Tropical Cyclogenesis and Ecoinformatics Methods - New Ecoinformatics Tools in Environmental Science: Applications and Decision-making - page 629

percolation cluster, it is necessary to establish the algorithm for the calculation of
the critical level of the renormalization probability p taking into consideration the
procedure of renormalization group transformations for more than two dimensions.
Only in this case the real-space renormalization process can make more precise the
moment of I m (t) transition from background zone

ʞ 1 to the signal zone ʞ 3

(Fig. 7.17 ).
Thus, the following conclusions are drawn:

(1) An application of percolation procedure gives the possibility to detect the
moment when the ocean-atmosphere system switches its phase state from the
background position to the state with high level of environmental instability.
(2) An introduced indicator of the ocean-atmosphere system instability allows the
detection of the beginning of a hurricane before its detection by existing
monitoring systems.
(3) An increase of the precision of the proposed algorithm is possible with the
optimization of parametrical structure in the indicator I m (t).
It is evident that the solution to the problem of reliable prediction of the
beginning of a hurricane requires the development of an ef

cient information
technology which must be used in the environmental monitoring systems. This
technology should include sections responsible for planning of measurements,
development of algorithms for complex processing of data from different sources,
and relevant risk assessment. The main dif culty in this lies in the spatial and

temporal heterogeneity of existing information obtained from different observa-
tional systems. This heterogeneity can be overcome by using the algorithms of the
space-temporal interpolation (Kondratayev et al. 2006; Krapivin and Varotsos
2007). Ultimately, however, the proof of the usefulness of a prediction algorithm
such as this one can only really be established by attempting to use it , i.e. in

real time, to make predictions of hurricanes in their initial stages. The serious task
consists in the prediction of environmental disaster under minimal false alarm. To a
considerable extent it is a matter of the successful de

“ live ”

nition of thresholds.
Thus, the structural scheme of the multi-channel monitoring system for search
and detection of the hurricanes in their beginning stage can be presented by the
scheme of Figs. 3.16 and 7.13 . One of these systems can be realized on the existing
meteorological stations located in the zones where hurricanes or typhoons are
beginning. Here Ti i and

* ,
respectively. Meteorological and geophysical data delivered by satellite sensors and
NDBC meteorological stations are processing for the calculation of I m or

ʾ i is the deterministic and stochastic components of I m or ʻ

* that are
analyzed in the framework of sequential procedure described in Chap. 3 . In fact, the
picture of Fig. 7.30 is realized at all times. The calculation of probabilities that
characterize the prognosis level for the anomaly moving direction (see Figs. 7.25
and 7.26 ) gives a possibility to have a cyclone track. This procedure can operate in
automatic mode.

ʻ