An assessment of a LiDAR-based approach for estimating Hydraulic Geometry regional and regime relationship curves for the Southern Driftless Area of the Midwest

Christopher P Haring

doi:10.17077/etd.005265

Stream channels form distinct geometric dimensions, patterns, and profiles based on natural physical processes that route water and sediment through the landscape. Natural stream channel characteristics can be assessed and quantified into variables from which empirical Hy-draulic Geometry Relationships (HGR) are developed for specific hydro-physiographic regions. The most commonly developed empirical HGR are regional and regime curves. Regional curves provide stream channel dimensions that relate bankfull channel dimensions and dis-charge to watershed drainage area. Regime curves relate bankfull stream channel width, mean depth, cross-sectional area, and mean velocity to bankfull discharge. The availability of both curve sets is essential for conducting basic research, hydrologic and hydraulic analyses in un-gaged watersheds, culvert and bridge design, stream bed and bank stabilization and restoration, wetland enhancements, and watershed assessments. Development of these curves requires the collection of a large amount of data using tra-ditional geomorphic field survey methods that are labor, time, and cost intensive. The re-sources for conducting the geomorphic, hydrologic, and hydraulic analyses and developing curves are often limited due to constraints on time, funding, and personnel. Limited project resources and time constraints often dictate that generalized HGR curves from other hydro-physiographic regions be applied to streams outside their region as substitutes. The substitu-tion of the curves from a different hydro-physiographic region introduces additional sources of error and in many situations are not interchangeable with those of the original analysis area. In addition, literature regarding stream restoration activities indicates that traditional geomorphic field survey methods for gathering basic bankfull channel dimensions may be error prone. The challenges and inaccuracies that are inherent in developing regional and regime curves using traditional geomorphic field-based data collection methods or utilizing curves from other hy-dro-physiographic regions make it imperative to seek and develop other more appropriate, re-peatable, cost-effective, and reliable methods for deriving physical channel morphological characteristics. In the Southern Driftless Area of the Midwest, there are no established regional or re-gime curves creating a lack of information that is vital for basic research, and for federal, state, and local governments, and private organizations. Given this need, a new approach was devel-oped and tested using Light Detection and Ranging (LiDAR) high resolution terrain data to de-fine bankfull regional and regime regression equations and curves for the Southern Driftless Area streams of the Midwest. LiDAR is an optical remote-sensing technique that uses laser light to densely sample the surface of the Earth, producing highly accurate x, y, and z meas-urements. The LiDAR-HGR approach integrates common traditional geomorphic field-based analysis techniques and data interpretation of high-resolution terrain data to save time, reduce cost, reduce potential field data collection error, and offer a relatively rapid approach for de-veloping regional and regime curves. LiDAR-HGR also provides the ability to more effective-ly observe and monitor temporal and spatial channel dimension, pattern, and profile change at the large watershed scale. In this study, 24 hydraulic geometry relationship equations and curves were developed for the Southern Driftless Area of the Midwest using the LiDAR-HGR approach. The resulting regression equations and curves were compared statistically to representative data from exist-ing HGR studies to assess viability of the method. The curves were also compared to a number of existing traditional geomorphic field surveys from 2009 and 2017 to determine relative ac-curacy between the field surveys and the LiDAR-HGR approach. In addition, LiDAR-HGR was compared to other regional curves of the Midwest for possible association and used to predict geomorphic stream channel dimensions from existing Indiana and Ohio USGS regional curve studies. The results of the analysis of the LiDAR-HGR approach are listed below. a) Based on the statistical analysis of 28 gage sites, there are three sets of eight regres-sion equations and curves established for three geographic areas; the Southern Drift-less Area region as a whole, the Lower Southern Driftless Area sub-region, and the Little Maquoketa River sub-region. b) Statistically, the regional and regime curves produced similar or better coefficient of determination values when compared to existing USGS regional curve studies from Indiana, Ohio, Michigan, and other regions of the United States. c) The average recurrence intervals based on annual maximum flood series analysis from this study resulted in values of 1.43 to 1.47-year values. Previous work by Dunne and Leopold (1978) and Rosgen (1996) found the average recurrence interval to be the 1.5-year event. d) Discharge, not drainage area, was found to be the basin characteristic that accounts for the largest portion of variance for the Southern Driftless Area curves. e) Drainage area was found to be relatively equivalent to discharge for the Lower Southern Driftless and Little Maquoketa River sub-regional curve sets. f) The comparisons of the LiDAR-HGR approach to a limited number of 2009 and 2017 traditional geomorphic field surveys resulted in errors ranging from zero to 23%. The results compare closely or better than reported error ranges from Wahl (1977), Roper et al. (2002), Lawlor (2004), and Eaton and Church (2007) of up to 30%. g) Bankfull channel data evaluated from other Midwest regional curve studies provided close comparisons between the Southern Indiana regional curves and the Lower Southern and Little Maquoketa River Driftless Area sub-regional curve sets. h) Bankfull channel dimensions from the Eastern United States regional curves do not match closely with the Southern Driftless Area regional curves and were shown to be marginally applicable to the Lower Southern Driftless Area and Little Maquoketa sub-regional curves. i) The LiDAR-HGR method was also tested against four Indiana and three Ohio USGS regional curve sites and closely predicted bankfull channel dimension values for six of seven sites. j) The results of this study suggest that the LiDAR-HGR approach is a suitable method to derive regional and regime curves for the Southern Driftless Area of the Midwest.

An assessment of a LiDAR-based approach for estimating Hydraulic Geometry regional and regime relationship curves for the Southern Driftless Area of the Midwest

Files and links (1)

Abstract

Details

Metrics