Revision History for SIR 2014-5177 Thomas M. Over, James D. Riley, Mackenzie K. Marti, Jennifer B. Sharpe, and Donald V. Arvin Estimation of Regional Flow-Duration Curves for Indiana and Illinois ----------------------------------------------------- Posted online October 29, 2014 ----------------------------------------------------- Revised and reposted April 5, 2022, version 2.0 The text version was modified. Author Mackenzie K. Marti was added. Limited minor changes unrelated to the substantial changes described below were made due to isolated errors within the manuscript (for example, misspelled words). Substantial changes to the original manuscript (posted online October 29, 2014) 1. Stream density was a basin characteristic used in the regression equations for IN region 2 and IL region 3 in the original manuscript. After publication, this basin characteristic was determined to be geographically inconsistent. New equations using a different combination of basin characteristics were developed for IN region 2 and IL region 3, resulting in changes to text, tables (1, 2, 3, 5, 9, 13, 14), and figures (3, 4, 5) for those two regions. Changes to values are small enough that changes to resulting figures are difficult or impossible to see. 2. All Illinois basin characteristic values have been updated to reflect values computed using delineations derived from high resolution elevation and stream network data. The original manuscript used high resolution delineations to compute basin characteristic data for Indiana but medium resolution delineations to compute basin characteristic data for Illinois. High resolution elevation and stream network data are being added to StreamStats for Illinois in the near future, and as such, the basin characteristic values have been updated in the revisions. This resulted in minor changes to all basin characteristic values for Illinois, which also minorly affected the values of regression coefficients, intercepts, and so on for all regions in Illinois. Resulting changes were implemented in relevant text, tables (3, 5, 11, 12, 13, 14, 17, 18), and figures (3, 5). Changes to values are small enough that changes to resulting figures are difficult or impossible to see. 3. Three stations were added to table 5, the Illinois flow statistics table. These three stations were included in developing regional regression equations for their respective regions in the original manuscript but were accidentally omitted from table 5. They have been correctly added to this table in the revisions, and minor changes to the text were also a result. Changes made to the manuscript On the cover page, “Prepared in cooperation with the U.S. Environmental Protection Agency, Region V” was revised to “Prepared in cooperation with the U.S. Environmental Protection Agency, Region V, and the Indiana Department of Environmental Management” On the title page, the authors line was revised to add Mackenzie K. Marti as the third author. On the page following the title page, the “Cover image” text was revised from “…Station 05554000, North Fork Vermilion River, Illinois, at low flow. Photo taken by Teresa Halfar, USGS Illinois Water Science Center…” to “…station 05554000, North Fork Vermilion River near Charlotte, Illinois, at low flow. Photo taken by Teresa Halfar, U.S. Geological Survey…” On page i, “Prepared in cooperation with the U.S. Environmental Protection Agency, Region V” was revised to “Prepared in cooperation with the U.S. Environmental Protection Agency, Region V, and the Indiana Department of Environmental Management” On page v, a sentence was added at the end of the first paragraph of the “Acknowledgments” section: “Funding for the revisions leading to version 2.0 was provided by the Indiana Department of Environmental Management.” On page 1, in the second paragraph of the “Abstract” section, the value 740 was changed to 745. On page 3, in the sentence following the Drain.Index equation of the “Computing Basin Characteristics” section, the text was changed from “…where the drainage descriptors were computed with ArcHydro Tools in ArcMap from State Soil Geographic…” to “…where the fractions are decimal fractions and were computed with ArcHydro Tools in ArcMap from the “dominant” condition as specified in the…” On page 3, in the QSS_PermB equation of the “Computing Basin Characteristics” section, “1*” was added in front of “fraction fine-grained stratified sediment” and “fraction glacial till”. On page 3, in the first sentence following the QSS_PermB equation of the “Computing Basin Characteristics” section, “the fractions are decimal fractions and” was added before the text “the weights,” and “(100, 1, and 0.1)” was added following the text “the weights”. On page 3, at the end of the last paragraph of the “Computing Basin Characteristics” section, the following text was added: "Notice the weights in equation 2c (25, 75, and so on) are the midpoints of the thickness intervals; because the thickness intervals are in units of feet, these weights are as well, and thus QSS_Thick also has units of feet." On