Setting sediment quality guidelines: a simple yet effective method

Deriving sediment quality guidelines (SQGs) for marine sediments is a difficult task. It will often be a trade off between reproducibility and relevance. One of the fundamental questions in ecotoxicology is to decide what one should measure to detect response in ecosystems exposed to human disturbance. In this paper we use field data to estimate threshold levels eliciting effects on soft bottom macrobenthos collected at different sediment types and depths on the Norwegian Continental Shelf and test these against natural levels occurring levels in reference conditions. SQGs are presented from multivariate analyses based on 121 gradients (represented with Ba, THC, Cd, Cu, Pb and Zn) incorporating more than 2000 species. Clear clusters with slightly disturbed communities related to contamination loadings were evident in 35% of the gradients. We found large variations in naturally occurring contamination concentrations and in the threshold levels electing effects on the fauna at different sediment types and depths. For example, an increase in depth of only 100 m can triple the Cu and Zn concentrations that elicit effects. Lowest background and threshold levels were found in shallow, sandy sediment. Our results suggest that current SQGs are too high. We hypothesised that setting a SQG of 4-times background concentrations will give sufficient protection for the fauna from metal contamination. The overall background concentration eliciting effects on metal was 3.6x.     

Remote sensing of stream depths with hydraulically assisted bathymetry (HAB) models

This article introduces a technique for using a combination of remote sensing imagery and open-channel flow principles to estimate depths for each pixel in an imaged river. This technique, which we term hydraulically assisted bathymetry (HAB), uses a combination of local stream gage information on discharge, image brightness data, and Manning-based estimates of stream resistance to calculate water depth. The HAB technique does not require ground-truth depth information at the time of flight. HAB can be accomplished with multispectral or hyperspectral data, and therefore can be applied over entire watersheds using standard high spatial resolution satellite or aerial images. HAB also has the potential to be applied retroactively to historic imagery, allowing researchers to map temporal changes in depth.We present two versions of the technique, HAB-1 and HAB-2. HAB-1 is based primarily on the geometry, discharge and velocity relationships of river channels. Manning's equation (assuming average depth approximates the hydraulic radius), the discharge equation, and the assumption that the frequency distribution of depths within a cross-section approximates that of a triangle are combined with discharge data from a local station, width measurements from imagery, and slope measurements from maps to estimate minimum, average and maximum depths at a multiple cross-sections. These depths are assigned to pixels of maximum, average, and minimum brightness within the cross-sections to develop a brightness–depth relation to estimate depths throughout the remainder of the river.HAB-2 is similar to HAB-1 in operation, but the assumption that the distribution of depths approximates that of a triangle is replaced by an optical Beer–Lambert law of light absorbance. In this case, the flow equations and the optical equations are used to iteratively scale the river pixel values until their depths produce a discharge that matches that of a nearby gage.R² values for measured depths versus depths estimated by HAB-1 and HAB-2 are 0.51 and 0.77, respectively, in the relatively simple Brazos River, Texas. R² values for HAB-1 and HAB-2 are 0.46 and 0.26, respectively, in the Lamar River, a complex mountain river system in Yellowstone National Park. Although the R² values are moderate, depth maps and cross-sections derived from the HAB techniques are consistent with typical stream geomorphology patterns and provide far greater spatial coverage and detail than could be achieved with ground-based survey techniques. Improved depth estimates can be achieved by stratifying the river into different habitat types that normalize for differences in turbulence and substrate.