COMPASS Friday - Archive

Request Info

FALL 2018
Fridays at 11:00 am, RSMAS Auditorium (unless stated otherwise)










Samantha Ballard (AMP)
Introduction to Wavelet Analysis of Temporal and Spatial Signals
to Identify Turbulent Scales and Structures in the Coastal Environment

Littoral regions encompass complex, unique, small-scale turbulent features such as atmospheric gravity waves, boundary layer rolls, etc. The goal of the Coastal Land Air Sea Interaction (CLASI) project is to utilize in-situ and satellite measurements to better understand air-sea interaction phenomena and to improve the Coupled Ocean Atmosphere Mesoscale Prediction System (COAMPS) forecasts of coastal features in littoral regions such as Monterey Bay, CA. Improving model forecasts is beneficial for naval operations such as ship maneuvering, forecasts and trafficability. Coastal wind fields and small-scale turbulent features, in particular, change rapidly due to interactions with topography, wind fronts, breaking waves in the surf zone, etc. and can, therefore, be difficult to resolve and forecast by models. Synthetic aperture radar (SAR) is useful for obtaining high-resolution coastal wind fields, which we derive for Monterey Bay using an X-band SAR wind retrieval algorithm. SAR-derived wind fields resolve atmospheric boundary rolls, Kelvin-Helmholtz waves, and wind fronts that are not present in COAMPS wind field results. Knowledge of these coastal turbulent features' temporal and spatial scaling is critical for significant model improvement. Therefore, the question "What are the small-scale turbulence structures from the coast to the open ocean?" needs to be answered. Wavelet analysis is a valuable tool for analyzing variations of power within a temporal or spatial series. Through the segmentation of a temporal or spatial series in wavelet space, either time-scale (usually frequency or period) or space-scale (usually wavenumber or wavelength), respectively, one can identify the dominant scales of non-stationary, or non-periodic, fluctuations in the temporal or spatial series and where these fluctuations occur in time or space. The traditional Fourier transform analysis method lacks the ability to identify the time or position, shape, or size of the fluctuations in a temporal or spatial series and is better suited for stationary, or periodic analysis. The current literature lacks a cohesive composite of identification, extraction, and analysis techniques for turbulence structures, particularly scales, from both temporal and spatial data sets through wavelet analysis. Therefore, the goal of this presentation is to provide a straightforward guide for wavelet analysis of time and spatial series, utilizing coastal tower and SAR-derived wind fields from the CLASI dataset as example series. Wavelet analysis enables us to combine the SAR-derived and coastal tower wind fields to retrieve consistent information on spatial and temporal scales of the wind variations in coastal environments, such as Monterey Bay.

Sanchit Mehta (AMP)
A Comparison of Fresh and Sea Water Spume Droplet Production
in High Wind Conditions

Under high-wind conditions, the presence of spume particles in the marine boundary layer may significantly mediate the exchange of heat and momentum at the air-water interface. While spray dynamics have been the focus of many investigations in the ocean, spray production in freshwater bodies has been far less studied and remains poorly understood. Here we present the findings of the first laboratory experiment to directly compare the rates of fresh and sea water spume production in wind speeds up to 54 m/s. An unobtrusive, optical technique was used to directly image, count, and size individual spume droplets along a profile above the intensely breaking waves. Significant differences were observed between the two water types in terms of the rates of production as well as the vertical transport of spume droplets. The sea water trials exhibited an integrated particle density 48% to 146% higher than for fresh water, for the lowest and highest wind speeds respectively. While sea water tended to produce substantially more spray for a given wind force, the vertical transport of these droplets was more constrained than for the fresh water cases. This translated to significantly different empirical relationships describing the vertical profiles of particle density across the two media. The results of this unique study demonstrate that salinity plays an important role in the spume production at high wind, which ultimately holds implications for characterizing spray-mediated air-water flux over varying water masses.

