Abstract
A new kernel of a stochastie equation for snowflake aggregation with the effects of variations of falling velocities is determined by observed data of falling snowflakes (Sasyo and Matsuo, 1980).
Based on a stochastic equation with a new kernel, the snowflake aggregations in cloud with uniform elemental snow crystals are simulated without assuming an initial size distribution of snowflakes.
The following results are obtained.
i) The new kernel for the combination of colliding snowflakes is not symmetrical function with respect to the difference between their falling velocities. This suggests that j-snowflake whose mass is less than that of i-snowflake can collide with the latter at its rear. These results are different from those for the order collision in which the kernel is a symmetrical function.
ii) The variation of falling velocity reduces considerably the time required to get a specified snowfall intensity compared with the case of the order collision but is not effective for increasing snowfall intensity.
iii) The snowfall intensity is mainly determined by the initial number concentration of elemental snow crystals in a cloud.
