Powerful turbidity currents driven by dense basal layers

Abstract

Seafloor sediment flows (turbidity currents) are among the volumetrically most important yet least documented sediment transport processes on Earth. A scarcity of direct observations means that basic characteristics, such as whether flows are entirely dilute or driven by a dense basal layer, remain equivocal. Here we present the most detailed direct observations yet from oceanic turbidity currents. These powerful events in Monterey Canyon have frontal speeds of up to 7.2 m s^-1, and carry heavy (800 kg) objects at speeds of ≥4 m s^-1. We infer they consist of fast and dense near-bed layers, caused by remobilization of the seafloor, overlain by dilute clouds that outrun the dense layer. Seabed remobilization probably results from disturbance and liquefaction of loose-packed canyon-floor sand. Surprisingly, not all flows correlate with major perturbations such as storms, floods or earthquakes. We therefore provide a new view of sediment transport through submarine canyons into the deep-sea.

Fig. 1

Instrument array. Schematic drawing (not to scale) showing monitoring instruments deployed within Monterey Canyon. Inset maps show location of Monterey Canyon with respect to California and area covered by map in Fig. 2. Moored ADCPs were positioned 65 to 70 m above bottom (mab)

Fig. 2

Velocity and backscatter during two through-canyon flow events. a Map shows mooring locations with repeat mapping areas outlined in white. b ADCP-measured velocities capture the arrival of the 15 January 2016 and 1 September 2016 flows as they reach successive moorings within the canyon. Unresolved velocity readings are in white. Maximum ADCP-measured velocities (white numbers) and transit velocities (black numbers) are shown for comparison. c ADCP-measured backscatter, beginning 5 min before the arrival of the event, shows the evolution of the flow at each mooring. Echo Intensity Counts (EIC) are proportional to decibels. The ADCPs captured a 65 to 70 m range above the seafloor when the moorings were stable and upright. Ranges to seafloor decreased when the moorings were pulled downwards during events

Fig. 3

Timing of flow events compared with wave height and Salinas River discharge. a Occurrence of sediment density flow events and runout depth shown with red vertical lines. Three flows ran past the last mooring, and their full runout distance is unknown. The gray lines show the depth range covered and timing of repeat bathymetric surveys of the canyon floor. The three six-month-long mooring deployments are noted. b Blue line shows the maximum wave heights (H_Max) measured by the wave height sensor (Fig. 2, Supplementary Data 6). c Salinas River discharge data (Supplementary Data 7) shown in purple. Dotted red lines in b and c indicate when flow events occurred

Fig. 4

24 November 2016 flow event records. a Thirty-five-minute-long record of ADCP-measured velocity and backscatter recorded on mooring MS1 during the 24 November 2016 flow event. b Images of a round BED and the 800-kg tripod frame with BED11 and AMT attached. Also shown is one foot of the tripod frame, exposed above the seafloor after the event. c Changes in depth of 6 BED instruments during the 24 November 2016 flow event for the same time period shown in a. Dotted line connects initial bed movements, which are used to calculate the transit velocity through this depth interval. Gray area in a and c indicates period when all six instruments were moving simultaneously. d Plot of depth and temperature vs. time from the AMT (sampled every 45 min) during the third deployment (Fig. 3). The range of water depth shown in c and d are equivalent. The red oval indicates the time interval shown in c

Fig. 5

Repeat mapping shows changes in canyon floor morphology. a Bathymetric maps for the upper canyon collected with an Autonomous underwater vehicle between 42 and 540 m water depths on 4 November 2015 and b on 28 January 2016 (Fig. 1). Initial and final positions of instruments that moved on 15 January 2016 are also shown in a and b, respectively. c Changes in seafloor elevation between surveys a and b. x, y, and z are enlarged sections of a, b, and c, respectively. CSB Crescent-shaped bedforms

Fig. 6

Conceptual drawings of sediment density flow events in Monterey Canyon. a The highest velocities (V₁) occur in a dense basal layer near the flow front. This dense basal layer forms via liquefaction or mechanical erosion of underlying loose-packed sand, and helps to generate trains of crescentic bedforms. b Increased turbidity in the water column is coincident with slowing of the remobilized layer. c The evolution of a flow as it progresses down canyon. (Stage 1) A failure in the canyon floor results in the liquefaction of the seafloor at the front of the flow, (Stage 2) it propagates down-canyon as a dense remobilized layer, (Stage 3) the fast flow progressively generates an expanding dilute turbulent cloud, (Stage 4) which continues as a dilute turbidity current

Fig. 7

Movement of massive 800 kg frame down canyon during the 24 November 2016 event. a Black line shows depth (converted from pressure) vs. time from BED 11 which was attached to the 800 kg frame (AMT/BED11; Fig. 4) during the 24 November 2016 event. Red line segments are polynomial fits to three sections of these data. BED 11 traveled at an average speed of 4 m s⁻¹. b Deviations (blue line) of BED 11 trajectory (a: black line) from the fitted polynomials (a: red line) show vertical oscillations between 1–3 m