The distribution of precipitation in time during storms constitutes a complex stochastic process. Mathematical modelling of the process is quite difficult. The process is conceived as a finite-duration, quantitized-data, continuous-variable, stochastic process that can be represented as a finite time series.
Broad outlines of the procedure have been given in an earlier paper (Chow & Ramaseshan 1965). This paper deals with the details of formulating and fitting a suitable stochastic model for the process and the steps are illustrated with an example. The conclusions derived on the basis of the model are also discussed. The methods discussed in this paper may be useful for the study of transient stochastic processes such as floods, droughts, earthquakes, squalls, etc., and for the design of engineering systems to resist or control such processes.