Joshua Wadler (MPO)
A Characterization of Turbulent Kinetic Energy and Its Generation
as Revealed in Idealized Tropical Cyclone Simulations

The distribution of turbulent kinetic energy (TKE) and its budget terms is analyzed in simulated tropical cyclones of various intensities that were realized through different ocean profiles in model initialization. For each oceanic profile, the atmospheric simulations were performed with and without TKE advection. Simulations with TKE advection create storms with an intensity ~5 ms–1 stronger than those without TKE advection, although overall wind field distributions were similar. TKE advection contributed negatively TKE in the inflow layer, but contributed positively in the eyewall. In all simulations, the TKE is maximized at low levels (i.e. < 1 km) and ~ 0.5 km radially inward of the azimuthal-mean radius of maximum wind speed at 1 km height. The TKE decreases with height in the eyewall, but more abruptly in simulations that do not include TKE advection. The two largest TKE budget terms are shear generation and dissipation, though variability in vertical turbulent transport and buoyancy production affect the net TKE tendency. In the eyewall, there is a net production of TKE below 500 m, but destruction of TKE above. The general relationships between the TKE budget terms remains consistent with radius. In terms of the asymmetric distribution, TKE is maximized in the front-right quadrant where the sea surface temperature (SST) is largest, and is minimized in the rear-right quadrant where the SST is the lowest. This TKE asymmetry is mainly tied to the asymmetry of air-sea enthalpy fluxes and low-level gradient in virtual potential temperature. These results may be utilized to target turbulence measurements in tropical cyclones using unmanned aircraft.


Andrew Smith (MPO)
Enthalpy Transfer and Energy Dissipation Observations
Adjacent to a Changing Sea Surface Structure

Within the study of air-sea interactions, the effectiveness or efficiency of transfers of energy between the atmosphere and ocean is a useful quantity in the context of upward-scaling phenomena (e.g. tropical cyclones, mesoscale convective systems, climate, etc.). In 2004, Donelan et al. conducted a seminal experiment in the Air-Sea Interaction Saltwater Tank (ASIST) at the University of Miami Rosenstiel School of Marine and Atmospheric Science to investigate whether the effectiveness of momentum transfer, described by the drag coefficient CD, which had been found to increase with wind speed, could be extrapolated to extreme wind conditions. I will present results from a subsequent experiment conducted in late September – early October this year in ASIST. This follow-up experiment was designed to confirm the Donelan et al. (2004) findings and repeat the conditions of their study in conditions of swell as well as the original conditions of wind-sea. Specifically, I have analyzed sonic and hot-film anemometer, laser slope gauge, 3D current profiler, and infrared (IR) camera data to (1) compare air and water-side turbulent kinetic energy (TKE) dissipation rates in wind-sea and swell conditions, (2) estimate the enthalpy transfer coefficient in these varied sea surface conditions as wind speed increases, and (3) to determine if the sea surface temperature structure captured by the IR camera provides insights on how the sea surface structure changes through roughening and breaking influences distribution of heat along the surface skin.

Rebecca Evans (ATM)
Diurnal Signals in the Hurricane Nature Runs

In the presence of an active ocean, North Atlantic sea-surface temperatures (SST) exhibit a lagged response to atmospheric forcing, in both models and observations. However, variable ocean currents are not necessary to reproduce the pattern and statistics of the Atlantic Multidecadal Oscillation (AMO). We examine these seemingly contradictory claims by estimating the magnitude and contribution of this ocean dynamics mechanism to multidecadal North Atlantic SST variability, in observations, observational products, and climate models. When testing a null hypothesis that accounts for spurious signals introduced by low-pass filtering, we find that the ocean response to the NAO is limited to the sub-polar gyre. Further, the lagged SST response to the NAO is small in magnitude and offers a ¬limited contribution to the AMO pattern, statistics, or predictability. We conclude that it is not necessary to invoke the full overturning of the Atlantic Ocean to explain this ocean dynamical response to the NAO.


