Regression equations, which may be used to estimate flood flows at select annual exceedance probabilities, were developed for ungaged streams in Pennsylvania. The equations were developed using annual peak flow data through water year 2015 and basin characteristics for 285 streamflow gaging stations across Pennsylvania and surrounding states. The streamgages included active and discontinued continuous-record stations, as well as crest-stage partial-record stations, and required a minimum of 10 years of annual peak streamflow data for inclusion in the study. Explanatory variables significant at the 95-percent confidence level for one or more regression equations included the following basin characteristics: drainage area, maximum basin elevation, mean basin slope, percent storage, and the percentage of carbonate bedrock within a basin. The State was divided into five regions, and regional regression equations were developed to estimate flood flows associated with the 50-, 20-, 10-, 4-, 2-, 1-, 0.5-, and 0.2-percent annual exceedance probabilities (which correspond to the 2-, 5-, 10-, 25-, 50-, 100-, 200-, and 500-year recurrence intervals, respectively). Although the regression equations can be used to estimate the magnitude of flood flows for most streams in the State, they are not valid for streams with drainage areas generally greater than 1,500 square miles or with substantial regulation, diversion, or mining activity within the basin. The regional regression equations will be incorporated into the U.S. Geological Survey StreamStats application (https://water.usgs.gov/osw/streamstats/). Additionally, annual peak flow data for 356 streamgages initially considered for inclusion in the analysis for development of updated flood-flow regression equations were analyzed for the existence of trends; estimates of flood-flow magnitude and frequency were also computed for these streamgages. Estimates of flood-flow magnitude and frequency for streamgages substantially affected by upstream regulation are also presented. The data release contains PeakFQ input and output files (.psf, .txt, .prt) for select regulated (non-reference) and unregulated (reference) streamgages in Pennsylvania and border areas of adjacent states. The version 2.0 revision includes a dataset representing the five regions across Pennsylvania related to the estimation of flood flows by means of regression equations at ungaged streams not subject to substantial flow regulation, diversion, or mining activity. NOTE: The previous version is available from the author; all of the data in the previous version can be found in version 2.0. First Posted - 2019 (available from author) Revised - August 2020 (version 2.0)

Dataset is a model archive containing all relevant files to create regression models that are discussed in the report: Koltun, G.F., 2019, Flood-frequency estimates for Ohio streamgages based on data through water year 2015 and techniques for estimating flood-frequency characteristics of rural, unregulated Ohio streams: U.S. Geological Survey Scientific Investigations Report 2019-5018. WREG R source code, regression results and associated output files are also included in the archive.

Streamflow was collected at various streamgages in western Pennsylvania in support of the scientific investigations report "Estimation of Base Flow on Ungaged, Periodically Measured Streams in Small Watersheds in Western Pennsylvania". Data observed at the streamgages for the period of August 1, 2014 through March 31, 2017 are considered. This dataset includes 1) all data used to develop prediction intervals for the titled estimation site based upon the titled index streamgage for the period of May 1, 2015 to March 31, 2017 and 2) evaluation of data observed before May 1, 2015. For prediction interval development, a Move.1 regression was developed between the titled estimation site and titled index streamgage. From this regression model, Pearson's correlation coefficient (r) is reported, and the Nash-Sutcliffe efficiency value is computed. The prediction interval development is included for the 95% prediction interval for both the nonparametric and parametric methods. For the parametric method, any prediction interval can be developed by changing Student's t critical value. Any data that were observed before May 1, 2015 were evaluated to see if they fell within the 95% nonparametric prediction interval. The data chosen for the regression development and evaluation is streamflow that is not influenced by runoff. Streamflow that occurred the day of precipitation and two days after is considered influenced by runoff and not included in the dataset. Not all streamgages have the same period of record; dates where data were observed both at the titled index streamgage and estimation site are described below in the "Time Period of Data" section.