page 3, in table 1, “stream density” was removed as an example basin characteristic for Morphometric data and “Soil permeability” was removed as an example basin characteristic for soil properties data. Another row was inserted in the table for a third “Soil properties” entry. This entry has Source: "Soils data derived from STATSGO (Wolock, 1997)," WWW reference: "https://doi.org/10.3133/ofr97656," and Example basin characteristics: "Soil permeability." On page 4, in the third sentence of the last paragraph of the “Selection and Testing of Streamgage Records and Computation of Flow-Duration Curves” section, “171” was changed to “174”, “128” was changed to “131”, “157” was changed to “159”, and “92” was changed to “91”. On page 5, in figure 1, the regional boundary in Shelby and Bartholomew Counties was modified. The original map used the incorrect boundary. On page 6, in figure 2, a segment at the northern part of the Rock River (a major river), was missing from the original figure, so a flowline was added in southern Wisconsin (streamgage 05430500 location). In the same area, a minor river, Turtle Creek (flowline immediately to the east, streamgage 05431486 location), was present but not needed and so was removed. On page 7, at the end the last paragraph of the “Defining Regions” section, and continuing onto page 8, the following text was added: "For the purpose of the analyses performed for this study and presented in tables 2-5, the region of a streamgage was defined as the region that contained most of the streamgage drainage basin, which is not necessarily the location of the streamgage itself. The regional boundaries shown in figures 1 and 2 were drawn to follow major basin divides to minimize the number of streamgages with basins in more than one region but eliminating all multiregion basins was not possible for the larger drainage basins.” On page 8, at the end of the first paragraph of the “Regression” section, the text “TIBCO Spotfire S+ 8.1 for Windows statistical analysis software package (TIBCO Software, Inc., 2008).” was replaced with “survival R package (Therneau, 2021).” On page 10, in the table 14 headnote, “NWI, National Wetlands Inventory; cm, centimeters;” was removed. “QSS_PermB, index of permeability of Quaternary surface sediments;” was added. On page 10, in table 14, the range of exceedance probabilities for the first two instances of IN 2 were changed from “99.9-40” to “99.9-20”, and the third and fourth instance were changed from “30-0.1” to “10-0.1”. The “Drainage area” values for all Illinois regions (IL 1, IL 2, and IL 3) were updated to reflect new basin characteristic values for all exceedance probabilities. The “Centroid longitude” values for IL 2, exceedance probabilities 40-0.1, were updated to reflect new basin characteristic values. The “Stream density” column was removed. The “Mar-Apr-May” column was removed. “Sep-Oct-Nov” values were added for IL 3 (exceedance probabilities 99.9-0.1). On page 11, in table 14, the “Dec-Jan-Feb” values for IN 2 (exceedance probabilities 99.9-20) were added. The “NWI palustrine emergent wetlands” column was removed. The “QSS_PermB” column was added, with values for IN 2 (exceedance probabilities 99.9-20). The “Available water content of soil” values were updated for the IL 2 minimum (exceedance probabilities 99.9-95). The “Soil drainage index” values were updated for IL 1 minimum (exceedance probabilities 99.9-0.1). The “STATSGO soil permeability” values were updated IL 2 (exceedance probabilities 90-50 and 40-0.1). The “PermBXThick” values were updated for IL 1 (exceedance probabilities 99.9-95 and 90-50). On page 11, in the second sentence of the first paragraph of the “Basin Characteristics and their Coefficient Values” section, “2.7” was replaced with “2.8” On page 12, in figure 3, in panel A: Illinois region lines were changed due to high resolution basin characteristics update. The changes were very small, resulting in minimal to no visual change. In panel B: minor changes in Indiana region 2 (purple) and Illinois region 3 (orange) lines were due to revised equations. Very small changes were made to Illinois region 1 and Illinois region 2 due to high resolution basin characteristics update, resulting in minimal to no visual change. On page 13, in the first sentence of the third paragraph of the “Basin Characteristics and their Coefficient Values” section, the text “stream density (Str. Den),”, “March-April-May,”, and “an indicator of the abundance of palustrine wetlands (NWI.PEM)” were removed. The comma after November was removed. In the same sentence, following “...STAT.Perm),” the following text was added: "an index of Quaternary surface sediment permeability (QSS_PermB),". On page 13, in the second sentence of the third paragraph of the “Basin Characteristics and their Coefficient Values” section, the text “moderate to” was removed. On page 13, in the third sentence of the “Basin Characteristics and their Coefficient Values” section, following the text “…coefficient value in these equations” the text “switches from positive to negative” was replaced by “is negative”. In the same sentence, the two following pieces of text were removed: “as the flows increase,” and “, whereas the increased infiltration would allow for increased flows at higher exceedance probabilities.”