Anne Barkley (ATM)
Deposition of Soluble Phosphorus to the Amazon from
Long-Range Transport of African Biomass Burning Emissions

Biologically mediated carbon sinks like the Amazon Rainforest play a vital role in the global carbon cycle. Despite a constant release of nutrients through plant decomposition, the soil stays highly phosphorus (P) depleted. The long-range atmospheric transport and deposition of P-rich aerosols is thought to help replenish P loss and ultimately sustain primary productivity rates that drive the drawdown of carbon dioxide. However, there is a lack of long-term measurements of P in aerosols collected in the Amazon and little is known about the seasonality, solubility, and long-range transported aerosol sources. This study provides the first ground-based seasonal measurements of transported aerosol and its associated total phosphorus (TP) and soluble reactive phosphorus (SRP) concentrations that provide insight into the aerosol sources of soluble P and are used to estimate the seasonal cycle of P deposition to the Amazon Basin. Most TP transport was due to the long-range transport of mineral dust. Enhanced P solubility was sometimes found, likely due to the co-transport of biomass burning aerosol from the Sahel with Saharan dust. Interestingly, elevated TP and SRP transport also occurred during boreal fall (September to November) when dust transport is not active. Using remote sensing data from the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP), we attribute the unexpectedly high SRP concentrations in the fall to the long-range transport of P-rich biomass burning aerosol from southern Africa. This is the first study to link P-rich biomass burning aerosol in southern Africa to increased TP and SRP transport to the Amazon. Implications for the global carbon cycle and ecosystem health are discussed.

Kurt Hansen (ATM)
The Genesis Potential Index on Sub-Seasonal Timescales

Previous studies such as Vitart 2010 have shown that dynamical subseasonal models have been able to skillfully predict tropical cyclone activity on basin wide scales more than two weeks in advance; either through a purely statistical approach or by assessing model produced tropical cyclone activity. This study hopes to find meaningful skill in subseasonal predictions of atmospheric (and oceanic) conditions such as vorticity, shear, humidity and potential intensity that are then related to tropical cyclone activity through the genesis potential index (GP). First the variability of the parameters, GP and tropical cyclone activity are quantified for subseasonal timescales. Then the correlation is calculated between GP and individual parameters, and various measures of tropical cyclone activity. This is in hopes of quantifying subseasonal tropical cyclone activity and discerning which variables are most closely associated with it. Ultimately model forecasts of GP in the SUBX project will be compared to re-analysis GP and observed tropical cyclone activity to determine how much skill is in these forecasts and for how long.

Shun-Nan Wu (MPO)
Prediction of Tropical Cyclone Intensification Rates
Using Satellite Measurements and Model Simulations

The relationship between the rate of tropical cyclone (TC) intensification and the amount and distribution of frozen hydrometeors is examined using satellite observations and numerical model simulations. Previous observational studies have demonstrated an overall enhancement of ice water content (IWC) for intensifying TCs, with the largest increases near the TC center. Such a signature of TC intensification is consistent with theoretical studies suggesting that latent heating within eyewall can effectively strengthen TC intensity. However, the relationship among different rates of TC intensification remains unclear due, in part, to the lack of measurements with consistent coverage of TCs to standardize the comparison of IWC. To investigate this issue, we used an idealized WRF model to examine the relationship between IWC and TC intensification rate. In numerical simulations, intensifying TCs have ~20% higher IWC within 200 km of their TC center than weakening TCs. The signature of TC intensification in the WRF simulations is consistent with that obtained from satellite observations, both in amount and spatial extent. Such consistency encourages the use of numerically based metrics to further investigate the relationship between IWC and TC intensification rates. Moreover, such relationships may serve as guidance to predict future TC intensification rates.


Hanjing Dai (AMP)
Surface Oil Film Effects on Short Wind-Generated Waves

A laboratory experiment was conducted to investigate the generation of waves by a turbulent wind at the surface with and without oil films in the SUSTAIN tank. The background short gravity waves with a JONSWAP spectrum were generated by a 12-element directional wave maker. Surface elevations were recorded by an ultrasonic distance sensor at a fetch of x = 72 cm. The spatial-temporal structure of the surface slope was analyzed by an optical method. The Polarimetric Imaging System had a field of view of 43.6 cm × 49.1 cm at fetch x Î [118, 162] cm, which permits wave-phase-resolved observation of surface waves. The PDF of the surface slope, wavenumber power spectral density, and mean square slope were calculated. At low wind speed, U10 = 8.5 m/s, the growth of gravity-capillary waves was not observed with oil films, because of viscous damping, larger surface tension and film elasticity. At high wind speed, U10 > 20 m/s, the spatially growing transverse waves become the dominant feature with the notable presence of group streaks. Interestingly, at even higher wind speed, U10 > 45 m/s, rapid local-damped waves are excited by the turbulent pressure field, however the growth of waves was apparently suppressed. Comparing the wind-generated surface waves with and without oil films helps us to gain better insight into the surface response to the pressure field and viscous damping.

