Applications of Passive and Active Microwave Sensors ...
465
We begin with a multispectral study of stratus clouds which employs the combination of cloud
liquid water estimates from microwave data with coincident visible and infrared measurements
(Miletta and Katsaros, 1995). We then describe research on midlatitude cyclones which began
before the launch of SSM/I emphasizing a recent study where the SSM/I signals were analyzed for their predictive value in identifying rapidly deepening cyclones off the U.S. east coast
(McMurdie and Katsaros, 1996). This study is followed by a look at Polar Lows with numerous
satellite systems. These cyclonic storms develop rapidly over oceanic regions where data other
than those obtained by satellite are rare (Claud et al., 1993 and 1994).
Finally, we present the combined use of active and passive microwave systems to study tropical
storms. Strong convergence lines in the surface wind field and very complex wave fields have
been identified as well as the role of organized precipitation patterns surrounding the eye acting
as precursors for storm intensification (Quilfen et al., 1993 and MS Thesis; Zhao, 1994). The
chapter ends with conclusions and an outlook on future work.
17.2 The Instruments
17.2.1 The active microwave instrument: dcatterometer and SAR
in wave mode
The European Remote Sensing Satellite 1, ERS-l, carries a complex radar system called the
Active Microwave Instrument (AMI) operating at a frequency of 5.3 GHz (C- band). It functions
in two mutually exclusive modes, either as a Wind Scatterometer, using three antennas or as a
Synthetic Aperture Radar (SAR), using a very large (10m by 1m) antenna, as in a configuration
illustrated by Figure 17.1. The SAR feature can be called upon during the scatterometer data
collection for an instant, while the central scatterometer antenna would have been sampling.
The SAR pixel is an area 5 km by 10 kmjust inside the scatterometer swath which is illuminated
by the radar during these bursts.
The resulting SAR image is called an 'imagette' and the data collection operation is called 'SAR
in wave mode'. (We do not discuss here regular SAR images 100 km by 100 km in size. These
are obtained with the scatterometer mode turned off for reasons of power constraints.) The
three antennas used for the AMI operated as a scatterometer generate radar beams looking 90°
from the flight track and 45° fore and aft. Due to the variations in the power returned from the
three looks at overlapping footprints, about 50 km by 50 km in size, wind direction as well as
wind speed are determined. The three looks are obtained at different incidence angles and with
a slight time delay. The backscattered power normalized by the emitted power form a ratio
termed the normalized radar cross-section. These triplets of information for each footprint are
entered into a mathematical algorithm to solve for the wind speed and direction. A certain
ambiguity in the wind direction remains, since up and down-wind returns differ only slightly,
often modified by signal noise, variability in the sea state or varibility in the wind over the footprint. To select one of the ambiguous wind vectors current methods use a 'first guess' approach
based on numerical weather prediction models. The data products provide information on the
multiple solutions to the users (ERS-l System, 1992). The French Processing and Archiving
Facility, CERSAT, distributes reprocessed AMI scatterometer data on CD-ROMS (CERSAT
NEWS, Issue 5, 1995). The wind is perhaps the most important parameter controlling air-sea
interactions. Scatterometry is one of four methods to observe the surface wind over the sea from
space. Table 17.1 lists the methods, the characteristics of current systems and the principles
behind the wind related signal.
465
We begin with a multispectral study of stratus clouds which employs the combination of cloud
liquid water estimates from microwave data with coincident visible and infrared measurements
(Miletta and Katsaros, 1995). We then describe research on midlatitude cyclones which began
before the launch of SSM/I emphasizing a recent study where the SSM/I signals were analyzed for their predictive value in identifying rapidly deepening cyclones off the U.S. east coast
(McMurdie and Katsaros, 1996). This study is followed by a look at Polar Lows with numerous
satellite systems. These cyclonic storms develop rapidly over oceanic regions where data other
than those obtained by satellite are rare (Claud et al., 1993 and 1994).
Finally, we present the combined use of active and passive microwave systems to study tropical
storms. Strong convergence lines in the surface wind field and very complex wave fields have
been identified as well as the role of organized precipitation patterns surrounding the eye acting
as precursors for storm intensification (Quilfen et al., 1993 and MS Thesis; Zhao, 1994). The
chapter ends with conclusions and an outlook on future work.
17.2 The Instruments
17.2.1 The active microwave instrument: dcatterometer and SAR
in wave mode
The European Remote Sensing Satellite 1, ERS-l, carries a complex radar system called the
Active Microwave Instrument (AMI) operating at a frequency of 5.3 GHz (C- band). It functions
in two mutually exclusive modes, either as a Wind Scatterometer, using three antennas or as a
Synthetic Aperture Radar (SAR), using a very large (10m by 1m) antenna, as in a configuration
illustrated by Figure 17.1. The SAR feature can be called upon during the scatterometer data
collection for an instant, while the central scatterometer antenna would have been sampling.
The SAR pixel is an area 5 km by 10 kmjust inside the scatterometer swath which is illuminated
by the radar during these bursts.
The resulting SAR image is called an 'imagette' and the data collection operation is called 'SAR
in wave mode'. (We do not discuss here regular SAR images 100 km by 100 km in size. These
are obtained with the scatterometer mode turned off for reasons of power constraints.) The
three antennas used for the AMI operated as a scatterometer generate radar beams looking 90°
from the flight track and 45° fore and aft. Due to the variations in the power returned from the
three looks at overlapping footprints, about 50 km by 50 km in size, wind direction as well as
wind speed are determined. The three looks are obtained at different incidence angles and with
a slight time delay. The backscattered power normalized by the emitted power form a ratio
termed the normalized radar cross-section. These triplets of information for each footprint are
entered into a mathematical algorithm to solve for the wind speed and direction. A certain
ambiguity in the wind direction remains, since up and down-wind returns differ only slightly,
often modified by signal noise, variability in the sea state or varibility in the wind over the footprint. To select one of the ambiguous wind vectors current methods use a 'first guess' approach
based on numerical weather prediction models. The data products provide information on the
multiple solutions to the users (ERS-l System, 1992). The French Processing and Archiving
Facility, CERSAT, distributes reprocessed AMI scatterometer data on CD-ROMS (CERSAT
NEWS, Issue 5, 1995). The wind is perhaps the most important parameter controlling air-sea
interactions. Scatterometry is one of four methods to observe the surface wind over the sea from
space. Table 17.1 lists the methods, the characteristics of current systems and the principles
behind the wind related signal.