The hydrologic regime of rivers and streams is a major determinant of habitat quality for fish and aquatic invertebrates. Long-term streamflow data were compiled and multidecadal streamflow trends and ecological flow (EFlow) statistics were calculated in support of the United States Geological Survey (USGS) Chesapeake Bay Science Initiative toward understanding fish habitat and health in the Chesapeake Bay Watershed (CBWS). A dataset comprising all streamgages (n = 409) reporting daily means of streamflow within the CBWS and remaining active as of September 30, 2018 (the end of Water Year [WY] 2018), independent of streamgage installation date, was retrieved from the USGS National Water Information System (NWIS). This dataset was then subset to include only those streamgages with a contiguous timeseries of streamflow data from a start date no earlier than April 1, 1939 (Climate Year [CY] 1940) and no later than October 1, 1999 (WY 2000). The R packages “EGRET” and "Eflowstats" were utilized together to determine streamflow trends and EFlow statistics from the subset (n = 243). Trends and EFlows were computed for the ranges 1940-1969 (n = 90), 1970-1999 (n = 167), and 2000-2018 (n = 243). Streamflow trends were computed for eight annual metrics (1-, 7- and 30-day minima [CY] and maxima [WY], mean and median [WYs]). These streamflow trends provide context for the 178 EFlow statistics (WY) which have been designated to characterize the magnitude, frequency, and duration of extreme high and low flows, the timing of seasonal flows, and the consistency of the historic regime. Files herein include the following Child Items: (1) a table summarizing streamflow trends for three time periods at a minimum of 90 and maximum of 243 streamgages and 500 time-series plots graphically representing those trends; (2) a table summarizing EFlow statistics and the change between each statistic for three time periods at a minimum of 90 and maximum of 243 streamgages; and (3) a GIS shapefile of the original 409 USGS streamgage locations, complete with NWIS attributes, active within the CBWS through September 30, 2018.

Ecological flow (EFlow) statistics have been designated to characterize the magnitude, frequency, and duration of extreme high- and low-flows, the timing of seasonal flows, and the consistency of the historic regime. This Child Item contains a table of 178 EFlows for the time periods 1940-1969, 1970-1999, and 2000-2018, with absolute and percent change between periods, where applicable. Statistics were computed by Water Year (WY) for all 178 metrics and absolute and percent change were calculated by comparing metrics between combinations of two of the three time periods (1940-1969 and 1970-1999; 1940-1969 and 2000-2018; 1970-1999 and 2000-2018). Streamgages from the original dataset (n = 409) were excluded from one or more time periods of analysis because of extensive data gaps that would yield incomplete EFlows; therefore, stations were indexed into the earliest possible time period relative to their installation date (for example, a streamgage with an operating start year of 1958 would be included in the analysis for the time periods 1970-1999 and 2000-2018), which resulted in different sample sizes for each period: 1940-1969 (n = 90), 1970-1999 (n = 167), and 2000-2018 (n = 243). Similarly, multiple stations were wholly excluded because of frequent discontinuities in the daily mean streamflow through all three time periods. Finally, a streamgage must have fallen within at least two time periods to have a change value. As such, not all stations are represented in the change analysis (change between 1940-1969 and 1970-1999 [n = 90]; change between 1940-1969 and 2000-2018 [n = 90]; change between 1970-1999 and 2000-2018 [n = 167]).

Ecological flow (EFlow) statistics have been designated to characterize the magnitude, frequency, and duration of extreme high- and low-flows, the timing of seasonal flows, and the consistency of the historic regime. This Child Item contains a table of 178 EFlows for the time periods 1940-1969, 1970-1999, and 2000-2018, with absolute and percent change between periods, where applicable. Statistics were computed by Water Year (WY) for all 178 metrics and absolute and percent change were calculated by comparing metrics between combinations of two of the three time periods (1940-1969 and 1970-1999; 1940-1969 and 2000-2018; 1970-1999 and 2000-2018). Streamgages from the original dataset (n = 409) were excluded from one or more time periods of analysis because of extensive data gaps that would yield incomplete EFlows; therefore, stations were indexed into the earliest possible time period relative to their installation date (for example, a streamgage with an operating start year of 1958 would be included in the analysis for the time periods 1970-1999 and 2000-2018), which resulted in different sample sizes for each period: 1940-1969 (n = 90), 1970-1999 (n = 167), and 2000-2018 (n = 243). Similarly, multiple stations were wholly excluded because of frequent discontinuities in the daily mean streamflow through all three time periods. Finally, a streamgage must have fallen within at least two time periods to have a change value. As such, not all stations are represented in the change analysis (change between 1940-1969 and 1970-1999 [n = 90]; change between 1940-1969 and 2000-2018 [n = 90]; change between 1970-1999 and 2000-2018 [n = 167]).

