SATELLITE MICROWAVE SCANNER DATA USING FOR REVEALING
OCEAN AREAS WITH MOST PROBABILITY OF TROPICAL
CYCLONES GENERATION BY CALCULATION OF TEMPERATUREHUMIDITY CRITERION
Vera V. Rostovtseva and Igor V. Goncharenko
P.P.Shirshov Institute of Oceanology RAS
117 997, 36, Nakhimovskii av., Moscow, Russia
Abstract
The problem of determination of the conditions resulting in tropical cyclones generation is of great
scientific and practical significance. The suggested by V.N.Pelevin temperature-humidity criterion
allows to estimate the ocean areas with high probability of tropical cyclone generation using values of
ocean surface temperature, water vapor amount above the ocean surface and Coriolis parameter
depended on the place latitude. All these parameters can be measured operatively by microwave
radiometry methods, satellite microwave scanner data giving the possibility to examine the distribution
of these physical parameters over different aquatoria of the World Ocean and to follow their alteration
within the day.
The program of the satellite information processing was developed for TMI scanners of TRMM
satellites system allowing to estimate the abovementioned criterion twice a day with the spatial
resolution of 25х25 km. Distribution of the suggested criterion in the tropical Atlantic was analyzed for
2002-2009, spatial and time characteristics of statistics of this distribution being revealed. Rather good
correlation between the criterion values and the frequency of tropical cyclones generation taking place
in Northern Atlantic at these time periods is shown.
As a result of daily analyses of satellite data over some aquatorium of interest during the cyclone
generation activity period some ways of further criterion development were proposed with cloudiness
taken into account. It was shown that stability of cloudiness to some extent can be the sign of still
weather that was known to be one of the necessary factors for tropical cyclone starting. Adding these
easily measured by satellite microwave scanner parameters we modified the criterion. The examples
of the criterion distribution in the tropical Atlantic were given several days before tropical cyclones
generation to demonstrate the efficiency of the developed approach.
INTRODUCTION
Tropical cyclones (TC) genesis has been widely studied using satellite, airborne, ship borne and other
data [1-4], as well as mathematical modeling [2, 5, 6]. A number of factors were proved to influence
TC development: high sea surface temperatures, low level relative vorticity and weak vertical wind
shear, presence of moist convective cloud systems, pre-existing disturbance in the troposphere, the
presence of a Tropical Upper Tropospheric Trough etc. Various indices were suggested by combining
three or more of these factors to predict TC genesis [7-9] giving the possibility to explain some
seasonal and spatial features of TC distribution. However, the development of these powerful
atmospheric phenomena is still unclear.
The approach developed by professor Pelevin [10] based on the fact that huge energy of TC (the
18
kinetic energy of TC is about 10 J [1]) is released during quite short time period (4-15 days). That
means that before a TC-starting there must be an extra energy stock in the appropriate oceanic
region. Considering water vapour as the main power source for TC development we introduced a
temperature-humidity criterion (TH-criterion) that enabled us to find such zones. It was shown that
seasonal and latitudinal distribution of the averaged criterion obtained from the climatic atlases
correlated quite well with the frequency of TC generation [10, 11]. All parameters necessary for the
TH-criterion calculation can be measured with success by microwave radiometry methods, satellite
microwave scanner data giving the possibility to examine its distribution over different aquatoria of
World Ocean and to follow its alteration within the day [12].
TEMPERATURE-HUMIDITY CRITERION CONSTRUCTION AND ITS CALCULATION FROM
SATELLITE DATA
The developed TC looks like a vertical tube in the troposphere through which the warm and moist air
from the air-layer adjacent to the ocean surface rushes upward. This vertical flux is twisted because of
Coriolis force influence and accelerated due to the angular momentum conservation law. As enormous
energy is released during quite short time-periods, it is obvious that for a TC realization the potential
energy stock is necessary which can transfer into the kinetic energy of great wind velocities. This
potential energy stock is accumulated gradually before the TC generation with the solar energy as a
source. The solar flux crosses the atmosphere almost without extinction and heats the upper layer of
the ocean. This water layer heats in its turn the adjacent air layer and makes it moist by the intensive
evaporation. As a result the lower 0.5-1 km air layer becomes warm and moist (the relative moisture
exceeds 85% at the 10-meter level above the water surface) and ready to go up.
We use the experimental data obtained by PIGAP program during the scientific voyages of r/v
“Akademik Kurchatov” and “ Professor Shtokman” in the equatorial zone (3N – 3S) to make the
mechanism of atmospheric instability generation clear. Typical vertical profile of the air temperature
averaged by more than 200 radio-sounds is used as stable tropical atmospheric stratification profile
(Fig.1). We calculated the temperature of an air sample lifted adiabatically from the ocean surface to
the tropopause, assuming the starting parameters as Ts=27C and ks=0 or ks=85%. The obtained
temperatures were compared at every height to the temperatures of the stable atmosphere: if the airsample temperature is higher than the one of the surrounding air, the sample will go upwards, in the
other case the air-sample rising will stop at this height. It is seen that the moist air with the given
temperature at the water surface is able to get to the tropopause (~ 100 mb).
