Analysis of Acoustic Scattering Layers In and Around Petermann Fjord, Northwest Greenland

TitleAnalysis of Acoustic Scattering Layers In and Around Petermann Fjord, Northwest Greenland
Publication TypeThesis
AuthorsHeffron, E
Degree and ProgramMaster of Science
DegreeEarth Sciences/Ocean Mapping
Number of Pages469
Date PublishedMay
UniversityUniversity of New Hampshire
LocationDurham, NH

In 2015 a major international collaborative expedition took place focused on understanding the processes associated with the recent rapid decline of the Greenland Ice Sheet (GIS) and the impact that this decline could have on global sea-level rise. The Petermann Expedition collected a broad range of data designed to characterize the Petermann Glacier system, a marine-terminating glacier with a floating ice tongue that has undergone dramatic changes in the last decade. During the expedition, sonars were used to map the seafloor and the water column, generating a continuous dataset over 30 days. The water column mapping revealed extensive acoustic scattering layers, so called because the components of the layer – typically zooplankton and fish – scatter acoustic energy when concentrated in layers in the water column. The scattering layer was observed to change depth in a geospatially consistent manner and corresponded to our general, but limited understanding of the complex circulation patterns in the study area. This unexpected observation became the research question investigated in this thesis: Is the distribution of the acoustic scattering layer observed in and around Petermann Fjord a proxy for spatial and temporal changes in water mass structure and interactions? In order to answer this question, we focused on four objectives: determine the geospatial distribution of the scattering layer, determine if light influences the scattering layer depth distribution, determine if there is a consistent relationship to water column structure and circulation, and investigate the components of the scattering layer for clues as to its make-up and subsequently any potential reasoning for its distribution. Understanding the distribution of water masses and their circulation patterns in Arctic fjords are critical to understanding the fate of floating ice shelves and the glaciers they buttress, as the most pronounced change is occurring where ice sheets are grounded below sea level due to enhanced interaction with warming ocean waters. However, our ability to predict future sea level rise is hampered by our limited knowledge of these glacial systems, including the regional water mass distribution and circulation responsible for that enhanced ocean-ice interaction. Indeed, quantification of melting processes at marine terminating glaciers represents the largest source of uncertainty in predicting global sea level rise (Church et al., 2013). Traditional methods of oceanographic observation provide relatively sparse information at high cost, whereas acoustic records are continuous and, if the observed relationship between scattering layer depth and regional hydrography holds true, can potentially provide information about circulation, productivity, and ocean dynamics over large areas from underway platforms. Evaluation of the scattering layer distribution focused on the continuous Simrad EK80 18 kHz split-beam echosounder sonar records (section The top of the scattering layer was manually picked on each echogram, providing the latitude, longitude, and depth for the top of each layer (section 3.2.1) that were then plotted to show the geospatial and depth distribution. The resulting distributions (section 4.1) showed a recognizable geospatial pattern that was consistent with our understanding of the distribution of water masses. Broadly, there was a scattering layer generally present in the fjord along the coast of Greenland (eastern Hall Basin) and ringing central Hall Basin, and absent in northern Hall Basin, along the coast of Ellesmere Island (northern Nares Strait and western Hall Basin), central Hall Basin, and southern Nares Strait. The top of the scattering layer was significantly shallower in the fjord and along the coast of Greenland, deepening in the central ring and western Hall Basin (when it was present). We evaluated whether there was a linear correlation between the scattering layer depth and the bathymetric depth and slope (sections,, 3.2.2), but no correlation was found (section 4.2.1). The second objective was to determine whether the scattering layer distribution was influenced by light rather than water mass distribution. This analysis was undertaken because of the typical association of scattering layers with daily migrations corresponding to daily light cycles as a means of predator avoidance (section 1.3.3). Though the expedition took place in Arctic summer during the ‘midnight sun’ regime of 24-hour light, there was enough daily change to discern a cycle in the ship-based radiation data collected by a Photosynthetically Active Radiation (PAR) Sensor mounted on the roof of the ship’s bridge (section The relationship between light levels and scattering layer depth was examined (section 3.2.3), finding no linear correlation (section 4.2.2). A second analysis was done to see if we could discern a difference in water clarity across the study area using satellite-derived Kd(490) data, the diffuse attenuation coefficient for downwelling irradiance at 490 nm (section 3.1.3), and evaluate its effect on the scattering layer depth. Though available data for this region was very limited and