. On page 13, in the last sentence of the third paragraph of the “Basin Characteristics and their Coefficient Values” section, the text “Similar to Indiana region 2, the” was revised to read “The”. The following text was added to the end of the same sentence: " which agrees with the Indiana region 2 coefficient at high flows and indicates that the higher infiltration indicated by higher permeability tends to increase the magnitude of lower flows.” On page 13, the fourth paragraph of the “Basin Characteristics and their Coefficient Values” section that discussed stream density was removed. On page 14, in the ninth sentence of the fifth paragraph of the “Basin Characteristics and their Coefficient Values” section, the text “a positive coefficient for precipitation during this period seems” was modified to read "the sign and trend in magnitude of this coefficient seem". On page 14, between the ninth and tenth sentence of the fifth paragraph of the “Basin Characteristics and their Coefficient Values” section, the following text was added "The equations for all exceedance probabilities in Illinois region 3 also use SON.Precip, in addition to drainage area. The SON.Precip coefficients have a trend that is opposite of that of Indiana region 3, decreasing from small positive values to larger negative values with increasing exceedance probability. Based on the geographic distribution of fall precipitation in Illinois, which increases to the south (not shown), this behavior indicates that SON.Precip is acting in this equation as a proxy for another variable that depends on geographic location. Because southern Illinois has generally less permeable soils and therefore less infiltration, recharge, and base flow (Singh, 1971) but larger SON.Precip values, the coefficient of SON.Precip is increasingly negative with higher exceedance probability.” On page 14, in the last sentence of the fifth paragraph of the “Basin Characteristics and their Coefficient Values” section, “decreasing” was replaced with “increasing”. On page 14, the following text was added to the end of the fifth paragraph of the “Basin Characteristics and their Coefficient Values” section: "The low- to moderate-flow quantiles for Indiana region 2 also use DJF.Precip, with the coefficient decreasing with increasing exceedance probability, indicating that the magnitude of lower flows is decreased by more winter precipitation. This apparently anomalous behavior, similar to that of SON.Precip in Illinois region 3, indicates that winter precipitation is acting as a proxy for some other basin property. Also appearing in the low- to moderate-flow quantiles for Indiana region 2 is the index of Quaternary sediment permeability (QSS_PermB), the coefficients of which increase with increasing exceedance probability, indicating that more permeable sediments increase the magnitude of low flows.” On page 14, in the fourth sentence in the sixth paragraph of the “Basin Characteristics and their Coefficient Values” section, the text “the magnitude of” was added. On page 14, in the second sentence of the first paragraph of the “Accuracy of Estimation Equations” section, the text “around of” was replaced with “of around”. On page 14, in the last sentence of the first paragraph of the “Accuracy of Estimation Equations”, “740” was replaced with “745”. On page 15, in figure 4, panel B: Indiana region 2 multiple regression line (orange) was changed due to new equation, resulting in minimal visual change. On page 16, in figure 5, panels A, B, and C: small changes were made in all multiple regression lines due to switch to high resolution data, resulting in minimal to no visual change. In panel C: changes to the multiple regression line were made due to equation revision in Illinois region 3, resulting in minimal visual changes. On page 18, in the first sentence of the third paragraph of the “Example Application” section, the text “(http://water.usgs.gov/osw/streamstats/illinois.html)” was replaced by "(https://streamstats.usgs.gov/ss/, accessed 2014)". On page 18, in the third sentence of the third paragraph of the “Example Application” section, the text “PermBXTHick” was replaced with “PermBXThick”. On page 18, in the third sentence of the fourth paragraph of the “Example Application” section, “67.02” was replaced by “67.0”. In the same sentence, “5.637” was replaced with “-5.361”. On page 18, in the equation within paragraph four of the “Example Application” section, the sum “5.637” was replaced with “-5.361” and “2.108” was replaced with “2.106”. In the same equation, the text “log10(67.02)” was replaced with “log10(67.0)“ and “-1.787” was replaced with “-1.782”. On page 18, in the last sentence of the fourth paragraph of the “Example Application” section, the text “Q (subscript 0.99)=10 (superscript -1.787)=0.0163” was replaced