Abstract
The interaction of internal tides with submarine canyons can generate pulsed delivery of continental slope waters onto the continental shelf, strongly influencing shelf sea physics and biogeochemistry. The continental slope of southeast New Zealand presents a unique opportunity to investigate this “tidal pumping” mechanism, where a system of seven canyons incises the slope, and where the nature of slope-shelf exchanges has received little attention. The present study utilises repeat CTD transects and temperature time series from moorings located at the head of Saunders Canyon between 2019-2021 to investigate timescales and drivers of physical variability in this system. Additionally, a collaboration with a local commercial fishing vessel enabled multi-week temperature and velocity time series to be obtained by using deep-sea fish pots as sampling platforms.
Analysis of three years of legacy CTD surveys revealed that Otago Submarine Canyons are typically occupied by a ‘tongue’ of Subtropical Water (STW) extending seaward across the upper slope to ~400 m depth. Water properties consistent with SubAntarctic Surface Water (SSAW) are present offshore and below the STW and extend across the deeper canyon slopes. The canyon water column displays some seasonal variability, becoming cooler and fresher during austral winter with an erosion of the STW tongue. A 13-hour repeat CTD transect through Saunders Canyon also revealed evidence of an internal tide, with the canyon deep thermocline at ~275 m being displaced vertically by > 50 m, lifting slope water across a near-critical bathymetric slope at the canyon head, and providing favourable conditions for the generation of internal tidal bores (ITBs).
Moored temperature observations collected during summer 2019/20 and 2020/21 revealed large fluctuations in temperature occurring in both higher (< 24 hour) and lower (> 24 hour) frequency bands. Lower frequency temperature variability was found to be correlated with local wind forcing, with temperatures at the canyon head decreasing (increasing) under upwelling (downwelling) favourable along-shelf wind stress, probably indicating a lifting (depression) of the thermocline up the canyon slope. Higher frequency temperature variability was found to be strongest in the semi-diurnal tidal band (12.4 hours) and related to rapid cooling consistent with propagation of ITBs onto the adjacent shelf. These ITBs generally occur around local high tide, extend vertically up to 40 m above the bottom, and account for a substantial part of a heat budget calculated at the canyon head. Variability in the occurrence of ITBs in the temperature records was also found to be related to local tide and wind forcing, with ITBs occurring more frequently during periods of elevated tidal sea level variability (e.g. perigean tides) and relaxed wind stress.
A three-week time series of seafloor temperature and velocity measurements collected at the head of Saunders Canyon in collaboration with a local fishing vessel supported the identification of rapid cooling as ITBs. Cooling signals were found to be associated with up-slope (down-slope) transport and cooling (warming) during their leading (trailing) edges. Upslope (downslope) flows were also found to be strengthened during periods of upwelling (downwelling) favourable wind stress along the continental shelf edge, suggesting the promotion of on-shelf flows at depth during these conditions.
Collectively, these results highlight an important and potentially predictable role internal tidal bores play in transporting slope waters from the Otago Submarine Canyons onto the adjacent continental shelf. Regular cold water intrusions and mixing produced by internal tidal bores are likely to substantially influence the local biogeochemistry around the Otago Submarine Canyons, and through the input of macronutrients to shelf waters, may help to support the elevated primary and secondary productivity reported in this region. Future efforts should focus on the collection of synchronised CTD and moored instrument data over longer timescales and across multiple canyons to better understand the origin and physical properties of the ITBs and the variability of slope-shelf transport across the Otago canyon system. High-resolution numerical models would also be advantageous in differentiating between the barotropic and baroclinic contributions to physical variability and slope-shelf transport.