US
JO(ST A. DCJSINGER AND S. P. S. ARYA
Fib. 10. The boundary luycr height defined in variour ways 0s a function of stability.
using the Great Plains data, and by Clarke (1970) using the Australian
expedition data. Unfortunately, their results are markedly different with too
much scatter of data points to permit any reliable estimate of these
supposedly universal functions. In a separate study the junior author has
undertaken to resolve thew differences and to investigate the possible effects
of thermal winds and aaleratioas. An analysis of the " Wangara" data
(Clarke et al.. 1971) by him has Icd to improved estimates of the above
similarity functions. There is still considerable scatter of individual data
points and only the best fitted third degae polynomials through them are
shown in Fig. 12. Note that the agreement with our theoretical curves is
good in view of the large scatter in the original data
The results of Bobyleva er d. (1%7) are ah0 shown in Fig. 12 for comparison with the premt model. Tbe agreement between the two is very close for
B(p+), but very poor for A(#.). Our model is in better agreement with the
empirical estimates of the latter. We also point out that the velocity defect
profiles calculated by Bobylcva et al. (1%7), although quite similar to ours
in the lower layer, had very sharp kinks just b l o w the top of the boundary
laycr. This may be indicative of the failure of some d their model assumptions in stable conditions.
JO(ST A. DCJSINGER AND S. P. S. ARYA
Fib. 10. The boundary luycr height defined in variour ways 0s a function of stability.
using the Great Plains data, and by Clarke (1970) using the Australian
expedition data. Unfortunately, their results are markedly different with too
much scatter of data points to permit any reliable estimate of these
supposedly universal functions. In a separate study the junior author has
undertaken to resolve thew differences and to investigate the possible effects
of thermal winds and aaleratioas. An analysis of the " Wangara" data
(Clarke et al.. 1971) by him has Icd to improved estimates of the above
similarity functions. There is still considerable scatter of individual data
points and only the best fitted third degae polynomials through them are
shown in Fig. 12. Note that the agreement with our theoretical curves is
good in view of the large scatter in the original data
The results of Bobyleva er d. (1%7) are ah0 shown in Fig. 12 for comparison with the premt model. Tbe agreement between the two is very close for
B(p+), but very poor for A(#.). Our model is in better agreement with the
empirical estimates of the latter. We also point out that the velocity defect
profiles calculated by Bobylcva et al. (1%7), although quite similar to ours
in the lower layer, had very sharp kinks just b l o w the top of the boundary
laycr. This may be indicative of the failure of some d their model assumptions in stable conditions.