with ”Q (subscript 0.99)=10 superscript) -1.782)=0.0165”. On page 18 in the fifth paragraph of the “Example Application” section, within the equation where the numerical values were inserted, the text “-10.24” was replaced with “-10.251”, “1.925” was replaced with “1.926”, “log (subscript 10) (67.02)” was replaced with “log (subscript 10) (67.0)”, and “1.627” was replaced with “1.631”. On page 19, in table 17, the Station identifier “DSPA-01 (05544300)” was updated to read “DSPA-01 (05554300)”. Additionally, the basin characteristics in the 6th row, “05554500 (values used to compute regression equations in this study)”, were updated to reflect new basin characteristic values. On page 20, in the fifth sentence of the sixth paragraph of the “Example Application” section, “492” was replace with “491”. On page 20, in the sixth sentence of the sixth paragraph of the “Example Application” section, the text “-1.787 ± 0.780 and -3.096 ± 0.593.” was modified to read “-1.782 ± 0.780 and -3.096 ± 0.594.” On page 20, in the seventh sentence of the sixth paragraph of the “Example Application” section, the text “Because -1.787 – 0.780 = -2.567 and -3.096 + 0.593 = -2.567...” was modified to read “Because -1.782 – 0.780 = -2.562 and -3.096 + 0.594 = -2.502,”. On page 20, in the eighth sentence of the sixth paragraph of the “Example Application” section, the text “515” was replaced with “514”, “595” was replaced with “593”, “20.5” was replaced with “20.2”, and “13.9” was replaced with “13.7”. In the same sentence, the text “2.712 ± 0.088 and 2.774 ± 0.060” was modified to read “2.711 ± 0.087 and 2.773 ± 0.059”. On page 20, in the last sentence of the sixth paragraph of the “Example Application” section, the text “595/515 = 1.155” was replaced with “593/514 = 1.154” and “2.712 + 0.088 = 2.800 and 2.774 - 0.060 = 2.714,” was replaced with “2.711 + 0.087 = 2.799 and 2.773 - 0.059 = 2.715,”. On page 20, in the first sentence of the second paragraph under the “Summary” section, “740” was replaced with “745”. On page 21, in figure 7, in panels A and B: all lines were changed due to updated equation in Illinois region 2 as a result to the switch to high resolution basin characteristics. Minimal to no visual change in figure 7. On page 24, the “Tasker, G.D., 1997…” reference was removed. On page 24, the “TIBCO Software Inc., 2008…” reference was removed. On page 24, the reference "Therneau, T., 2021, A Package for Survival Analysis in R: R package version 3.1-12, accessed August 21, 2020, at https://CRAN.R-project.org/package=survival." was added. On page 24, the "Wolock, D.M., 1997, STATSGO soil characteristics for the conterminous United States: U.S. Geological Survey Open-File Report 97-656, accessed September 15, 2020, at https://doi.org/10.3133/ofr97656. " reference was added. Table changes In table 2, the headnote was updated to include QSS_PermB and the "stream density" column was removed. The QSS_PermB column was added. Reference “Schwarz and Alexander, 1995” was removed from the footnotes and reference “Wolock, 1997” was added to the footnotes. In table 3, the headnote was updated to included QSS_PermB. NWI was removed from the headnote, the "stream density" column was removed. The NWI column was removed. Reference "Cowardin and others, 1979" was removed from the footnotes and “Wolock, 1997” was added to the footnotes. The QSS_PermB column was added. All values in the table were updated. In table 5, values were updated in the "drainage area" column. Three rows were added for stations 05442000, 05445500, and 05448000. In table 9, the headnote was updated and values in the "fraction of correct censoring estimates," "MSR," "RMSR," "MSR as percent," "intercept i," "SE of intercept," "coefficient a (subscript 1)," "SE of coefficient a (subscript 1)," "basin characteristic A (subscript 2)," "coefficient a (subscript 2)," "SE of coefficient a (subscript 2)," "basin characteristic A (subscript 3)," "coefficient a (subscript 3)," and "SE of coefficient a (subscript 3)" columns were updated. The footnotes were revised. In table 11, values in the "MSR," "RMSR," "MSR as percent," "intercept i," "SE of intercept," "coefficient a (subscript 1)," "SE of coefficient a (subscript 1)," "coefficient a (subscript 2)," and "SE of coefficient a (subscript 2)" columns were updated. In table 12, values in the "fraction of correct censoring estimates," "MSR," "RMSR," "MSR as percent," "intercept i," "SE of intercept," "basin characteristic A (subscript 1)," "coefficient a (subscript 1)," "SE of coefficient a (subscript 1)," "coefficient a (subscript 2)," "SE of coefficient a (subscript 2)," "coefficient a (subscript 3)," and "SE of coefficient a (subscript 3)" columns were updated. In table 13, the headnote was updated and values in the "fraction of correct censoring estimates," "MSR," "RMSR," "MSR as percent," "intercept i," "SE of intercept," "coefficient a (subscript 1)," "SE of coefficient a (subscript 1)," "basin characteristic A (subscript 2)," "coefficient a (subscript 2)," and "SE of coefficient a (subscript 2)," columns were updated.