Abstract
Streamflow simulation is usable in water resources design for the purpose of extracting pertinent information from streamflows at recorder locations and using that information to generate additional values for those locations, and values for additional locations, that could as reasonably occur in the future as could repetition of past events. An electronic computer procedure is described that extracts a maximum degree of pertinent information from monthly streamflow data and generates values whose statistical characteristics are consistent with those of the observed monthly streamflow. Multiple linear regression of the data transformed to unit normal standard deviates is used. A procedure is described for using data that are not simultaneous at all recorder locations and that would permit formulating an interrelated-streamflow generator for ungaged locations.

This publication has 0 references indexed in Scilit: