Chemistry Reference
In-Depth Information
In the field, experimental problems arise from the fact that the net heat
flux density at the water surface cannot be artificially controlled and thus
only the surface temperature is directly accessible with the infrared camera
and not the skin-bulk temperature difference required for estimating the
transfer coefficient for heat. Practical implementations use the statistical
properties of the sea surface temperature distribution assuming a surface
renewal model (Schimpf et al. 1999) in order to estimate the skin-bulk
temperature difference.
5 Results
Two conditions were realized in the wind wave facility: in the first set of
experiments the interface was clean, whereas in the second set a surfactant
(Triton-X-100, concentration: 3 ppm) was used. Image sequences of dura-
tion of 1 second (60 frames at 60 Hz) were recorded every 10 seconds for a
time period of 50 minutes. This procedure was repeated at four different
wind speeds. Using the controllable air ventilation system in the wind
wave facility, the latent heat flux was switched on and off every 5 minutes.
The net loss of heat at the water surface was determined by the rate of
change of the overall water bulk temperature (Schimpf 2000).
Figure 6 shows the mean surface temperature (left) and variance of tem-
perature (right) of the imaged footprint at the water surface versus time.
When there is no heat exchange between water and air phase (closed air
circulation mode), the temperature increases and temperature fluctuations
at the surface no longer occur. The infrared camera detects the bulk tem-
perature. The variance of the surface temperature is of the order of the
noise level of the infrared imager. Immediately after the latent heat flux is
switched on (open air circulation mode), the surface temperature drops
down a few tenths of a degree and the patterns of near surface turbulence
appear in the images. The variance in the temperature increases and is sig-
nificantly above the noise level of the infrared imager (26 mK). The infra-
red camera now detects the surface temperature.
If a surfactant is present, the values of the skin-bulk temperature differ-
ence are higher than for a clean surface (Fig. 7, right). This is due to the
fact that near surface turbulent mixing is generally dampened in presence
of a surfactant. The values of the skin-bulk temperature difference also
show a significant trend, decreasing rapidly with increasing wind speed, in
line with the formation of capillary waves (Schimpf 2000). If the interface
is clean, the values of the temperature gradient all are below 0.1 Kelvin
and they do not alter significantly as the wind speed increases. This is the
result of two competing effects at higher wind speed: enhanced turbulent
