Loading...
Please wait, while we are loading the content...
Similar Documents
Determination of baseline periods of record for selected streamflow-gaging stations in and near Oklahoma for use in modeling applications
| Content Provider | Semantic Scholar |
|---|---|
| Author | Esralew, Rachel A. |
| Copyright Year | 2010 |
| Abstract | Use of historical streamflow data from a least-altered period of record can be used in calibration of various modeling applications that are used to characterize least-altered flow and predict the effects of proposed streamflow alteration. This information can be used to enhance water-resources planning. A baseline period of record was determined for selected streamflow-gaging stations that can be used as a calibration dataset for modeling applications. The baseline period of record was defined as a period that is least-altered by anthropogenic activity and has sufficient streamflow record length to represent extreme climate variability. Streamflow data from 171 stations in and near Oklahoma with a minimum of 10 complete water years of daily streamflow record through water year 2007 and drainage areas that were less than 2,500 square miles were considered for use in the baseline period analysis. The first step to determine the least-altered period of record was to evaluate station information by using previous publications, historical station record notes, and information gathered from oral and written communication with hydrographers familiar with selected stations. The second step was to indentify stations that had substantial effects from upstream regulation by evaluating the location and extent of dams in the drainage basin. The third step was (a) the analysis of annual hydrographs and included visual hydrograph analysis for selected stations with 20 or more years of streamflow record, (b) analysis of covariance of double-mass curves, and (c) Kendall's tau trend analysis to detect statistically significant trends in base flow, runoff, total flow, and base-flow index related to anthropogenic activity for selected stations with 15 or more years of streamflow record. A preliminary least-altered period of record for each stream was identified by removing the period of streamflow record when streams were substantially affected by anthropogenic activity. After streamflow record was removed from designation as a least-altered period, stations that did not have at least 10 years of remaining continuous streamflow record were considered to have an insufficient baseline period for modeling applications. An optimum minimum period of record was determined for each of the least-altered periods for each station to ensure a sufficient streamflow record length to provide a representative sample of annual climate variability. An optimum minimum period of 10 years or more was evaluated by analyzing the variability of annual precipitation for selected 5-, 10-, 15-, 25-, and 35-year periods for each of 20 climate divisions that contained stations used in the baseline period analysis. The distribution of annual precipitation was compared for each consecutive overlapping 5-year period to the period 1925–2007 by using a Wilcoxon rank-sum test. The least-altered period of record for stations was also compared to the period 1925–2007 by using a Wilcoxon rank-sum test. The results of this analysis were used to determine how many years of annual precipitation data were needed for the selected period to be statistically similar to the distribution of annual precipitation data for a long-term period, 1925–2007. Minimum optimum periods ranged from 10 to 35 years and varied by climate division. A final baseline period was determined for 111 stations that had a baseline period of at least 10 years of continuous streamflow record after the record-elimination process. A suitable baseline period of record for use in modeling applications could not be identified for 58 of the initial 171 stations because of substantial anthropogenic alteration of the stream or drainage basin and for 2 stations because the least-altered period of record was not representative of annual climate variability. The baseline period for each station was rated “excellent”, “good”, “fair”, “poor”, or “no baseline period.” This rating was based on a qualitative evaluation of the approximate degree of basin alteration for the least-altered period of record, and whether or not the least-altered period was long enough to be representative of long-term climate variability. Baseline periods of record were rated as “excellent” for 22 stations, “good” for 42 stations, “fair” for 24 stations, and “poor” for 23 stations. Introduction Managing rivers and streams to maintain physical, chemical, and biological integrity is a challenge for resource Determination of Baseline Periods of Record for Selected Streamflow-Gaging Stations in and near Oklahoma for use in Modeling Applications By Rachel A. Esralew 2 Determination of Baseline Periods of Record for Selected Streamflow-Gaging Stations in and near Oklahoma managers in Oklahoma and nationwide. Knowledge of how anthropogenic activity and climate change affect the flow regime can be helpful for water supply regulation and planning. Currently (2010), the Oklahoma 50-Year Comprehensive Water Plan