Figure 1. Calculated PT-dependences for an air-sample with Ts=27C and different relative surface moisture
(ks=85% and ks=0) lifting from the ocean surface in comparison to the real atmospheric stratification curve.
Figure 2. Air temperatures and relative air moisture above the water surface which are necessary for TC generation
(zone A), the area where TC are impossible (zone B)
Having carried out calculations for various values of air temperatures and relative moisture at the sea
surface we obtained the basic conditions for the deep convection existence that is the conditions at
which the air from sea surface adjacent layer can rise to the tropopause by itself (Fig. 2).
Thus, there are mainly two factors causing the atmospheric deep convection instability: high air
temperature at the water surface (TS) and high relative moisture above water surface (kS). We assume
that the more the linear deviation  of the point characterizing the aquatorium of interest in TSkS –
coordinates from the “TC zone-bordering” curve is, the more the possibility of deep convection
instability generation is, and so the more is TC generation probability. Using linear approximation of
the “TC zone-bordering” curve for the interval of most possible TC developing temperatures (from
25.5grad to 29.5grad C, Tc = 27.5C, Tmax=4, corresponding relative moisture interval kc = 67%, 
kmax =24%), we obtain for the linear deviation (the constant multiplier is omitted):
( TS, kS ) = (TS - Tc)/ Tmax + (kS - kc)/ kmax
TC never occur on the equator though the temperature and air humidity above the water are often
within the required conditions. It occurs because of the second basic condition for TC generation that
is not fulfilled on the equator: there is no horizontal Coriolis acceleration. The latter is known to depend
on sin ( is the value of latitude of the place no matter south or north). Taking that into account we
form the TH-criterion using temperature and relative air moisture values above the ocean surface as
well as Coriolis parameter depending on the latitude:
0
for (TS, kS) ≤ 0
(TS, kS) sin
for (TS, kS) > 0
=
Here  = 0 in the areas where no TC can develop and as negative linear deviation  stands for such
areas we put 0 there. We showed [10] the suggested criterion estimated from climatic data was in
good correlation with the TC frequency that means it could be used for estimating TC forming
probability.
The daily values of air temperature and air humidity necessary for the temperature-humidity criterion
can be measured by microwave radiometry methods, satellite microwave scanner data giving the
possibility to examine the distribution of these physical parameters over different aquatoria of World
Ocean and to follow their alteration within the day. We chose the data from TRMM satellites system
obtained by TMI microwave scanner which observes the Earth from 40N to 40S twice a day with the
spatial resolution of 25х25 km. SST stand for estimates of Ts, while water vapor amount divided by the
maximal value in tropical areas stands for ks .
Examining the daily distribution of the criterion over the North Atlantic we confirmed the enhanced
values of the criterion are observed both in the areas of acting TC and in the area about three days
before a new TC starting. An example of such distribution is given on the three days before TC
Hermine (28N, 66W) and Gaston (32N, 78W) starting, their areas being clearly indicated by the
maximal values of the criterion (Fig. 3).
Figure 3. TH-criterion distribution on August, 24-27, 2004 calculated from the satellite data. The two TC are shown on
the image with clouds on the day of their starting.
COMPARISON OF TEMPERATURE-HUMIDITY CRITERION DISTRIBUTION AND TROPICAL
CYCLOGENESIS ACTIVITY
We calculate the TH-criterion distribution for two years demonstrating quite different TC generation
activity – 2004 and 2009. The chosen seasons differ from each other greatly: there were 15 cyclones
in 2004 and only 9 in 2009 with much less intensity and duration (Fig. 4). The values of TH-criterion
were averaged for every month for the North Atlantic and compared to the number of TC starting
during the same month. One can see the TH-criterion predicts the seasonal distribution of TC-genesis
as well as the variations of TC activity from year to year.
Figure 4. Monthly averaged TH-criterion for the North Atlantic calculated from the satellite data for 2004 and 2009
compared to the temporal distribution of the number of TC in the corresponding year and distribution averaged for 16
years.
Figure 5. TH-criterion averaged for the chosen zones in the North Atlantic (see the schematic drawing at the top) for
two-month term compared to the number of days including TC-starting and TC-supporting processes in the
corresponding zones related to their squares.
