An Integrated Stream Classification System for the State of Texas
Eric S. Hersh, CRWR
Table of Contents
Motivation and Background
There is a recognized need for a methodology to scale up the results of site-specific environmental flow studies (typically habitat-based) into a larger framework for management, regulation, and implementation; integrated stream classification systems are viewed by many to be a promising avenue to accomplish this task. To preserve ecological relevance and thus ensure the protection of a “sound ecological environment,” the development of stream classes must take into account the importance of multiple systems and processes typically grouped into the four disciplines of instream flows: (1) Hydrology & Hydraulics (including climatology); (2) Water Quality; (3) Geomorphology & Physical Processes; and (4) Biology.
The value and need for stream classification was recognized in the May 2006 Texas Instream Flow Program (TIFP) Draft Texas Instream Flow Studies: Technical Overview:
…it is anticipated that classification tools will be developed to aid in the application of instream flow standards to the state’s myriad rivers and streams. It would be a near-impossible task to individually study all the state’s 191,000 river miles. Derivation of hydrologically, ecologically, and geomorphologically similar aquatic ecosystem units would enable the establishment and application of streamlined methods for developing instream flow recommendations.
A current project underway at CRWR makes use of Geographic Information System (GIS) technology to organize existing information relevant to the understanding of Texas streams and rivers (i.e., water quality, geologic and geomorphic, hydrologic, and biological data) and develop a classification scheme such that particular classes or regions of streams and rivers could be recognized as having a common identity. Accordingly, conclusions drawn from instream flow studies in particular river reaches might have a wider applicability than the particular study site. Moreover, this project could help prioritize areas for future instream flow studies.
A qualitative regionalization of Texas streams and rivers is presented in the National Research Council Committee (2005) review of the Texas Instream Flow Program; this regionalization and its boundaries were examined using quantified criteria. The State of Texas was partitioned into five regions: East Texas, South-Central Texas, Lower Rio Grande Basin, West Texas, and North-Central Texas via a series of qualitative parameters by basin.
Based on a review of stream classification literature (documented elsewhere), meetings, discussions with stakeholders and peers, and a review of available data, a series of quantitative parameters were selected for evaluation in the stream classification system.
Summary of quantitative variables.
Channel bed slope is an important driver of habitat availability as it is a primary factor in determining flow velocity and mesohabitat (pool, riffle, run, etc). Bed slope for each linear stream reach is included in the FlowlineAttributesFlow table in NHDPlus and is calculated by taking the difference in the upstream and downstream node elevations from the National Elevation Dataset divided by the reach length from the National Hydrography Dataset. The marriage of NHD and NED in NHDPlus allows for the examination of the longitudinal elevation profiles of water bodies.
Longitudinal elevation profile for the Colorado River, Texas. Note the step-shaped reaches between kilometers 900 and 1150, which are the Highland Lake system reservoirs and dams, with the largest vertical reach (approximately kilometer 1000) being Mansfield Dam at Lake Travis.
The United States Geological Survey (USGS) Water Watch program provides a near real-time snapshot of streamflow across the country on its website, grouped by percentile classes http://water.usgs.gov/waterwatch/ (Figure 24). CRWR recently acquired a dataset of the daily streamflow distribution at 22,380 gages across the nation by percentile of historic flow (1, 5, 10, 20, 25, 30, 40, 50, 60, 70, 75, 80, 90, 95, and 99% along with minimum, maximum, mean, and standard deviation). The dataset is updated annually and includes gages with greater than 30 years of daily data. In addition to daily flow distributions, the data files each contain an annual summary of flow distribution; this data is included in the NHDPlus in the StreamGageEvent layer.
USGS Water Watch near real-time streamflow summary map and legend.
This type of data can be conceptualized using the four-dimensional data cube representing:
-
space (a USGS gaging station),
-
time,
-
variable (streamflow), and
-
probability distribution.
Plotting a time series of this daily streamflow distribution data enables one to visualize and gain an understanding of the flow variation in a stream. For example, the Sabine River near Gladewater, Texas gage (USGS #08020000) is subject to a high degree of regulation from hydropower production at Toledo Bend Reservoir and Dam and thus exhibits a wide variation in daily streamflow distribution; aggregating the daily distribution to a monthly time step by using a centered moving average can provide a clearer picture of the flows.
One-month centered moving average of daily streamflow distribution for USGS Gage #08020000.
The USGS maintains an extensive nationwide network of real-time streamflow gaging stations across the United States; these stations measure real-time stage and that information is combined with an established rating curve to estimate the volumetric streamflow. The rating curves are continually calibrated and verified by USGS employees who visit the gage sites to measure the stream velocity and the flow cross-sectional area. This field measured data is also recorded and available via the National Water Information System at: http://nwis.waterdata.usgs.gov/usa/nwis/measurements.
A similar distribution of available stream habitat can be generated by using the percentiles of width and depth from the field measurement data. The accuracy of these distributions might be increased by combining the frequency distribution of the measured field data with the percentile streamflow distributions using a look-up table to assign flow width and depth to each of the daily streamflow percentile values for a given gage.
A limitation of this method is that the gaging stations are often sited at altered river cross sections, such as at bridges or culverts, or the gage itself causes alteration of the river, such as via the installation of a weir. Thus, the wetted width and depth distributions at a gage may not be representative of those for a broader reach or river. Nonetheless, techniques for the understanding and characterizing the probability distributions of available habitat (via width and depth) may prove useful in gaining an understanding of the linkage between hydrology and biology.
Monthly average depth distribution, Gage #08020000.
A slew of recently developed and recently publicized data sources have made the present an exciting time for the study of riverine systems, and the near future looks to offer more of the same. As such, an integrated stream classification system for Texas will benefit greatly.
National Research Council Committee. (2005). The Science of Instream Flows: A Review of the Texas Instream Flow Program. Committee on Review of Methods for Establishing Instream Flows for Texas Rivers, National Research Council. The National Academies Press, Washington, D.C. http://www.twdb.state.tx.us/instreamflows/pdfs/NAS_Report.pdf
Texas Instream Flow Program. (2006). Draft Texas Instream Flow Studies: Technical Overview: http://www.twdb.state.tx.us/InstreamFlows/pdfs/TechnicalOverview-Draft052206.pdf
A recent progress report from this project can be found online at: http://www.crwr.utexas.edu/gis/gishydro07/InstreamFlows/StreamClassification_files\CE394K2_Hersh_final.pdf
Eric S. Hersh
Graduate Research Assistant
Center for Research in Water Resources
University of Texas at
E-mail: ehersh@mail.utexas.edu