An Analysis of Two Schemes to Numerically Solve the Stochastic Collection Growth Equation

Abstract

Two schemes for the numerical solution of the stochastic collection growth equation for cloud drops are compared. Their numerical approaches are different. One (the Berry/Reinhardt method) emphasizes accuracy; the other (the Bleck method) emphasizes speed. Our analysis shows that for applications where the number of solutions (time steps) does not exceed 104 the accuracy-oriented scheme is faster. For larger, repetitive applications, such as a comprehensive cloud model, an objective analysis can be made on the merits of exchanging accuracy for computational time.