Streamflow was collected at various streamgages in western Pennsylvania in support of the scientific investigations report "Estimation of Base Flow on Ungaged, Periodically Measured Streams in Small Watersheds in Western Pennsylvania". Data observed at the streamgages for the period of August 1, 2014 through March 31, 2017 are considered. This dataset includes 1) all data used to develop prediction intervals for the titled estimation site based upon the titled index streamgage for the period of May 1, 2015 to March 31, 2017 and 2) evaluation of data observed before May 1, 2015. For prediction interval development, a Move.1 regression was developed between the titled estimation site and titled index streamgage. From this regression model, Pearson's correlation coefficient (r) is reported, and the Nash-Sutcliffe efficiency value is computed. The prediction interval development is included for the 95% prediction interval for both the nonparametric and parametric methods. For the parametric method, any prediction interval can be developed by changing Student's t critical value. Any data that were observed before May 1, 2015 were evaluated to see if they fell within the 95% nonparametric prediction interval. The data chosen for the regression development and evaluation is streamflow that is not influenced by runoff. Streamflow that occurred the day of precipitation and two days after is considered influenced by runoff and not included in the dataset. Not all streamgages have the same period of record; dates where data were observed both at the titled index streamgage and estimation site are described below in the "Time Period of Data" section.

Streamflow was collected at various streamgages in western Pennsylvania in support of the scientific investigations report "Estimation of Base Flow on Ungaged, Periodically Measured Streams in Small Watersheds in Western Pennsylvania". Data observed at the streamgages for the period of August 1, 2014 through March 31, 2017 are considered. This dataset includes 1) all data used to develop prediction intervals for the titled estimation site based upon the titled index streamgage for the period of May 1, 2015 to March 31, 2017 and 2) evaluation of data observed before May 1, 2015. For prediction interval development, a Move.1 regression was developed between the titled estimation site and titled index streamgage. From this regression model, Pearson's correlation coefficient (r) is reported, and the Nash-Sutcliffe efficiency value is computed. The prediction interval development is included for the 95% prediction interval for both the nonparametric and parametric methods. For the parametric method, any prediction interval can be developed by changing Student's t critical value. Any data that were observed before May 1, 2015 were evaluated to see if they fell within the 95% nonparametric prediction interval. The data chosen for the regression development and evaluation is streamflow that is not influenced by runoff. Streamflow that occurred the day of precipitation and two days after is considered influenced by runoff and not included in the dataset. Not all streamgages have the same period of record; dates where data were observed both at the titled index streamgage and estimation site are described below in the "Time Period of Data" section.

Streamflow was collected at various streamgages in western Pennsylvania in support of the scientific investigations report "Estimation of Base Flow on Ungaged, Periodically Measured Streams in Small Watersheds in Western Pennsylvania". Data observed at the streamgages for the period of August 1, 2014 through March 31, 2017 are considered. This dataset includes 1) all data used to develop prediction intervals for the titled estimation site based upon the titled index streamgage for the period of May 1, 2015 to March 31, 2017 and 2) evaluation of data observed before May 1, 2015. For prediction interval development, a Move.1 regression was developed between the titled estimation site and titled index streamgage. From this regression model, Pearson's correlation coefficient (r) is reported, and the Nash-Sutcliffe efficiency value is computed. The prediction interval development is included for the 95% prediction interval for both the nonparametric and parametric methods. For the parametric method, any prediction interval can be developed by changing Student's t critical value. Any data that were observed before May 1, 2015 were evaluated to see if they fell within the 95% nonparametric prediction interval. The data chosen for the regression development and evaluation is streamflow that is not influenced by runoff. Streamflow that occurred the day of precipitation and two days after is considered influenced by runoff and not included in the dataset. Not all streamgages have the same period of record; dates where data were observed both at the titled index streamgage and estimation site are described below in the "Time Period of Data" section.

Streamflow was collected at various streamgages in western Pennsylvania in support of the scientific investigations report "Estimation of Base Flow on Ungaged, Periodically Measured Streams in Small Watersheds in Western Pennsylvania". Data observed at the streamgages for the period of August 1, 2014 through March 31, 2017 are considered. This dataset includes 1) all data used to develop prediction intervals for the titled estimation site based upon the titled index streamgage for the period of May 1, 2015 to March 31, 2017 and 2) evaluation of data observed before May 1, 2015. For prediction interval development, a Move.1 regression was developed between the titled estimation site and titled index streamgage. From this regression model, Pearson's correlation coefficient (r) is reported, and the Nash-Sutcliffe efficiency value is computed. The prediction interval development is included for the 95% prediction interval for both the nonparametric and parametric methods. For the parametric method, any prediction interval can be developed by changing Student's t critical value. Any data that were observed before May 1, 2015 were evaluated to see if they fell within the 95% nonparametric prediction interval. The data chosen for the regression development and evaluation is streamflow that is not influenced by runoff. Streamflow that occurred the day of precipitation and two days after is considered influenced by runoff and not included in the dataset. Not all streamgages have the same period of record; dates where data were observed both at the titled index streamgage and estimation site are described below in the "Time Period of Data" section.