Heather Hunter (AMP)
Automated Object Detection and Classification in High-Resolution SAR Imagery
Using Region Proposals and Convolutional Neural Networks

Automated detection of objects in satellite images is an on-going challenge, especially for synthetic aperture radar (SAR). Current state-of-the-art object detection methods based on convolutional neural networks (CNNs) have produced good results in the recognition and classification of targets in spaceborne optical sensor images, but are much less exploited for such tasks in SAR images. These methods use algorithms which first hypothesize the location of an object of interest, then provide the classification model with pixel coordinates that represent the potential bounding box of the object. Object detection in SAR is typically performed separately from classification by using a constant false alarm rate (CFAR) algorithm, which distinguishes the target from the background, and classification is subsequently performed on a small portion of the image centered on the target. SAR images, however, are very large – much larger than the images on which most state-of-the-art CNNs are designed to operate – and techniques that depend on sliding windows to search for objects, like CFAR, result in long run times. Additionally, the complex scattering mechanisms of a radar signal can preclude target recognition by classical computer vision techniques. This work applies a combination of a region proposal network (RPN) and CNN, called Faster R-CNN, to detect and classify specific targets in high-resolution SAR imagery. Results of this work show that a Faster R-CNN, trained on both optical and SAR images, can be successfully applied to SAR images for detection and classification of objects that vary in size and are imaged from different satellite orbits.

Lisa Nyman (AMP)
Using Doppler Marine Radar to Detect Surface Currents
and an Application to Internal Waves

During three research cruises in the Pacific in 2016 and 2017, members of the University of Miami remote sensing team operated Helmholtz-Zentrum Geesthacht's rotating X-Band, VV polarization, Doppler Marine Radar (DMR) onboard the R/V Roger Revelle and the R/V Oceanus. Ocean currents with marine radar are typically acquired with the use of the dispersion shell technique, which measures the velocity of the waves and compares it to the theoretical wave velocity in absence of a current. However, dispersion shell currents cannot resolve fine-scale current variations, such as the interface of a current front. To fill this gap, the coherent marine radar data were used in a different way. The raw data from the DMR can be converted to phase information, which can be used to extract radial Doppler velocities at a resolution of 7.5 m in range and 0.09° in azimuth. The Doppler velocity itself includes contributions from a multitude of ocean surface processes. When radial Doppler velocity from a full rotation of the radar is plotted with respect to radar look direction, it has a very obvious sinusoidal shape because the radar sees the strongest positive Doppler velocity when looking in one particular direction and the strongest negative velocity when looking in the opposite direction. Therefore, a sinusoid is fitted to the data every 1.5 minutes, which can then be converted into meridional and zonal velocities. From this, a linear regression was set up to determine a functional relationship between Doppler velocity and surface current. Furthermore, the Doppler Velocity Synthesis algorithm (DoVeS) was developed to synthesize radial surface current information from a moving ship, essentially fitting a piece of a sinusoid from radial Doppler data into small boxes paired with the linear regression to get a 2D vector surface current field around the ship track. Finally, to showcase the fine-scale capabilities of this method, the surface convergence and divergence from internal waves off the coast of California is compared with echo sounder data, which shows the shape of the internal wave under the ship.