As August and September are the main two months for every season of TC activity (in 2004 there
were 8 TC in August and 4 in September, in 2009 - 4 and 3 correspondingly) we carried out the
calculations for these two months. For more detailed comparison we divided the North Atlantic
aquatorium in 5 zones: the southern part that includes the IZC (Intratropical Zone of Convergence) in
summer and autumn (1), the central part in the North Atlantic (2), the part at the south-east coast of
North America (3), the Gulf of Mexico (4) and the Caribbean Sea (5) (Fig. 5). For estimating TC activity
the number of TC starting in every zone is calculated for the chosen months. Then as the energy
potential of troposphere above the ocean is used both for starting and for supporting TC we calculate
the number of days for every TC to cross the zone. To compare the intensity of TC-generation process
in every zone we divide the number of TC-days in each zone by the square of the zone. The obtained
results are given in Fig. 5.
During the hurricane season of 2004 the best conditions for TC generating were in the zone 1: for the
mean criterion value of about 20 almost 7 TC days activity per the square of 10°x10° for two months
were observed. In 2009 the TH-criterion was twice as low and the amount of TC-days came down to
3,6. The effect is quite similar for the zone 2.
For the zones 3-5 the criterion also dropped twice as much but the number of TC days reduced
dramatically. That means in those zones the conditions for TC generating were different and there
were some more important factors (for example the wind distribution) besides from temperature and
humidity factors that should be taken into account. The comparison of different hurricane seasons
according to this method will give the possibility to reveal the zones and the years of such effect and to
determine the appropriate factors and modify the TH-criterion..
MODIFICATION OF TEMPERATURE-HUMIDITY CRITERION
One of the well known factors preventing TC forming is vertical wind shear. Assuming that stable
cloudiness cannot exist in the areas with strong wind shear we consider it as a additional predictor of
TC generation. Multiplying the TH-criterion by the cloudiness factor Cl, which is equal to 1 inside the
cloudy area, while in the other places it is much less than 1 (for example, 0.1). Using the modified THcriterion enables to reveal the areas of TC starting more easily. It is illustrated in Fig. 6.
Figure 6. Distribution of TH-criterion and Modified TH-criterion, which uses the factor of cloudiness, for the period of
three days before TC ISAAC starting.
CONCLUSION
The suggested temperature-humidity criterion (TH-criterion) calculated from satellite microwave
radiometry data was shown to correlate to the probability of TC starting. The seasonal and spatial
statistics of the TH-criterion distribution analyzed with comparison to the TC generation activity in the
North Atlantic for 2004 and 2009 gives the explanation to the cyclogenesis intensity variety. The daily
distribution of TH-criterion or its modification enables to determine the areas of TC forming for about
three days before its starting.
The work was carried out with the support of grant ISTC N 3827.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Golitsin G.S., (1997), Statistics and energetic of tropical cyclones, Dokladi Akademii Nauk,
354, №4, pp 535-538.
Bister M., Emanuel K.A., (1977), The Genesis of Harricane Gulliermo: TEXMEX Analyses and
a Modeling Study, Monthly Weather Review, 125, pp 2662-2682.
Knaff J.A., Zehr R.M., Kidder S.Q., (2000), An Example of Temperature Structure Differences
in Two Cyclone Systems Derived from the Advanced Microwave Sounder Unit, Weather
Forecast, 15, pp 476-483.
Mitnik L., Chen K.S.,Wang J.-T., Mitnik M., Hsu M.-K., (2002), Satellite Microwave
Observations of Typhoon Herb (1996) near Taiwan, The Global Atmosphere and Ocean
System, 8, №1, pp 19-39.
Challa M., Preffer R.l., Zhao Q., Chang S.W. (1998), Can Eddy Fluxes Serve as a Catalyst for
Hurricane and Typhoon Formation?, J. Atm. Sci., 55(12), pp 2201-2219.
Yaroshevich M.I., Ingel L.X., (2004), Tropical Cyclone as an Element of Ocean- Atmosphere
System, Dokladi Akademii Nauk, 399(3), pp 397-400.
Gray W.M., (1998), The formation of Tropical Cyclones, Meteor. Atmos. Phys. 67, pp 37-69.
Ward J.F.A., (1995), Prediction of Tropical Cyclone Formation in Terms of Sea-Surface
Temperature Vorticity and Vertical Wind Shear, Aust. Meteor . Mag., 44, pp 61-70.
DeMaria M., Knaff J.A. and Connell B.H., (2001), A Tropical Cyclone Genesis Parameter for
the Tropical Atlantic, Weather Forecast. 16, pp 219-233.
Pelevin V.N. and Rostovtseva V.V., (2004), New temperature-humidity criterion estimating the
possibility of tropical cyclone generation, Atmospheric and ocean optics, 17, №7, pp 563-568.
Atlas of the oceans. Atlantic and Indian oceans, Moscow, (1977), 306 pp.
Pelevin V.N., Rostovtseva V.V. and Goncharenko I.V., (2005), Estimation of possibility of
tropical cyclone generation on base of satellite microwave scanner data using new
temperature-humidity criterion, Issledovanie Zemli iz Kosmosa, №1, pp 1-7.