requires development of stream water allocation models to predict how proposed alterations may affect water supply for consumptive and nonconsumptive uses (Robert S. Fabian, Oklahoma Water Resources Board, oral and written commun., July 2008). These models are part of the process for development of a water-permitting policy for Oklahoma. The use of historical streamflow data from a least-altered period of record can enhance water resources planning. These data can be used to calibrate various modeling applications that are used to characterize unregulated flow and predict the effects of proposed streamflow alteration. Streamflow data commonly are needed from streams that are difficult or costprohibitive to gage. Streamflow can be estimated for many ungaged stream locations by using digital watershed or regression models. The following are examples of frequently used digital watershed modeling applications in which verification or calibration by using long-term data from streamflow-gaging stations (referred to as “stations” in this report) are required: Hydrologic Simulation Program-Fortran (HSPF) (Donigan and others, 1984), Spatially Referenced Regresssion on Watershed Attributes (SPARROW) (Preston and others, 2009), Precipitation Runoff Modeling System (PRMS) (Leavesley and others, 1983), MIKE_11 [Havnø and others, 1995], and TOPMODEL [Wolock, 1993]). Regression models often are used to estimate flow statistics at ungaged stream locations (for example, the U.S. Geological Survey, StreamStats program, U.S. Geological Survey, 2009a), and many regression applications require use of a least-altered period of streamflow data (or streamflow record) for model calibration (Esralew and Smith, 2009). A least-altered period of streamflow record was used to create the U.S. Geological Survey (USGS) Hydroecological Integrity Assessment Process for Oklahoma (HIP) (Turton and others, 2009). This model can be used to estimate how proposed flow alteration may affect flow parameters that are vital to aquatic ecological function. For this report, continuous streamflow data from streamflow gaging-stations are referred to as “streamflow record” and a specified period during which continuous daily streamflow data are calculated is referred to as a “period of record”. Ideally, modeling applications that use data from stations to characterize unaltered flow need to include data from a long-term period of record. This period of record should reflect the least-altered condition of streamflow available. Demand for water increases with population, urban, and agricultural development increases, and increasing recreational use. Typical anthropogenic activity that affects streamflow may include streamflow regulation, irrigation, diversion and groundwater withdrawal for industrial and consumptive water supply, effluent discharge, and urban development (Stankowski 1972; Fitzpatrick and others, 1999; Konrad and Booth, 2002; Smith and Wahl, 2003). The unpredictable nature of current and future patterns in climate also can affect streamflow. Streamflow data have been collected for streams in and near Oklahoma during periods ranging from a few years to nearly a century (USGS National Water Information System, http://waterdata.usgs. gov/nwis, accessed November 1, 2007). The least-altered period of record for some stations (referred to in this report as the “least-altered period”) may be short because of few streamflow data and an increase in urban or agricultural development of a stream with time. Shorter periods of record may coincide with aberrant climate and streamflow patterns that are not considered typical. Longer periods of record are more likely to provide a representative sample of central tendencies and variability of streamflow. A sufficient streamflow record length (at least 10 years, for example) would likely increase the probability that variability of the daily hydrograph, caused by recurrent climate cycles, is included in the streamflow data used for model calibration. Therefore, longer periods of streamflow record can help to minimize statistical bias and random error in modeling applications. A baseline period of record (referred to in this report as a “baseline period”) may be determined for stations with a sufficient streamflow record. The baseline period can be defined as a period in which streamflow is least-altered by anthropogenic activity and has sufficient streamflow record length to represent annual climate variability. By this definition, some stations cannot be assigned a baseline period because streamflow at these stations is substantially altered or the length of the period of record is short. To address the need for a least-altered period of record for use in modeling applications and water-resources planning, the USGS, in cooperation with Oklahoma State University and the Oklahoma Water Resources Board, conducted a study to determine baseline period of record for selected streamflow-gaging stations in and near Oklahoma. Streamflow at few if any |
| File Format | PDF HTM / HTML |
| DOI | 10.3133/sir20105106 |
| Alternate Webpage(s) | https://pubs.usgs.gov/sir/2010/5106/pdf/SIR10-5106.pdf |
| Alternate Webpage(s) | https://permanent.access.gpo.gov/gpo3404/SIR10-5106.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