Nektaria Ntaganou (MPO)
The Interaction of West Florida Shelf Anticyclones (WFAs)
with the Gulf of Mexico Mesoscale Activity

The Gulf of Mexico (GoM) is a region with high mesoscale variability that is characterized by the Loop Current (LC) and associated eddies. The features that mostly contribute to the LC variability are the anticyclonic Loop Current Eddy (LCE), cyclonic Loop Current Frontal Eddies (LCFEs), and Cuba Anticyclones (CubANs). In addition, we identify a certain type of anticyclonic eddies, which have not been fully investigated before, that are shed from the LC after it impinges on the West Florida Shelf and propagate following the shelf isobaths. We seek to investigate the contribution of these West Florida Anticyclones (WFAs) to the overall LC variability. High-resolution numerical simulations as well as satellite observations were used to identify cases of WFA formation and detachment for the period between 2010 and 2016. Favorable conditions for WFA formation are identified and connected to cyclonic LCFE activity, and the LC and Florida Current (FC) extensions. By combining several processes that take place in the GoM in each of the WFA cases, a possible connection between the cyclonic variability in the Florida Straits and the WFA cases is suggested.

Jeremy Klavans (MPO)
Variable External Forcing Obscures the Weak Relationship Between the NAO
and North Atlantic Multi-Decadal SST Variability

North Atlantic sea-surface temperatures (SST) exhibit a lagged response to the North Atlantic Oscillation (NAO) in both models and observations, which has previously been attributed to changes in ocean heat transport. However, variable ocean heat transport is only one of a few components of the Atlantic Multidecadal Variability (AMV) and has not been conclusively linked to the AMV in observations. We therefore examine the lagged relationship between the NAO and AMV in the context of the two other major components of the AMV, atmospheric noise and external forcing. In pre-industrial control runs, we generally find that the SST response to the NAO is only statistically significant in the subpolar gyre, when accounting for spurious signals introduced by filtering. Further, the lagged SST response to the NAO is small in magnitude and offers a limited contribution to the AMV pattern, statistics, or predictability. When climate models include variable external forcing, the relationship between the NAO and AMV becomes inconsistent; any lagged relationship that exists between the NAO and AMV become obscured. In these historically forced runs, knowledge of the prior NAO no longer offers predictability. Inconsistency in historically-forced ensembles allow for the possibility that the observed NAO-AMV relationship is due to chance and adds evidence that external forcing plays a meaningful role in producing the AMV.

Mariana Bernardi Bif (OCE)
Ocean-Atmospheric Oscillations Impact Carbon Production
in the North Pacific Ocean

Here, we assessed data from Bio-Argo floats equipped with nitrate sensors deployed in the Northeast Pacific Ocean between 2009 and 2018 in order to (1) estimate high-resolution net community production (NCP) in mol C m–2; and (2) evaluate the impact of oceanic-atmospheric oscillations, such as El Niño Southern Oscillation (ENSO) and the warm anomaly known as the Blob, on the magnitude of carbon production. The climate oscillations occurred between 2013-2016 and induced positive surface temperature anomalies, resulting in strong stratification that persisted through the winter months. The restriction of vertical mixing combined with a deeper halocline limited the nutrient renewal in the euphotic layer, and impacted organic carbon production in the period. We show that the Blob event, which started in October of 2013, amplified the 2014 ENSO, and the teleconnections added persistence to the anomalies in 2015. The highly-stratified winter of 2013 had no immediate impact in the magnitude of NCP, that reached ~1.6 mol C m–2 in the 2014 spring (June) and ~2.2 mol C m–2 in the summer (August). However, the nutrient renewal restriction for another consecutive winter, here as 2014, prevented carbon production in the following 2015 spring, as NCP = –0.02 mol C m–2. As a consequence, new production was delayed towards summer, and reached the lowest magnitude among all analyzed years at ~0.3 mol C m–2. The vertical nutrient exchange was subsequently restored in the winter of 2015, and NCP reached 0.6 mol C m–2 and 2 mol C m–2 in spring and summer of 2016, respectively. A composite analysis showed significant concurrent and delayed time between the development of warm anomalies and NCP, in agreement with our observations. This study highlights the sensitivity of marine biogeochemistry to extreme events such as ocean-atmospheric oscillations, and give us insights on the possible consequences of a warming ocean to carbon sequestration.

