AbstractThere is general consensus that climate is undergoing change but whether climate change is occurring or not is still being debated in certain scientific, political, and religious quarters. Hydrologic variability influences the design of civil works and assessment of long-term climate change would help improve design criteria. To this end, long-term variability of streamflow was estimated using Shannon entropy. Three statistical tests were applied to determine trends in annual and seasonal daily streamflow with 5% two-sided confidence limit. Daily streamflow data spanning 70 years (from 1943 to 2012) from 669 stream gauge stations located in 23 states in the northeastern part of United States of America, covering six different water regions were employed. The time variability of annual and seasonal daily streamflow was assessed using the Mean Decadal Apportionment Disorder Index (MDADI). Analysis showed that in all cases minimum and maximum streamflows had higher variability than average and median streamflows. A significant number of stations exhibited trends. Considering annual minimum, average and median daily streamflows, approximately 50% of the stations followed trends and for almost all these stations trends were increasing. Only for annual maximum daily streamflow, 15% of the stations showed increasing trend and 10% decreasing trend. In terms of geographical distribution, the stations with increasing trend were essentially located along the Atlantic coast and near Great Lakes and in the Upper Mississippi Water Region. Similar considerations apply for seasonal time series as well.