Both a simple model and a complex mesoscale simulation model are used
to simulate and better understand the wind observations in the Kali Gandaki
valley reported in the first part of this paper. The Kali Gandaki river
originates in Nepal near Tibet, flows southward through the Mustang basin,
crosses the Himalayas in a gorge to descend to the lowlands of Nepal. Extremely
strong diurnal upvalley flow in the gorge and the basin alternates with
rather weak drainage flow in the night. Accorgin to the simple model the
Mustang basin acts as an elevated heat source during the day, where the
inflow from the south is hampered by the Himalayan barrier. This leads
to the formation of a jet at the northern end of the gorge which protrudes
into the basin. The simple model fails to fully explain both the observed
strength of the inflow and the weakness of the nocturnal drainage flow.
The mesoscale model is successful in simulating almost all aspects
of the observations. The simulations strongly suggest that the observed
acceleration of the upvalley winds near the entrance to the Mustang basin
is linked to supercritical flow. Moreover, gravity waves induced by the
ridges protruding into the valley appear to contribute to the acceleration.
Humidity is found to be essential for simulating the observed day-night
asymmetry because of its impact on the boundary layer structure above the
Himalayan foothills and on the nocturnal cooling rates. In addition, advection
of relatively stable air from the foreland into the basin is important
for the formation of the jet and explains also part of the asymmetry. The
Plateau of Tibet appears to have a small but positive impact on the flow
speeds in the valley.