Precipitation and streamflow trends in Michigan, USA

Document Type


Publication Date



Department of Civil, Environmental, and Geospatial Engineering


Precipitation and streamflow trends may be changing due to changing climate. Therefore, data for these in the state of Michigan, USA, are examined through the use of statistical methods. Data from 117 precipitation stations were used, along with data from 143 streamflow gages. Data time periods varied among the stations with the longest record dating back to 1901. These methods include the linear regression best-fit line for the whole data set and also for before and after a two-sample change point analysis, moving mean, and moving standard deviation. It was found that mean precipitation for 90% of the locations and mean streamflow for 76% of the locations increased over the period of record. The moving standard deviation for precipitation increased for 54% of the locations, while 28% of the streamflow locations had an increase. Values of precipitation P(T ≤ t) two-tail, precipitation linear regression slope, and streamflow P(T ≤ t) two-tail at a 0.05 significance level occur in concentrated regions. 97% of the precipitation data sets and 92% of the streamflow data sets exhibited a distinct change. These results have implications for future management of flood control, recreation, water supply, and irrigation.

Publication Title

Sustainable Water Resources Management