28
DAVID Y. GATES
----.-- 30% 00%
1
I0.C. 00%
1
0
1 2 3 4 5 6 7 0
Characteristic dimension 5 crn
Rodtation absorbed 0 0 cai
Diffusion resntance (internal) I0 0 scc cm-'
Wind speed, (m p h )
rnin-'
FIG. 10. Trtbmpkation rate es a function of wind speed for e leaf of chereoteristic dimension 6 om, intend diffusion reeiStance of 10.0 BBO cm-1, and 0.8 4 cm-* min-1 of
absorbad redietion at v&ou8 air temperaturea and relative humidities.
Theso diagrams merit a few other interesting comments concerning
the influence of wind. For humid regions such aa tropical forests or for
island vegetation like on Hawaii where there are often persistent trade
winds, one can speculate concerning the impact of wind. The wind may
produce less evapotranspiration than at timea without wind. However,
wind w i l l produce cooler vegetation due to convective oooling and w i l l
in addition produce more mixing of air and of carbon dioxide into a
dense canopy which is beneficial for photoeynthesis. If the wind,
through mixing, diminiahea the humidity within the canopy, compared
with still air conditions, then this could cauae an increase of evapotranspiration. It is entirely likely that becauae of compensating factors
wind will have relatively little influence on the vegetation under generally warm humid conditions.
By cDmparison, a dry desert environment, which is often windy,
would have an increaae in evapotranspiration with wind. The desert
would Lave about twice aa much water lose per unit area of vegetation
than wmld the humid tropical forest. The water loss per unit area of
ground surface would, of course, be much greater for the humid, tropioal
forest than for the desert because of the sparaity of vegetation 011 the
Précédent

- 41/297

Suivant