Squamish Prodelta Experiment

banner delta front
Project Overview
The Squamish Prodelta experiment is ultimately designed to test the limits to which real seabed morphologic change can be detected using integrated multibeam sonar systems. The experimental objectives are twofold:
  1. to detect and understand the extent of, timing of, and mechanisms involved in, seabed mass wasting on the prodelta
  2. to improve acoustic imaging methods to the point where such processes can be observed adequately.
lip animationThe Ocean Mapping Group (OMG) at UNB specializes in developing improved methods for survey, processing, analysis and presentation of seabed mapping information, primarily through the use of integrated multibeam sonar systems. To that end, research topics in positioning, orientation, integration, refraction, seabed backscatter and water column imaging are continually undertaken. As part of these developments, the OMG undertake field mapping programs to support these engineering research aims. Those programs provide an opportunity to test out and improve these research developments. The largest program is the CCGS Amundsen Arctic Archipelago Seabed Mapping> Program. The most flexible program, however, revolves arounds the operation of the CSL Heron with her CFI-funded suite of positioning and acoustic sensors. The Squamish Prodelta Program utilizes the CSL Heron and is an opportunity to apply research developments directly to questions about seabed sedimentary processes.




Upper Howe Sound
above sill bathymetryThe basin under investigation is the upper section of Howe Sound north of the Porteau Cove Sill. This sill is probably a terminal moraine and separates the southern shallower section of Howe Sound from the deeper northern section. Within this northern section, the fjord basin acts like a natural laboratory for mass wasting studies, It is ~ 16 km long and fed almost entirely by sediment input from the Squamish Delta. Secondary sources of sediment have been input from the Britannia Beach and Woodfibre locations, but volumetrically they are unimportant.

The southern section of this upper basin is almost flat at a depth of about 285m, probably reflecting the ponding of distal turbidity currents originating from the Squamish prodelta slope to the north. A series of active and fossil prodelta channels are clearly visible in the multibeam surveys of the region.  The fossil channels predate the installation of a training dyke in 1971. The active channels are the main focus of this survey.

The Squamish Delta Location
annotated deltaThe Squamish River drains a watershed within the Coast Mountain Range of British Columbia and debouches at the head of Howe Sound. During the summer freshet, a discharge of > 500m3/s is usually maintained for a period of more than 4 months during which high suspended sediment concentrations are present, and has resulted in the development of a delta just west of the town of Squamish. Peak discharge events of over 1000m3/s  do occur within the summer freshet. Notably however, even larger events (although of much shorter duration) are common in the autumn due to local runoff.

A training dyke, installed in 1971 has concentrated the full Squamish River discharge into a 500m section on the north west side of the fjord. The purpose was in part to be able to create the Squamish Port facility. Prior to the dyke, the main river flow actually ran through the current port facility location and, prior to 1921, actually flowed through what is now termed the Mamquam Blind Channel. Since installation of the dyke, the delta top has prograded several hundred metres to the south and now extends out partially cutting off the entrance to the port. Howe Sound is a macrotidal fjord. The tidal range varies from about 5m during Springs to 3m during Neaps.

The delta top consists of two intertidal sand flats that are developed on either side of the main subtidal river/estuary channel The river channel is deepest immediately adjacent to the training dyke with a depth of ~2-3m below chart datum.  As the channel extends beyond the training dyke it shoals and widen slightly as it approaches the delta lip. At the lip, the channel is about 200m wide and less than a metre below chart datum. During the winter-time low discharge period, the salt water of Howe Sound penetrates up the river, forming a seasonal estuary extending several kilometres upstream. During the freshet, the salt water is effectively expelled from the river.

Morphology of the Prodelta Slope
Seaward of the lip of the delta, the prodelta slope extends over 2000m and drops down to more than 180m depth. Superimposed on that slope are several prominent channels. Based on biannual, annual and semi-annual multibeam surveys over the past 6 years, three major conduits are noted to be currently active. They are herein termed the northern,the active channels central and southern channels. They have been the focus of the 2011 experiment which involved daily resurveys of these three channels to look for evidence of activity.

Southeast of the active delta is a pair of relict channels. The oldest one extends from the mouth of the Mamquam Blind Channel and, although predominantly buried, clearly used to have an incised meanderingall summer difference character. The younger one drained from the location of the current port facilty and was presumably the main active channel prior to 1971. Since the dyke installation it is clear that it has been completely abandoned and is being swamped by overbank sedimentation.

The net change in the prodelta morphology from April to August 2011 is presented in the figure to the left. As can be seen the three main channels were indeed the loci of activity. The yellow star indicates the location of the bottom mounted ADCP (see discussion later), which was installed to detect turbidity current events. As can be seen it was right on the edge of the region of detectable change. The instrument was in fact buried by one flow for a period 20 days.

Daily Seabed Change Detection
While the figure above illustrates the cummulative change over the summer, there remains a question about the scale of individual example eventmass wasting events and their periodicity. For that purpose, the resurvey frequency was designed to establish change over a 1 or 3 day period (every weekday).
The animation to the right shows the seafloor the day before and the day after one of the largest mass wasting events. As can be seen there is a clear shift in the transverse bedforms patterns on the channel floor. The difference map, highlights the location of the change.
The animation at the head of this report shows the daily evolution of the morphology of the lip of the delta and the prodelta slope immediately below. As can be seen, the events were intermittent, with pauses of several days between activity at times. During the peak activity however, daily events were common with a clear net upslope migration of those channel-transverse features. Also one can see the lip of the delta active prograding as a result of rapid dumping of suspended sediment as the flow decelerates off the lip of the delta. The apparent noise in the data is largely sonar mistracking on the gas plumes that were strongly but ephemerally developed on the prodelta slope.

