Understanding longitudinal ADCP measurements to determine water velocities for open channel flow

Banjavcic, Scott David

Permalink

https://hdl.handle.net/2142/100936 Copy

Description

Title

Understanding longitudinal ADCP measurements to determine water velocities for open channel flow

Author(s)

Banjavcic, Scott David

Issue Date

2018-04-09

Director of Research (if dissertation) or Advisor (if thesis)

Schmidt, Arthur R.

Doctoral Committee Chair(s)

Schmidt, Arthur R.

Committee Member(s)

• Parker, Gary
• Kalita, Prasanta K.
• Best, James
• Cox, Amanda L
• Jackson, Patrick R

Department of Study

Civil & Environmental Eng

Discipline

Civil Engineering

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• accoustic Doppler current profiler
• ADCP
• open channel flow, depth-averaged velocity, velocity profile, bed shear stress, longitudinal ADCP

Abstract

Despite the advances in technology, the traditional standard of practice of measuring velocities in open channels relies on the historically proven methodology that was primarily focused on accurately determining the discharge of streams and rivers. The traditional standard of practice relies on time-averaged stationary measurements and transect measurements (performed for cross sections roughly perpendicular to the flow) to describe complex hydraulic processes. This research seeks to understand a longitudinal acoustic Doppler current profiler (ADCP) measurement approach as an alternative to the present ADCP measurement practice (collecting stationary measurements or measurements along transects). The growing need for precisely interpolated velocities at all points throughout a river reach necessitates a more spatially diverse approach to data collection. The author explores the hypothesis that longitudinal measurements can predict velocities for a river reach scale more effectively than collecting stationary measurements or measurements along transects. More precise velocity mapping for river reaches could have a direct impact on engineering topics such as determining areas of scour and deposition, estimating habitat suitability, calculating dispersion of pollutants, calibrating 2D and 3D hydrodynamic models, and more. The following research objectives increase the understanding of the longitudinal measurement approach in three specific areas: determining depth-averaged velocity for locations in a river reach, extending velocity interpolation to vertical velocity profiles, and utilizing the interpolated velocities to estimate bed shear stress for a river reach. The first research objective focusing on exploring depth-averaged velocity for locations in a river reach presents interpolated transect and longitudinal velocities and compares them to the known transect and stationary data. The two main variables explored to compare the data collection methods were data density and data collection effort. Data density was determined by theoretical transect spacings, which correspond to the standard practice of interpolating between transect measurements to determine depth-averaged velocities for an entire river reach. The transect spacing was defined in terms of dimensionless RiverWidths to allow the results from this research to be scaled to rivers of different sizes. For each data density (denoted in RiverWidths) that was investigated, the data collection effort time was determined for the transect and longitudinal data collection methods. It was very important to ensure equivalent measurement time (defined in this research as data collection effort time) and effort for all comparisons to present an accurate comparison of the stationary, transect, and longitudinal data collection methods. The results from this research objective indicate that the longitudinal interpolated velocities match the known data and mimic the cross section velocity trend better than the interpolated transect depth-averaged velocities. In addition, the interpolated longitudinal velocities were analyzed to determine an optimal data density. The following recommendations were developed to guide data collection efforts for transect and longitudinal ADCP data to reduce the average absolute relative error when interpolating depth averaged velocities. 1. The relationship between RiverWidth and the absolute relative error implies an increase in error of 1% to 2% per RiverWidth for the interpolated longitudinal depth-averaged velocities, and an increase in error of 2.5% to 5% per RiverWidth for the interpolated transect depth-averaged velocities. 2. The absolute relative error for interpolating longitudinal velocities is the same as for interpolating between transects when the longitudinal measurement passes are the same distance apart as the transects. 3. An increase in the average absolute relative error from 0.1% to 0.4% per meter spacing between longitudinal measurement passes was observed for the Pecatonica and St. Joseph rivers. The second research objective explores using transect and longitudinal ACDP measurements to interpolate velocities throughout the water column. Both dimensionless depth layers (which favor interpolating in layers that utilize velocities in similar positions in the water column) and elevation-difference defined depth (horizontal layering) were explored to test the effect of channel shape on velocity interpolation. In general, the difference between velocities interpolated using dimensionless depth versus elevation defined depth was less significant than the difference between interpolated velocities developed using the longitudinal data collection technique as opposed to the transect data collection method. The following observations provide guidance for using transect and longitudinal ACDP measurements to interpolate velocities throughout the water column. 1. The longitudinal measurement technique is a better alternative to interpolation between transect measurements for describing velocities at various depths and locations in a river reach. 2. The dimensionless depth layering approach is marginally better than layering by elevation difference (horizontal) for river reaches with significant bathymetric variation. The final research objective presents potential hydraulic applications for the depth-averaged velocities and vertical velocity profiles. Many hydraulic applications including: determining habitat suitability, characterizing secondary flow, estimating dispersion, determining bed shear stress, and others may benefit from ADCP data collection utilizing the longitudinal method along with the stationary and transect-focused methodology. This research objective focused on using the dimensionless depth velocity profiles developed by the previous research objective to estimate the bed shear stress. Bed shear stress is an important hydraulic parameter that affects deposition and scour, which directly impact navigation (among other things) in open channel flow. The following information can be gleaned from the bed shear stress estimates developed using the interpolated dimensionless depth longitudinal vertical velocity profiles. 1. Areas where the bed shear stress estimates were small were consistent with the low moving-bed velocities determined for the studied rivers. 2. There is significant error in the bed shear stress estimates in areas with complex bathymetry, which is consistent with the increased error observed in the vertical velocity profiles. 3. The error in the bed shear stress estimates suggests that longitudinal ADCP measurements should be used with caution in areas of complex bathymetry. Overall, this research concludes that the longitudinal data collection approach is very useful for generating depth-averaged velocities and vertical velocity profiles, and has significant value as an important tool for a hydrologist’s toolbox to better solve complex open channel flow problems.

Graduation Semester

2018-05

Type of Resource

text

Permalink

http://hdl.handle.net/2142/100936

Copyright and License Information

Copyright 2018 Scott David Banjavcic

Owning Collections