The Penn State-NCAR mesoscale model MM5 is used to simulate and better unde
rstand the wind observations in the Kali Gandaki Valley reported in the fir
st part of this paper. The Kali Gandaki River originates in Nepal near Tibe
t, flows southward through the Mustang Basin, crosses the Himalayas in a go
rge, and descends to the lowlands of Nepal. Extremely strong diurnal upvall
ey flow in the gorge and the basin alternates with rather weak drainage flo
w in the night. As proposed in Part I, the Mustang Basin and the Tibetan Pl
ateau can be considered as an elevated heat source driving the upvalley flo
w during the day. However, the extreme strength of the diurnal upvalley win
ds and the order-of-magnitude asymmetry between day and night cannot be exp
lained with a simple plateau circulation theory.
The model is successful in simulating almost all aspects of the observation
s. The simulations strongly suggest that the observed acceleration of the u
pvalley winds near the entrance to the Mustang Basin is linked to a supercr
itical-like flow pattern. Gravity waves induced by the ridges protruding in
to the valley appear to contribute to this flow structure. Humidity is foun
d to be essential for simulating the strength of the observed day-night asy
mmetry because of its impact on the boundary layer structure above the Hima
layan foothills, especially due to the evaporation of rain. In addition, ad
vection of relatively stable air from the foreland into the basin is import
ant for the formation of the gravity waves and also explains part of the as
ymmetry. The Plateau of Tibet appears to have a small but positive impact o
n the flow speeds in the valley.