Milne Technologies Home Products and Services Recent Projects Publications and Project Reports Senior Staff Research Collaborations and Data Sharing Contact Us
 
 
 

Multibeam Sonar Lake Bottom Imaging and Mosaics

Milne Technologies provides sonar imaging of the nearshore lake and river bottom habitat using a side- or forward-looking Kongsberg Mesotech M3 multibeam sonar system and the Oceanic Imaging Consultants “Stand-Alone Mosaicking Module (SAMM)” sonar image processing software to generate acoustic “pictures” of the bottom and stitched together into a habitat mosaic. Bottom substrate features (rock, cobble, wood debris, sand-bars, etc.) and sub-surface structures (water intake pipes and caps, dredge cuts, etc.) can be readily identified and geo-referenced.

Figure 1. Multibeam sonar habitat mosaic from Lake 626, Experimental Lakes Area, Ontario.


Lake Bottom Substrate Roughness, Hardness, and Statistical Classification

Milne Technologies uses hydroacoustics to analyze bottom substrate habitat in three ways as outlined below.

1. E1 and E2 Analysis

The analysis, first described by Chivers et al. (1990) and Kloser et al. (2001), provides a measure of relative substrate roughness and relative hardness from echo integration of the primary (E1) and secondary (E2) bottom reflections. Secondary bottom traces are observed where hard bottom substrates cause the primary echo to be reflected from the survey vessel hull or surface, back to the bottom, and then returned to the transducer face. For this reason, the second echo observed on an echosounder is approximately twice the depth. A simplified illustration of the E1 and E2 principles is shown in figure 1.


Figure 1. Illustration demonstrating the principles of E1 and E2 analysis. Shown in both panels is an echogram from the hydroacoustic survey. Within each panel are two transducers (“T”) positioned over a soft bottom and a hard bottom. The first panel (a.) represents the hypothetical pathway of the primary echo. In this case, sound is transmitted from the transducer face (1) and intercepts the bottom. If the bottom is soft, much of the energy will be lost to absorption and only a small fraction will be reflected away (3) or back to the transducer face (2). However, if the substrate is hard, much of the energy will be reflected away from the bottom resulting in a strong echo return back to the transducer face (2). Echoes from very hard substrates will be of high energy amplitude and will bounce from the water surface (4), back to the bottom to be reflected again upwards (5) with a small proportion of the secondary echo being detected by the transducer (6). Secondary echoes from soft substrates will be weak with most of the remaining energy being absorbed by the bottom. In this case no secondary trace is observed.


Integration of the acoustic backscatter through the primary echo region (E1) provides a measure of the substrate roughness. Very smooth substrates such as compacted sand, smooth shales, and bedrocks often act as an acoustic mirror and reflect much of the returning echo away from the transducer face. This results in an echo amplitude waveform that has a very sharp rise but deteriorates rapidly with no tail (figure 2a). Substrates such as cobble, gravel, pitting and fissures in bedrock consist of many interstitial spaces and irregular surfaces that force much of the reflected sound to reverberate and multi-path back to the transducer face. The irregular surfaces increase the proportion of the total transmitted sound that is reflected back to the transducer; this delays components of the echo and thus increases the echo return-time to the transducer. The delay in the response can also be attributed to nonlinear effects and is accentuated within the side lobes and at the beam edge. The result for these hard substrates is an echo amplitude waveform that rises quickly but deteriorates slowly, resembling a long tail (figure 2b). Echo integration analysis through the E1 layer is limited to include only the tail region of the waveform.

Figure 2. Illustration of the E1 echo amplitude wave from reflections off (a) smooth and (b) rough substrates. Note that the curve of the wave form from the smooth substrate is sharp with a narrow tail and is in contrast to the shape of the amplitude curve from the very rough substrate.

 

Integration of the acoustic backscatter through the secondary echo region (E2) provides a measure of the substrate hardness. The second echo return is a measure of the acoustic absorption or impedance of the bottom substrate. Bottoms of mud, detritus, or materials of high water content will absorb a large proportion of the transmitted acoustic energy thus reducing secondary echo amplitude. Hard substrates tend to reflect a greater proportion of the transmitted energy resulting in high amplitude secondary echoes.

E1 and E2 analysis layers are constructed using the virtual line function in Echoview (figure 3). The E1 and E2 layers of the echogram were calculated from the bottom line pick following the methodology described in Kloser et al (2001).

2. E1 peak Sv

The “peak Sv” is calculated from the primary echo (E1) and is defined as the observed maximum volumetric backscatter (Sv) of all acoustic samples (or pixels) within the layer between the sounder detected bottom and the bottom of the E1 layer (figure 3).

Figure 3. Echogram from a segment of the Lac La Biche survey showing the three substrate analysis layers used to generate relative substrate hardness and roughness values. The shaded areas within each echogram represent the echo integration from the (a.) E1, (b.) E2, and (c) Peak Sv analyses.

 

The E1, E2 and Peak Sv analyses provide only a relative measure of substrate roughness and hardness. E1 and E2 measurements are heavily influenced by the acoustic foot print size, thus limiting the ability to resolve substrate changes with increasing water depth. E2 measurements of bottom hardness rely on sound reflection from the lake surface and are therefore influenced by wind and wave conditions. Mean Sv values from echo integration through the E2 layer over a known substrate type have been observed to vary with surface condition therefore limiting the ability to compare E2 values between surveys.

3. QTC EchoImpact Waveform Analysis

Statistical classification of the lake bottom substrate can also be completed using Quester Tangent’s Impact seabed classification analysis software. The commercial software uses a multivariate statistical analysis of the waveform shape from the primary bottom reflection and corrects for biases associated with depth, slope, and the acoustic footprint. Of 166 measurements of the wave form, the software uses a multivariate approach to identify the 3 principal components (Q1, Q2, and Q3) that best explain the variation across all survey pings. From plots of the Q-values in 3D, “Q-space” k-means clustering techniques are used to assign each ping to a substrate class based on similarities between the waveform shapes. Once classified and assigned a known substrate type (from cores, ekman dredge samples, etc.) a seabed catalogue is constructed so that the substrate classification can be easily assigned to all future acoustic surveys. Further details are available from the manufacturer’s website.

Figure 4.  Quester Tangent EchoImpact substrate classification of the lake bed from EY500 120 kHz surveys within the coastal region of Parry Sound, Ontario Canada.  The survey area shown is a 10 km X 20 km segment of the Georgian Bay coastline just west of the Frying Pan Islands.

 

 

Figure 5.  Supervised lake bottom substrate classification of Scott Lake for 13 substrate types. Substrate classification was generated from the 2009 EK60 ES120-7G hydroacoustic survey data using QTC Impact bottom classification software for 15 classes. QTC substrate classes were interpolated using the QTC CLAMS ordination method. Physical substrate types were assigned to each QTC class using bottom sampling observations (Ekman grabs and video).

 

References:

Chivers R.C., Emerson N. and Burns D.R. 1990. New acoustic processing for underway surveying. The Hydrographic Journal, 56:8-17.

Kloser R.J., Bax N.J., Ryan T., Williams A., Barker B.A. 2001. Remote sensing of seabed types in the Australian Southeast Fishery; development and application of normal-incident acoustic techniques and associated ground truthing. Marine and Freshwater Research 52:4475–489.

 
 
Milne Technologies Home Products and Services Contact Us