The amount of daily rainfall for a solitary wet day (Class 0) is normally less than the rainfall for a
wet day that is enclosed by wet days on either side (Class 2). The same characteristics hold true for
the case of the first or last day of wet spells (Class 1) and there fore the rainfall here is found to be
smaller than the Class 2 rainfall. In many regions, there exist large differences between these rainfall
classes. In this study, the daily rainfall data from 5 rain gauge stations in the north and northwest of
Iran are classified to classes 0, 1 and 2 according to the number of adjoining wet days. To examine
the impact of adjoining wet days on the distribution of rainfall amounts, different distributions were
tested with 3 different data sets: (1) all rainfall classes grouped together, (2) solitary wet days fitted
separately from other wet days, and finally (3) all rainfall classes separately. The results showed that
the models which treat the rainfall classes separately are a better fit for all the 5 stations. The Akaike
Information Criterion (AIC) values are significantly reduced as the rainfall data are split into rainfall
classes. Also, among 4 different distributions considered (Gamma, Lognormal, Mixed Exponential
and Weibul), mixed exponential distribution in most cases and lognormal distribution in one case
showed better performances.