This Child Item contains (1) a table of trends in eight annual streamflow statistics for the time periods 1940-1969, 1970-1999, and 2000-2018, and (2) a .zip file of plots of the eight statistical trends for the three time periods. The annual streamflow statistics and their trends were computed by Climate Year (CY) for 1-,7- and 30-day minima and by Water Year (WY) for 1-,7- and 30-day maxima, median and mean. Streamgages from the original dataset (n = 409) were excluded from one or more time periods of analysis because of extensive data gaps that would interrupt trend lines; therefore, streamgages were indexed into the earliest possible time period relative to their installation date (for example, a streamgage with an operating start year of 1958 would be included in the analysis for the time periods 1970-1999 and 2000-2018), which resulted in different sample sizes for each period: 1940-1969 (n = 90), 1970-1999 (n = 167), and 2000-2018 (n = 243). Similarly, multiple streamgages were wholly excluded because of frequent discontinuities in the daily mean streamflow used to derive the annual mean streamflow through all three time periods.

This Child Item contains (1) a table of trends in eight annual streamflow statistics for the time periods 1940-1969, 1970-1999, and 2000-2018, and (2) a .zip file of plots of the eight statistical trends for the three time periods. The annual streamflow statistics and their trends were computed by Climate Year (CY) for 1-,7- and 30-day minima and by Water Year (WY) for 1-,7- and 30-day maxima, median and mean. Streamgages from the original dataset (n = 409) were excluded from one or more time periods of analysis because of extensive data gaps that would interrupt trend lines; therefore, streamgages were indexed into the earliest possible time period relative to their installation date (for example, a streamgage with an operating start year of 1958 would be included in the analysis for the time periods 1970-1999 and 2000-2018), which resulted in different sample sizes for each period: 1940-1969 (n = 90), 1970-1999 (n = 167), and 2000-2018 (n = 243). Similarly, multiple streamgages were wholly excluded because of frequent discontinuities in the daily mean streamflow used to derive the annual mean streamflow through all three time periods.

This data set contains U.S. Geological Survey (USGS) streamgage identification numbers, begin and end years of the periods of streamflow record tested, Sen slope trends in the annual minimum 7-day streamflow for the period of record tested, the p-values (significance) of the trends, and the trend Sen slopes standardized by the standard deviations of the residual errors defined as the difference between observations and the Sen slope lines, for 174 USGS streamgages with 56 to 75 years of record in the Chesapeake Bay watershed, Mid-Atlantic U.S.

This data set contains U.S. Geological Survey (USGS) streamgage identification numbers, begin and end years of the periods of streamflow record tested, Sen slope trends in the annual minimum 7-day streamflow for the period of record tested, the p-values (significance) of the trends, and the trend Sen slopes standardized by the standard deviations of the residual errors defined as the difference between observations and the Sen slope lines, for 174 USGS streamgages with 56 to 75 years of record in the Chesapeake Bay watershed, Mid-Atlantic U.S.

The Chesapeake Bay is greatly affected by freshwater flows from streams and rivers draining its watershed. Variations in the amount and timing of streamflow can change water temperature, salinity, suspended sediment and contaminants levels, and affect the amount of nutrients in freshwater streams. Additionally freshwater flows affect the amount of in-stream habitat available to freshwater flora and fauna. Metrics characterizing streamflow extremes such as magnitude, frequency and duration of high and low flows are helpful indicators of long-term patterns, shifts in streamflow regimes, and the interpretation of their effects on stream biota. Generalized additive models were used to analyze long term changes in annual metrics of streamflow extrema from water year 1985 through water year 2022. Detailed data preparation information, analytical methods and results are presented and discussed in the associated Scientific Investigative Report (https://doi.org/10.3133/sir20255072).

