|Document Type:||Journal Article|
|Title:||Crew variability in topgraphic surveys for monitoring instream fish habitat: a case study from the Columbia River Basin|
|Author:||Sara Bangen, Joe Wheaton, Nicolaas Bouwes, Chris E. Jordan, Carol Volk, Michael B. Ward|
|Journal:||River Research and Applications|
|Keywords:||Columbia River Basin,|
Digital elevation models (DEMs) derived from ground-based total station surveys have become ubiquitous in the field of fluvial geomorphology. Their wide application in spatially explicit analysis includes hydraulic modeling, habitat modeling, and sediment budgeting. However, we still lack an understanding regarding the repeatability and precision of DEMs derived from ground-based surveys conducted by different, and inherently subjective, observers. This is of particular concern when we consider the proportion of studies and monitoring programs that are implemented across multiple sites and over time by different observers. Here we used a case study from the Columbia River basin, where seven field crews sampled the same six sites, to quantify the magnitude and effect of observer variability on DEMs interpolated from ground-based total station surveys. We were interested in quantifying the degree to which DEM-derived metrics and measured geomorphic change were precise and accurate, or if discrepancies were due to inherent noise in how or what different crews surveyed. We observed several large magnitude elevation discrepancies across crews (on the order of up to 3.62 m), however many of these i) tended to be highly localized, ii) were due to systematic errors (e.g. triangular irregular network (TIN) busts), iii) did not significantly affect the precision of DEM-derived water depths, and iv) in many cases can be identified and corrected post-hoc. We modeled two plausible geomorphic change scenarios and calculated net erosion and deposition volumes at a 95% confidence interval. We found several crews failed to accurately detect the ‘true’ budget result simply due to how far they chose to survey out on to the floodplain. While total stations are a tractable method of sampling stream topography, this study highlights that in order to reduce noise in DEM-based datasets it is crucial to establish consistent surveying methods (both through time and among field technicians) along with measures of data accuracy and precision.