Abstract
Many statistical models exist for modelling precipitation. One difficulty is that two issues need to be addressed: the probability of precipitation occurring, and then the quantity of precipitation recorded. This paper considers a family of distributions for modelling the quantity of precipitation, including those observations in which exactly no precipitation is recorded. Two examples are then discussed showing the distributions model the precipitation patterns well. Copyright © 2004 Royal Meteorological Society

This publication has 22 references indexed in Scilit: