Request Info

Fridays at 11:00 am, Rosenstiel School Auditorium / Virtual Auditorium
(unless stated otherwise)

Jan 19: Dr. Allison Wing
Invited Speaker of the Department of Atmospheric Sciences
Werner A. and Shirley B. Baum Professor and Associate Professor
Department of Earth, Ocean and Atmospheric Science, Florida State University, Tallahassee

Acceleration of Tropical Cyclone Development by Cloud-Radiative Feedbacks
Recording Available at COMPASS ON DEMAND

A complete understanding of the development of tropical cyclones (TC) remains elusive, and forecasting TC intensification remains challenging. This motivates further research into the physical processes that govern TC development. I find that a "cloudy greenhouse effect" accelerates TC development, in which anomalous infrared warming in areas of deep thunderstorm clouds near the center of theTC drives rising motion in the storm, which helps moisten the atmosphere and aids in the formation of the TC's circulation. Improving the representation of cloud-radiative feedbacks in forecast models therefore has the potential to yield critical advancements in TC prediction, but this requires a better understanding of cloud-radiative feedbacks in observed TCs. Thus I also present ongoing work that investigates these processes using satellite (CloudSat) retrievals of cloud properties. We examine the vertical and radial structure of ice and liquid clouds and the implications of the heterogeneous cloud distribution for spatial variability in radiative heating rates. Radiative transfer calculations using the CloudSat cloud property retrievals indicate that most of the spatial variability in longwave heating rates comes from ice clouds, as in the idealized simulations. The longwave cloud radiative feedback is found to be strongest in the inner core of TCs undergoing rapid intensification.

Jan 26: Dr. Lichuan Wu
Guest of Peisen Tan, Department of Ocean Sciences
Department of Earth Sciences, Uppsala University, Sweden

Development of Polar Lows in Future Climate Scenarios Over the Barents Sea
Recording Available at COMPASS ON DEMAND

Polar lows (PLs) are intense mesoscale cyclones that form over polar oceans during colder months. Characterized by high wind speeds and heavy precipitation, they profoundly impact coastal communities, shipping, and offshore activities. Amid the substantial environmental changes in polar regions due to global warming, PLs are expected to undergo noteworthy transformations. In this study, we investigate the response of PL development in the Barents Sea to climate warming based on two representative PLs. Sensitivity experiments were conducted including the PLs in the present climate and the PLs in a pseudo-global warming scenario projected by the late 21st century for SSP 2-4.5 and SSP 3-7.0 scenarios from CMIP6. In both warming climate scenarios, there is an anticipated decrease in PL intensity, in terms of the maximum surface wind speed and minimum sea level pressure. Despite the foreseen increase in latent heat release in the future climate, contributing to the enhancement of PL intensity, other primary factors such as decreased baroclinic instability, heightened atmospheric static stability, and reduced overall surface heat fluxes play pivotal roles in the overall decrease in PL intensity in the Barents Sea under warming conditions. The augmentation of surface latent heat flux, however, results in increased precipitation associated with PLs by enhancing the latent heat release. Furthermore, the regional steering flow shifts in the warming climate can influence the trajectory of PLs during their development.

Feb 02: Dr. Qianjiang (Bryant) Xing
Department of Ocean Sciences, Rosenstiel School

How the Complexity of Continental Breakup Controls
Early Antarctic Circumpolar Current
Recording Available at COMPASS ON DEMAND

Around 30 million years ago, the globe underwent a climate transition from "greenhouse" to "icehouse" conditions, usually termed the Eocene-Oligocene transition (EOT). Coevally, Southern Ocean horizontal circulation changed from a two-gyre pattern with a weak proto-Antarctic Circumpolar Current (ACC) to a more modern regime of ACC dominance. The mechanisms driving the changes in Southern Ocean horizontal circulation through the EOT have been investigated for decades. Two main factors are emphasised in the paleoceanography community; the opening of ocean gateways and wind stress adjustments. However, previous studies show that neither deepening of ocean gateways nor wind stress modification alone obtain a strong proto-ACC with transport comparable to the modern ACC. Most past studies have focused on the effect of the gateway deepening and wind stress and / or have been run with a coarse-resolution model, with the parameterization of turbulent processes, such as mesoscale eddies. In addition, surface buoyancy (heat or salt) forcing, a key driver of deep-water formation, also likely controls the modern ACC, but has remained understudied in the context of EOT circulation changes. To understand the dynamics of the EOT ocean circulation changes, especially turbulent processes in the proto-ACC and their ability to transport heat, I use a high-resolution ocean model with realistic late Eocene bathymetry. These models reveal that when the maximum westerly winds align with both the deep Tasmania Gateway (TG) and Drake Passage (DP), the circumpolar proto-ACC initiates, with a maximum DP transport of 38.3 Sv, ~27% of the modern strength. Proto-ACC transport still fails to reach the value of modern ACC even under a doubled wind stress. However, when applying modern surface buoyancy (heat / salt) forcing, a vigorous proto-ACC with a DP transport of 139 Sv, closely aligning with modern ACC strength, is formed. I, therefore, hypothesise that deep convection due to change in surface buoyancy forcing is the primary mechanism driving the transition from a weak proto-ACC to the modern ACC provided that deep Southern Ocean gateways allow the weakening of subpolar gyres and onset of proto-ACC.

