuracy has yet to be developed. Bankert and Tag ( 2002 ) described a
microwave imagery technique based on a nearest neighbor approach that could be
fully automated. However, the average errors were about twice as large as those
from the Dvorak technique. The use of microwave imagery for TC intensity
estimation remains an area of active research.
10.4 Other Wind Estimation Techniques
Wind estimates from satellites have application to many phenomena in addition to
tropical cyclones. Wind vectors can be estimated in the atmosphere by tracking
features in subsequent images from geostationary satellites. This technique can also
be applied at high latitudes with polar-orbiting imagery, since the temporal coverage
from those satellites is much higher there (Dworak and Key 2009 ). Development of
feature track wind algorithms began shortly after the availability of the first satellite
observations, and improvements continue to be made (e.g., Velden and Bekda 2009 ).
Surface winds over the ocean can also be estimated from scatterometers on LEO
satellites. These techniques are described in more detail in Chap. 8 . The scatterometer
winds have proved to be very useful for tropical cyclone intensity and structure
analysis.
The AMSU retrieval technique described in Sect. 10.3 can be generalized to
provide three-dimensional estimates of the horizontal wind. For that application,
the gradient wind balance assumption is replaced by the more general nonlinear
balance equation (Bessho et al. 2006 ). The three-dimensional AMSU winds are
useful for TC structure analysis and are also being applied to other phenomena,
including the atmospheric response to the Gulf Stream (O'Neill et al. 2010 ).
The three wind estimate techniques (feature track winds, scatterometer winds,
and generalized AMSU retrievals) can all be used for tropical cyclone intensity
estimation. However, low-level feature track winds are usually not available near
the center of tropical cyclones due to the cirrus canopy. The scatterometer signal
can attenuate at very high wind speeds, and there are some complications due to
rain contamination and the footprint size. The generalized AMSU winds also have a
limitation due to the instrument resolution. Thus, these techniques have application
to TC intensity estimation, but generally can provide a lower bound estimate, and
are used in combination with other information.