This U.S. Geological Survey (USGS) data release contains estimated daily streamflow and base flow for HUC12 in the nontidal areas of the Chesapeake Bay watershed, monthly average streamflow and base flow, flow statistics, MATLAB scripts, and a document that describes how to create similar datasets in other watersheds. Daily streamflow was estimated for all the nontidal parts of the Chesapeake Bay watershed with the program "Unit Flows in Networks of Channels" (UFINCH; Holtschlag, 2016), together with the observations of measured streamflow at gages at the downstream ends of major rivers. The estimated streamflow was aggregated at the HUC12 level and reformatted as an Optimal Hydrograph Separation (OHS) input file using MATLAB scripts. Base flow was calculated at each HUC12 outlet using the base flow index (BFI) hydrograph separation methods developed by Wahl and Wahl (Wahl and Wahl, 1988; Wahl and Wahl, 1995) and by Eckhardt (Eckhardt, 2005) with the parameter estimation method developed by Collischonn and Fan (Collischonn and Fan, 2013) which are incorporated into the OHS program (Raffensperger and others, 2017). This data release supports the following publication: • Buffington, P.C., and Capel, P.D., 2020, Estimating streamflow and base flow within the nontidal Chesapeake Bay riverine system: U.S. Geological Survey Scientific Investigations Report 2020-5055, 26 p., https://doi.org/10.3133/sir20205055. References cited: • Collischonn, W. and Fan, F.M., 2013, Defining parameters for Eckhardt's digital baseflow filter: Hydrological Processes, v. 27, no. 18, p. 2614-2622, https://doi.org/10.1002/hyp.9391. • Eckhardt, K., 2005, How to construct recursive digital filters for baseflow separation: Hydrological Processes, v. 19, no. 2, p. 507-515, https://doi.org/10.1002/hyp.5675. • Holtschlag, D.J., 2016, UFINCH-A method for simulating unit and daily flows in networks of channels described by NHDPlus using continuous flow data at U.S. Geological Survey streamgages: U.S. Geological Survey Scientific Investigations Report 2016-5074, 17 p., https://doi.org/10.3133/sir20165074. • Raffensperger, J.P., Baker, A.C., Blomquist, J.D., and Hopple, J.A., 2017, Optimal hydrograph separation using a recursive digital filter constrained by chemical mass balance, with application to selected Chesapeake Bay watersheds: U.S. Geological Survey Scientific Investigations Report 2017-5034, 51 p., https://doi.org/10.3133/sir20175034. • Wahl, K.L., and Wahl, T.L., 1988, Effects of regional ground water declines on streamflows in the Oklahoma Panhandle, in Symposium on Water-Use Data for Water Resources Management, Tucson, Arizona, American Water Resources Association, p. 239-249. • Wahl, K.L., and Wahl, T.L., 1995, Determining the flow of Comal Springs at New Braunfels, Texas, Texas Water '95: San Antonio, Texas, American Society of Civil Engineers, p. 77-86, http://www.usbr.gov/tsc/techreferences/hydraulics_lab/pubs/PAP/PAP-0708.pdf.

This U.S. Geological Survey (USGS) data release contains estimated daily streamflow and base flow for HUC12 in the nontidal areas of the Chesapeake Bay watershed, monthly average streamflow and base flow, flow statistics, MATLAB scripts, and a document that describes how to create similar datasets in other watersheds. Daily streamflow was estimated for all the nontidal parts of the Chesapeake Bay watershed with the program "Unit Flows in Networks of Channels" (UFINCH; Holtschlag, 2016), together with the observations of measured streamflow at gages at the downstream ends of major rivers. The estimated streamflow was aggregated at the HUC12 level and reformatted as an Optimal Hydrograph Separation (OHS) input file using MATLAB scripts. Base flow was calculated at each HUC12 outlet using the base flow index (BFI) hydrograph separation methods developed by Wahl and Wahl (Wahl and Wahl, 1988; Wahl and Wahl, 1995) and by Eckhardt (Eckhardt, 2005) with the parameter estimation method developed by Collischonn and Fan (Collischonn and Fan, 2013) which are incorporated into the OHS program (Raffensperger and others, 2017). This data release supports the following publication: • Buffington, P.C., and Capel, P.D., 2020, Estimating streamflow and base flow within the nontidal Chesapeake Bay riverine system: U.S. Geological Survey Scientific Investigations Report 2020-5055, 26 p., https://doi.org/10.3133/sir20205055. References cited: • Collischonn, W. and Fan, F.M., 2013, Defining parameters for Eckhardt's digital baseflow filter: Hydrological Processes, v. 27, no. 18, p. 2614-2622, https://doi.org/10.1002/hyp.9391. • Eckhardt, K., 2005, How to construct recursive digital filters for baseflow separation: Hydrological Processes, v. 19, no. 2, p. 507-515, https://doi.org/10.1002/hyp.5675. • Holtschlag, D.J., 2016, UFINCH-A method for simulating unit and daily flows in networks of channels described by NHDPlus using continuous flow data at U.S. Geological Survey streamgages: U.S. Geological Survey Scientific Investigations Report 2016-5074, 17 p., https://doi.org/10.3133/sir20165074. • Raffensperger, J.P., Baker, A.C., Blomquist, J.D., and Hopple, J.A., 2017, Optimal hydrograph separation using a recursive digital filter constrained by chemical mass balance, with application to selected Chesapeake Bay watersheds: U.S. Geological Survey Scientific Investigations Report 2017-5034, 51 p., https://doi.org/10.3133/sir20175034. • Wahl, K.L., and Wahl, T.L., 1988, Effects of regional ground water declines on streamflows in the Oklahoma Panhandle, in Symposium on Water-Use Data for Water Resources Management, Tucson, Arizona, American Water Resources Association, p. 239-249. • Wahl, K.L., and Wahl, T.L., 1995, Determining the flow of Comal Springs at New Braunfels, Texas, Texas Water '95: San Antonio, Texas, American Society of Civil Engineers, p. 77-86, http://www.usbr.gov/tsc/techreferences/hydraulics_lab/pubs/PAP/PAP-0708.pdf.