there was some evidence of higher attenuation in the fjord where the shallower scattering layers were typically located, no correlation between scattering layer depth and Kd(490) values was found (section 4.2.3). Thus, neither light levels nor water clarity were responsible for the depth distribution of the scattering layer. The third objective was to determine if there was a consistent relationship between water mass properties and scattering layer depth beyond that established by initial observations (section 1.1). Profiles of conductivity, temperature, and depth (CTDs) were collected at 46 sites during the expedition to provide information on water mass properties and facilitate interpretation of regional circulation (section Plots of temperature versus salinity (T-S diagrams) and temperature and salinity versus depth were generated for each CTD location, and the average depth of the scattering layer for that location was overlain on the plots (section 3.2.5). Examination of the T-S diagrams revealed a pattern in scattering layer preference for specific sections of the water column (section 4.2.4). Of the 38 profiles with an associated scattering layer, 22 had scattering layers with a preferred depth range that fell in the deeper, warmer, saltier portion of the water column associated with Atlantic Water, where salinity and temperature (and therefore density) values were steady – we called this group the ‘homogeneous preference’ scattering layers, in reference to the lack of change or stratification in the water column. Twelve of the profiles had scattering layers with a preferred depth range that fell in the shallower, cooler, fresher portion of the water column associated with Winter Water (or more generally, the Arctic outflow), where salinity and temperature (and therefore density) were changing relatively quickly with depth – we called this group the ‘heterogenous preference’ scattering layers, in reference to the changing, or stratified, water column. Four of the profiles had scattering layers that fell right at the location where the water column properties were moving from stratified to steady. This group we refer to as ‘transitional’ scattering layers. The homogeneous preference scattering layers were found primarily in Hall Basin and the western side of the fjord mouth, areas associated with inflow of Winter Water and Atlantic Water from the Arctic Ocean to Nares Strait. The heterogeneous preference scattering layers were found in the fjord, an area associated with the influence of meltwater from the glacier (Petermann Glacier Water) and outflow from the glacier face through the fjord. Transitional preference layers were found primarily on eastern side of the fjord mouth, an area associated with meltwater-influenced outflow moving up along the coast of Greenland. Six of the profiles did not have an associated scattering layer, and all were found along the western edge of Nares Strait/Hall Basin, a region associated with low oxygen, cold, fast flow from the Arctic Ocean moving south through Nares Strait. These results show a clear relationship between the scattering layer depth and regional water column structure and circulation (Conclusions, Chapter 5). The final objective was to investigate, if possible, what the scattering targets in the layers were (section 3.2.6). Target strength analysis of individual targets visible in and around the scattering layers in the EK80 data showed average target strengths of -42.04 to -44.04 dB (section 4.3.2). Estimates of volume scattering for larger sections of the scattering layer were fairly weak, 57.17 to -81.70 dB (section 4.3.2). The high individual target strengths and visual observations of single targets in the echograms (section 4.3.1) seem to indicate larger targets, with a strong possibility being Boreogadus saida, polar cod. The low volume scattering values and density estimates made using the volume scattering and individual targets strength values, however, do not seem to indicate that the visually dense scattering layers in the echograms were composed entirely of these fish, so we believe the scattering layers may be a mix of fish interspersed with smaller fish and zooplankton (Conclusions, Chapter 5). All analyses described in this thesis was complicated by the fact that this was a ‘dataset of opportunity’, i.e., the objectives of this study were not at all part of the original work plan of the expedition. Water column sonar data were collected continuously, but that collection was focused on the search for gas seeps and secondary to the many other data collection efforts taking place on the expedition. Ship radiation data were collected as a matter of course but via an uncalibrated instrument not intended for this expedition. Lack of water clarity data led to the use of remotely sensed data to attempt to estimate this parameter, and lack of biological sampling pushed us to dig into the echograms for clues as to the scattering layer components, as no ground truthing of either parameter was available. Despite these complications and imitations, we were able to extract useful information from the data and clearly demonstrate that acoustic records such as these can be used to show patterns in water mass distribution and circulation and provide clues to biological communities in this region. Optimizing water column profiling for these objectives opens up the potential of using a rapidly-acquired acoustic remote sensing technique to provide critical information on water mass distribution as a standard underway tool.