IAPRS & SIS, Vol.34, Part 7, "Resource and Environmental Monitoring", Hyderabad, India,2002
RADIATIVE-TEXTURE BASED DELINEATION OF OCEANIC HYDROMETEORS IN
MICROWAVE RADIOMETRIC IMAGES FROM TRMM-TMI, DMSP-SSM/I AND
IRS-P4/MSMR
B.S. Gohil, B.G. Vasudevan, A.K. Mathur and Vijay K. Agarwal
Oceanic Sciences Division, Meteorology & Oceanography Group, Space Applications Centre (ISRO),
Ahmedabad- 380015, India, E-mail: bsgohil 9 yahoo.com
KEY WORDS: TMI, SSM/I, MSMR, oceanic hydrometeors, radiative-texture, rain classification
ABSTRACT:
An approach is presented to delineate hydrometeors (clouds and precipitation combined) in brightness temperature images from
microwave radiometers launched onboard TRMM, DMSP and IRS-P4 satellites through a new parameter termed as hydrometeor
index (HI). The HI is obtained from two independent parameters, one is a radiative parameter obtained by taking channel average of
inverse normalized polarized brightness temperature difference (AINPD), and the other is a physical texture parameter which is a
spatial gradient (GRD) of AINPD. Only low frequency channels at 19, 21 or 22 and 37 GHz have been used in case of TRMM and
DMSP while in the case of IRS-P4, only 18 and 21 GHz channels have been used for the purpose. Significant correspondence
between the TMI data derived HI and the TMI finished products of rain rate and cloud liquid water have been observed. Moreover,
attempt has also been made to use AINPD and GRD parameters in classifying the hydrometeor index representative of precipitation
in three categories, viz. statiform, convective and mixed. Analysis of TMI, SSMI-I and MSMR data over Orissa super cyclone in
Bay of Bengal during October 27-28, 1999 has been presented. Cloud and precipitation features of the cyclone have been well
brought out in TMI data by using the present approach. SSM-I data also depicts the desired features, while MSMR data shows the
same but with less detail due to its relatively coarser spatial resolutions and the channels being less sensitive to the clouds and
precipitation as compared to 37 GHz channel. The study indicates the potential of present approach in identifying the clouds and
precipitation features required for screening and retrievals.
1. INTRODUCTION
Cloud and precipitation play important role as heating source
for the atmospheric circulation affecting global weather and
climate. Precipitation is also an important exchange process in
hydrological cycle. Satellite sensors operating in visible,
infrared and microwave regions of electromagnetic spectrum
are used for remote sensing of clouds and precipitation. Visible
and infrared imageries for precipitation measurements have
been extensively used globally and have their own advantages
and limitations. Microwave remote sensing of clouds and
precipitation has significantly advanced with the launch of state
of art microwave radiometers like SSM/I and TMI launched
onboard DMSP and TRMM satellites and microwave
precipitation radar launched onboard TRMM satellite
(Kummerow and Weinman, 1988; Kummerow et al, 2001;
Olson, et al 2001; Spencer et al, 1989; Bauer and Schluessel,
1993; and several other). Extraction of geophysical parameters
from microwave radiometric measurements is quite complex
particularly precipitation. Prior knowledge of hydrometeors is
very useful in quantitative estimation of precipitation and in
comparison of various rain retrieval algorithms (Ferraro et al,
1998; Hong et al, 1999; Olson et al, 2001; Kummerow et al,
2001). Unlike visible and infrared radiation scattered by
hydrometeors (clouds and precipitation), the microwave
radiations, depending upon the frequency, are affected by
hydrometeors through the processes of absorption, scattering or
by both. Thus, the threshold techniques for the detection of
cloud and precipitation is less effective in the case of
microwave measurements due to the reasons that the same
amount of cloud or same intensity of rain at different altitudes,
different latitudinal regions of the globe and with different
textural characteristics has different microwave signatures. In
this respect, a different approach is needed for the detection of
clouds and precipitation in the microwave radiometric images.
Several approaches have been developed based on either
radiative or physical or both the properties of clouds and rain
(Ferraro et al, 1998; Hong et al, 1999; Olson et al, 2001;
Kummerow et al, 2001 and others) for the detection of clouds
and precipitation using SSM-I and TMI data
In the present study, attempt has been made to qualitatively
interpret microwave radiometric measurements in terms of
radiative and physical texture properties of clouds and
precipitation in view of its usefulness in screening (Grody,
1991) and retrievals of clouds and precipitation. The approach
used here determines a hydrometeor index based on radiative
and spatial distribution properties of clouds and precipitation
and by using only low frequency channels like 19, 21 and 37
GHz of TMI and 19, 22 and 37 GHz of SSM/I. In the case of
microwave radiometer (MSMR) launched onboard Indian
satellite IRS-P4, only 18 and 21 GHz dual polarized channels
out of 6, 10, 18 and 21 GHz channels have been used for the
purpose. Efforts have also been made to classify the
hydrometeor index for different rain situations. Variation of HI
with TMI rain and cloud finished products have been examined
to extend its application to SSM-I and IRS-P4/MSMR data.
2. ALGORITHMS
The parameters used in the present algorithms for detecting
clouds and precipitation can be explained with the help of
expression for total brightness temperature emitted by the earth
and atmosphere under non-scattering condition as given below
TB(f,0,p) = T(f,0).T,.[1-r,(f,0,p)] + T(F,0).[T(F,0).TB,(A) +
TBan(f,0)].r,(f,0,p) + TBup(£,0)
IN
TB
TB
Thi
use