Goal 1: Quantitatively measure the three-dimensional structure of the submesoscale features responsible for vertical exchange
How does this goal help address the hypothesis?
Model studies (Fig. 1.2) and limited observations (Fig. 1.3) indicate that submesoscale vertical exchange is concentrated near km-scale fronts and eddies. High-resolution simulations have outpaced our observational capabilities, but observational techniques have matured rapidly over the past decade. This investigation will make a comprehensive set of measurements of the dynamical variables needed to validate the high-resolution simulations and test their performance.
What needs to be done to achieve this goal and why?
Wide-swath DopplerScatt aircraft remote sensing is the centerpiece of the observation strategy. By measuring surface currents at submesoscale sampling resolution (1 km) over a wide area (100 km) on a daily basis, it will allow the in situ components to be targeted at the important frontal features, while providing spatial context for these measurements. The key measurement is current gradients, from which vorticity, divergence and strain can be calculated, with an accuracy of 2×10−5 s−1=0.2f at 1 km sampling, sufficient to resolve the typical frontal values. In particular, mass conservation and the incompressibility of water ensure that the horizontal current divergence equals the vertical derivative of vertical velocity and thus links the surface current measurements to vertical transport. Simultaneously, DopplerScatt will measure wind speed, which is the most important component of atmospheric forcing and a key determinant of frontal structure. Accuracy of about 1.5 m s−1 will resolve the typical 3-10 m s−1 wind speeds.
Figure 1.2 - Surface vorticity off Central California in two submesoscale-resolving models in March (left column) and September (right column). ROMS (top row 500 m resolution) and JPL-MITgcm (bottom row, ~ 2km resolution) show different structures and different seasonal evolution. The rectangle in the upper left shows our experiment domain. In all panels, the vorticity has been normalized by the local value of the Coriolis parameter. The inset in the upper-right panel is the normalized vorticity measured by DopplerScatt in the Gulf of Mexico (Fig. 1.4), shown with the same distance scaling and color scale as in the model fields-- we now have an exciting new capability for mapping ocean vorticity and divergence.
Figure 1.3 - Visible Infrared Imaging Radiometer Suite (VIIRS) false color image of proposed study region offshore of San Francisco shows complex pattern in February 2016 due to mesoscale and submesoscale currents acting on a large phytoplankton bloom. The proposed campaign will use a combination of detailed quantitative mapping of velocity, chlorophyll, temperature, and other properties from aircraft with targeted in situ measurements to overcome the observational challenges involved in sampling such a complex field in 3D (NASA image by Norman Kuring, NASA’s Ocean Color Web.)
SST and ocean color from aircraft remote sensing provide critical targeting information on submesoscale frontal features, which will have strong gradients in both SST and color, and thus complement DopplerScatt. The observations in Fig. 1.5 were targeted in this way. Observations should coincide with scatterometry imaging to one hour, allowing velocity and SST/Color data to be aligned to much better than 1 km for typical currents of 0.1 m s-1. SST resolution of better than 0.05°C at 100m horizontal scale will resolve the typical 0.5ºC temperature difference across 5 km fronts while showing their internal structure. Similarly, Chlorophyll concentration resolution of 0.05 mg m-3 will resolve frontal differences of 0.5 mg m-3.
Figure 1.4 - Vorticity (left) and divergence (right) of surface currents measured by the aircraft-mounted DopplerScatt instrument in the Gulf of Mexico in April 2017 (Rodriguez et al. 2018). The ‘plume’ of positive vorticity in the left panel is associated with the Mississippi River outflow. The DopplerScatt swath width is about 25 km and the center of each swath is blanked out here. The associated surface currents are presented and discussed in Rodriguez et al. (2018). Vorticity and divergence have been normalized by the local value of the Coriolis parameter (f=0.7× 10−4 s−1).
Ocean horizontal velocity, temperature and salinity on a range of scales are needed to place vertical velocity measurements in a proper context. Although theory and models predict that submesoscale vertical exchange occurs within specific frontal features, the details of these features, and thus the magnitude of the exchange, vary with details of model physics. To understand this physics, and properly model it, we need to measure the structure of both the submesoscale features and the larger mesoscale features in which they are embedded. Two sampling schemes will be used. First, the submesoscale hydrographic and velocity fields will be measured by rapid shipboard surveys following a Lagrangian float, an approach that has been demonstrated to yield successful measurements of submesoscale structures even in strong currents (D’Asaro et al. 2011, 2018; Thomas et al. 2015). The ship will carry an acoustic Doppler current profiler (ADCP) to measure velocity and a towed profiling probe, measuring temperature and salinity. These will provide a survey resolution of 1 km, a span of 5 km, a depth of 100m, and repeat every one to two hours. Previous work demonstrates that this resolution and standard oceanographic sensors with temperature and salinity accuracies of 0.005°C and 0.01 g kg-1 will be sufficient, as will a standard RDI 300 kHz ADCP with accuracies of ~0.01 m s-1 with 60 s averaging. Wave Gliders and Saildrones will also measure upper ocean velocity profiles and near-surface temperature and salinity in tight 1-km arrays to resolve submesoscale gradients. Second, the mesoscale will be measured using an array of autonomous profiling Seagliders surveying a surrounding 25 km region, repeating a survey every two days with the spatial resolution of 5 km. These will measure temperature and salinity to a depth of 1,000 m. The greater depth is needed to capture the deeper mesoscale structures. Velocity will be measured from an ADCP on the Seagliders (Todd et al., 2017).
Ocean vertical velocity is a central focus of this program and must be measured. Two approaches will provide redundancy and different space-time sampling. First, Lagrangian floats with ADCPs have a demonstrated ability to directly measure vertical velocity associated with submesoscale vertical exchange at fronts (Fig. 1.5). Float location to an accuracy of 300 m several times per hour is necessary to place these measurements in proper context. Second, arrays of four Wave Gliders and at least six Saildrones, two types of autonomous ocean surface vehicles, will be piloted in formations with 1 km separation and will carry downward looking ADCPs to measure divergence as a function of depth from depths of a few meters to about 70 m. Integrating this divergence from its surface value, will yield continuous profiles of vertical velocity. Divergence accuracy of better than 2×10-5 s-1, will resolve typical 1 km divergences of 10-4 s-1. This same array will measure vertical vorticity and strain, also important frontal parameters, with similar accuracy.
Figure 1.5 - New observational methods are effective in measuring submesoscale vertical exchange (D’Asaro et al. 2018). Here, aircraft SST maps were used to locate a submesoscale eddy. Surface drifters were deployed into the eddy. Their patterns (a- black dots) were used to guide a ship survey (a- plan view white circles), which showed density front (b). A water following Lagrangian Float (red lines) was deployed near the front. The float moved into the front and measured the sinking of dense water (b,c) at the front both from its motion and from an ADP mounted on the float