Nov 23: NO SEMINAR (Thanksgiving Recess)


Samantha Kramer (MPO)
Seasonal Variability of Saharan Dust Transport

In the summer months Saharan dust aerosols reach the southeast United States, affecting sea surface temperatures, radiative balance, and human health. Downwind dust concentrations are a function of emission, deposition, and transport. While dust emissions are often the focus of research studies, discriminating and attributing dust variability to transport versus emissions has yet to be calculated in a meaningful manner. Understanding seasonal and sub-seasonal variability of dust concentrations requires identifying wind flow patterns across the tropical Atlantic which move Saharan Air Layers (SAL) efficiently once emission has taken place. Daily wind changes over the ten-day crossing from the African coast to the southeast United States can alter the dust concentrations downwind. In order to address multiple temporal scales, the Hybrid Single Particle Lagrangian Integrated Trajectory Model [HYSPLIT] is used to analyze transport patterns. HYSPLIT backward trajectories are run from Miami, FL for 240 hours, twice per summer day between 1974-2016 using the NCEP Reanalysis. Analysis of these trajectories will shed light on dust transport, as they can be compared to the daily dust in-situ concentrations at the RSMAS sampling site. By quantifying dust transport efficiency further steps may be taken to determine if there is seasonal predictability for dust aerosols based on physical mechanisms and climate connections.

Kayleen McMonigal (MPO)
Temperature Variability of the Agulhas Current

The South Indian Ocean has heated rapidly in recent years. To understand this heating, we need to diagnose possible changes to the heat transport into and out of the basin. However, Indian Ocean heat transport across 32°S has only been calculated from observations at three points in time, leaving its variability on all time scales unknown. The Agulhas Current is the strongest and most variable current in the South Indian Ocean, and we expect that it may dominate the basin-wide heat transport variability. To begin to quantify the Agulhas Current temperature transport variability, we analyze the variability of 56 temperature records from a 26 month mooring deployment. These are the first temperature time series of the Agulhas Current. There is strong temperature variability on tidal, mesoscale, and seasonal time scales. Covariance length scales are 600–3000 dbar in the vertical and 60–110 km in the horizontal. Length scales are shorter near the surface and close to shore, where the current is strongest. Temperature variance is largest near the surface. In the future, we will combine the temperature data with velocity and acoustic travel time data to quantify the temperature transport of the Agulhas Current.

Marybeth Arcodia (ATM)
Tropical Intraseasonal Variability and the Effects on the Northern Mid-Latitudes

This study aims to understand and diagnose the teleconnection pattern associated with the Madden-Julian Oscillation (MJO) from Rossby wave forcing. "Active" MJO periods are defined using multivariate empirical orthogonal functions (EOF) of daily OLR and zonal wind vectors at 850 mb and 200 mb obtained from the NCAR-NCEP Reanalysis data set. Upper-level divergent flow from deep convection produces stationary Rossby waves which can extend into higher latitudes. Thus, a teleconnection associated with the tropical rainfall of the MJO is determined by a signal in the subsidence region, the subtropics. Composite analysis shows a rainfall perturbation in the United States, which indicates that the tropical-extratropical coupling has an influence in the North American region. ENSO and other low-frequency variability are strategically removed to highlight the teleconnection produced solely from the MJO. The physical processes driving this teleconnection are investigated and compared to those of ENSO. The MJO is further broken down by phase according to the Wheeler and Hendon (2004) Index, and the response associated with different phases is explored. North American rainfall is analyzed by phase during MJO active periods and combined with El Niño and La Niña separately to diagnose if the signals constructively or destructively interfere with each other. This research will help improve global sub-seasonal forecasts.


Tiago Bilo (MPO)
Interior Pathways of Labrador Sea Water in the North Atlantic
from the Argo Perspective

The pathways and transports of Labrador Sea Water (LSW) within the southward lower limb of the Atlantic Meridional Overturning Circulation are studied using 12 years of Argo profiles and subsurface Argo drift data. Consistent with previous studies, the results show clear evidence for interior pathways of LSW that separate from the western boundary near the Grand Banks and flow eastward and then southward around a large-scale deep anticyclonic gyre in the northern subtropical Atlantic. While most of the LSW exported into the interior recirculates in the Newfoundland Basin (7.3±2.0 Sv), approximately 2.6±0.2 Sv crosses the Mid-Atlantic Ridge and flows southward east of the Azores. The latter branch feeds a westward quasi-zonal pathway that recrosses the Ridge and returns to the western boundary around 30°N.