is much 
^ound- 
ig ab- 
1ge 
fig.2 
g water 
anch 
: di- 
ric ar- 
ng tide, 
on as 
al. If 
we would 
dynamic 
not con- 
is us- 
sidera- 
ication 
pro- 
mor- 
aver 
ing a 
with 
e of 
      
- 1157 - 
this, the distribution of the flow of water not only remáins 
unchanged, but even more evident. 
By comparing bathymetric maps, it is quite evident that 
the pattern is bounded to the bottom geometries (fig. !), in 
s ow 4 y Our 
  
   
  
  
  
  
  
   
   
   
  
    
  
Mie A 
N 4$ jf; ers ^ CENA 
M V. M Gras 
Z : i € = 
NEE, (7? : T ddl s CA' ZANE d A ; v 
seu © ) p VALLE DI CA' ZAN N e x [7 
/ 
#4 VALLE LA 
t. 
JELLA CENTREGA | 
&*. 
orenzo , V "A 
oe Motta S. Lgyenzo 
i 1 a -— M9 oo 
; 13.74 
zy ^ a 
4 * 
' PALUDE 3 f 
Circulation of rising tide. Shaded areas indi- 
eate "hummoek" areas; dotted line represents the 
channels; and the black arrows indicate the main 
flow pattern. 
fact the lines, very pronounced by harmonic analysis, followthe 
eourse of the smaller, natural canals. loreover, these areas 
which are slowly submerged by the tide are detectable because 
they present themselves as an obstacle to the flow pattern 
eausing a slight thermal perturbation on the upper surface. 
So even if the well-known infra'ed technique only surveys 
the phenomena on their surfaces, it seems to be able to givean 
analysis in a vertical sense. Of course this fact is caused by 
the drift of the electrical signal due to the elaboration used. 
Since the thermal signal was used in a non thermal way, the boun 
dary layer effect must be confined to a shallow area and short 
distances. It is for this reason, we are not concerned with 
the amplitude of a thermal signal. To contain the contrast be- 
tween the information from ground and water data, an ulterior 
logarithmic compression is relayed by signals.