Streamflow was collected at various streamgages in western Pennsylvania in support of the scientific investigations report "Estimation of Base Flow on Ungaged, Periodically Measured Streams in Small Watersheds in Western Pennsylvania". Data observed at the streamgages for the period of August 1, 2014 through March 31, 2017 are considered. This dataset includes 1) all data used to develop prediction intervals for the titled estimation site based upon the titled index streamgage for the period of May 1, 2015 to March 31, 2017 and 2) evaluation of data observed before May 1, 2015. For prediction interval development, a Move.1 regression was developed between the titled estimation site and titled index streamgage. From this regression model, Pearson's correlation coefficient (r) is reported, and the Nash-Sutcliffe efficiency value is computed. The prediction interval development is included for the 95% prediction interval for both the nonparametric and parametric methods. For the parametric method, any prediction interval can be developed by changing Student's t critical value. Any data that were observed before May 1, 2015 were evaluated to see if they fell within the 95% nonparametric prediction interval. The data chosen for the regression development and evaluation is streamflow that is not influenced by runoff. Streamflow that occurred the day of precipitation and two days after is considered influenced by runoff and not included in the dataset. Not all streamgages have the same period of record; dates where data were observed both at the titled index streamgage and estimation site are described below in the "Time Period of Data" section.

Streamflow was collected at various streamgages in western Pennsylvania in support of the scientific investigations report "Estimation of Base Flow on Ungaged, Periodically Measured Streams in Small Watersheds in Western Pennsylvania". Data observed at the streamgages for the period of August 1, 2014 through March 31, 2017 are considered. This dataset includes 1) all data used to develop prediction intervals for the titled estimation site based upon the titled index streamgage for the period of May 1, 2015 to March 31, 2017 and 2) evaluation of data observed before May 1, 2015. For prediction interval development, a Move.1 regression was developed between the titled estimation site and titled index streamgage. From this regression model, Pearson's correlation coefficient (r) is reported, and the Nash-Sutcliffe efficiency value is computed. The prediction interval development is included for the 95% prediction interval for both the nonparametric and parametric methods. For the parametric method, any prediction interval can be developed by changing Student's t critical value. Any data that were observed before May 1, 2015 were evaluated to see if they fell within the 95% nonparametric prediction interval. The data chosen for the regression development and evaluation is streamflow that is not influenced by runoff. Streamflow that occurred the day of precipitation and two days after is considered influenced by runoff and not included in the dataset. Not all streamgages have the same period of record; dates where data were observed both at the titled index streamgage and estimation site are described below in the "Time Period of Data" section.

Streamflow was collected at various streamgages in western Pennsylvania in support of the scientific investigations report "Estimation of Base Flow on Ungaged, Periodically Measured Streams in Small Watersheds in Western Pennsylvania". Data observed at the streamgages for the period of August 1, 2014 through March 31, 2017 are considered. This dataset includes 1) all data used to develop prediction intervals for the titled estimation site based upon the titled index streamgage for the period of May 1, 2015 to March 31, 2017 and 2) evaluation of data observed before May 1, 2015. For prediction interval development, a Move.1 regression was developed between the titled estimation site and titled index streamgage. From this regression model, Pearson's correlation coefficient (r) is reported, and the Nash-Sutcliffe efficiency value is computed. The prediction interval development is included for the 95% prediction interval for both the nonparametric and parametric methods. For the parametric method, any prediction interval can be developed by changing Student's t critical value. Any data that were observed before May 1, 2015 were evaluated to see if they fell within the 95% nonparametric prediction interval. The data chosen for the regression development and evaluation is streamflow that is not influenced by runoff. Streamflow that occurred the day of precipitation and two days after is considered influenced by runoff and not included in the dataset. Not all streamgages have the same period of record; dates where data were observed both at the titled index streamgage and estimation site are described below in the "Time Period of Data" section.