CSL Heron and her instrumentation
heron underwayCSL Heron is a 10m H-class survey launch on loan to UNB from the Canadian Hydrographic Service. From 2002 until 2009 she operated with a suite of instruments centred around a 300 kHz multibeam sonar (EM3002). During that time she supported OMG researchHerom EM710 configuration objectives in New Brunswick coastal waters and the Canadian Arctic Archipelago (deployed from CCGS Amundsen and CCGS Henry Larsen).   In 2010 she was upgraded to a newer suite of instrumentation centred around a 70-100 kHz multi-sector multibeam sonar (EM710). It is this configuration that has been used for the Squamish prodelta experiment.

The EM710 is ideal for imaging not just the seabed over the range of depths (0-285m), but has sufficient signal to noise left at those slant ranges to provide useful water column imagery of the scattering field. In addition to the mapped seabed morphologic change, the identification of gas plumes and active water body intrusions, using the water column imagery was an unexpected highlight of the experiment.
inside heron
The Heron is both a research platform and a training platform for graduate students. The Squamish prodelta program forms the basis for 2 PhD and 3 MScEng theses. She can take a full complement of 6 persons, athough for typical survey operations, the skipper and two surveyors are the usual load. From the Heron we can also undertake oceanographic surveys (see below) and deploy current meters.
 


Positioning Issues
antennasThe fidelity of the apparent surface difference maps depends on the component uncertainties in the integrated mapping system. As part of this, the horizontal positioning confidenceand vertical accuracies need to be assessed. Flavours of GPS including: Inertially smoothed DGPS, RTG, PPP and PPK have been explored. These provide horizontal positioning accuracies in the range of metres, decimetres and centimetres. Vertically, the PPP and especially the PPK solutions allow us a few or a single decimetre level uncertainty (95%).
Such levels of achievable 3D positioning accuracy have to be combined with a matching level of offset and alignment surveys to integrate the antenna solutions with the sonar-relative angles and ranges.
Any horizontal positioning biases, will show up on the steep delta front as false apparent depth changes when one survey is compared to another. Thus one of the geomatics research aims of this program, is to minimise these 3D positioning errors so that the natural change in the morphology can be best observed.

Oceanographic Observations
mvp profilingA river mouth emptying into a macrotidal salt water body represents an area of highly dynamic oceanography. The mixing of salt and fresh water,Estuary MVP Section modulated by the basin and delta-top morphology, the tides and the variable river discharge result in rapid spatial variations in sound speed (a function of temperature and salinity). The sounds speed variability in turn, affects the accuracy of the range and angles derived from the sonar system. 

To address this, a rapidly cycling, towed CTD sensor was deployed (a Brooke Ocean MVP) which, through the use of longitudinal sections, allowed us to look at the the temporal and spatial variability of the estuarine mixing around the delta mouth.The figure to the right illustrates a temperature, salinity and optical backscatter section along the Squamish river estuary, extending about 2000m upstream from the lip of the delta. These sections were obtained twice weekly. Additional sections through the river plume and along the length of the fjord to Porteau Cove were obtain on a similar sampling scale.

Bottom Mounted ADCP Observations.
adcp launch
To directly detect the presence of turbidity current activity, an acoustic doppler current profiler (ADCP) was deployed in adcp data150m+ of water immediately downstream of the terminus of the northern channel. The aim was to record current and suspended sediment surges that might result from the slump events events apparent from the morphological change seen upslope.

Twenty discrete turbidity current events were indeed detected. They were characterized by peak velocities of up to 1.5m/s (more typically 0.5m/s) for a duration of over an hour. The turbid plumes were seen to ascend up to 40m above the seabed and be detectable for more than 8 hours after the nose of the event passed..

Two example events are illustrated in the figure to the right.




Water Column Imaging of Hyperpycnal Intrusions
water column events
Using multibeam acoustic imaging of the volume scattering in the water mass we are able to detect whether the deep scattering layer, developed in the fjord, was perturbed by intrusive events. Every day, transects were run along the three channel axes. The hope was that intrusive activity could be detected and correlated with the observed morphologic change and the turbidity current events. Several intrusions were indeed detected. Unless, however, the intrusion was still taking place at the time of the survey (usually a 45 minute window), there may be no evidence of the events that caused the morphologic change within the 24 hour window between the previous and this survey.

An optical backscatter probe was lowered into the channel axes, and evidence of enhanced suspended sediment levels were detected, correlated with intrusions, evidenced by the zooplankton displacement. At this time, the optical backscatter probe is not calibrated, so exact suspended sediment loads are not yet known…


Publications:
Participants:
  • Gordon Allison  – skipper CSL Heron (CCG retired)
  • John E. Hughes Clarke
  • Steve Brucker
  • James Muggah
  • Ian Church
  • Doug Cartwright
  • Travis Hamilton
  • Pim Kuus
  • Danar Pratomo
  • Brad Eisan
page maintained by JEHC