Feb 09: NO SEMINAR (Recruitment Weekend)

Feb 16: NO SEMINAR (Rosenstiel School Faculty Meeting)


James Christie (ATM)
Implications of the Shrinking Great Salt Lake on Regional Air Quality

The volume of the Great Salt Lake has decreased by ~50% exposing saline lakebed, increasing the emission of playa dust into the air – which is a known source of coarse particulate matter (PM) in the Salt Lake Valley. Current air quality models neglect the chemical reactions that are likely to occur on the surface of playa dust particles, which can lead to the formation of the criteria air pollutants fine PM and ozone. The nighttime surface reaction with dinitrogen pentoxide (N2O5) can lead to the production of nitryl chloride (ClNO2). In the daytime, ClNO2 photolyzes to produce chlorine radicals, which facilitates the formation of both ozone and fine PM. To better understand how playa dust from the Great Salt Lake can contribute to these criteria air pollutants, we measured the reactive uptake of N2O5 and production of ClNO2 on playa samples collected from different regions of the Great Salt Lake using an aerosol kinetic flow tube. We also identify potential chemical properties of these sediments to better understand what is affecting uptake chemistry. To determine the potential contribution of playa dust to observed ClNO2 concentrations near the lake, we also used a zero-dimensional box model to estimate ClNO2 production from playa dust during the 2017 UWFPS campaign, showing the importance of this reaction to ambient halogen production. This work highlights the need for air quality models to consider reactions which occur on the surface of playa dust when predicting ozone and / or fine PM pollution.

Alexis Wilson (ATM)
Influence of the Caribbean Low-Level Jet and Topography
on the Pre-Genesis Environment of Hurricane Ida (2021)

The Caribbean Low-Level Jet (CLLJ) represents an area of enhanced trade winds extending from the surface up to 600 mb. It is often cited to have a climatologically negative influence on tropical cyclogenesis in the Caribbean by enhancing vertical wind shear and low-level moisture divergence. However, few studies have examined the dynamical and thermodynamical impact the CLLJ has on individual cases on a convective to synoptic scale, including the potential to positively contribute to cyclogenesis. To test this, the strength of the CLLJ was examined from four days pre-genesis until genesis for all Caribbean genesis cases from 1972 to 2021. At 700 mb, the CLLJ was found to be significantly stronger than average but weakened leading up to genesis, particularly in the zonal direction. To exemplify this, a case study on Hurricane Ida (2021) using brightness temperature and reanalysis data revealed that genesis was preceded by a strong CLLJ that quickly weakened. Simultaneously, a weak tropical wave amplified, which was not captured in operational forecast models. Additionally, a topographically induced mesoscale convective system (MCS) offshore of South America formed 36 hours prior to genesis and was advected northward within the tropical wave. This MCS was sustained by low-level moisture convergence at the exit region of the weakening CLLJ and resulted in the genesis of Ida. These results demonstrate how the CLLJ and the local topography that supports it can positively contribute to cyclogenesis.


Rachel Sampson (MPO)
Analyzing Historical ARGO Float Data to Understand the Dynamics
of Agulhas Leakage in the Cape Cauldron

The Cape Cauldron is a region off the coast of South Africa near the Agulhas current retroflection. In this region, the subtropical Indian ocean waters (warmer, saltier) interact with the subtropical South Atlantic waters (cooler, fresher) and move into the South Atlantic basin. This injection of heat and salt is known as Agulhas leakage, and it is part of the upper arm of the Atlantic Meridional Overturning Circulation (AMOC). It has primarily been thought to be carried by large mesoscale eddies called Agulhas Rings. However, recent studies have shown that because of the turbulent nature of the cauldron, most of these rings can lose 1/3 of their heat and salt anomalies, indicating that the leakage may be occurring outside of these rings. For the first time in this region, using a combination of Eulerian and Lagrangian in-situ observations, we plan to observe and characterize the sub mesoscale features that are generated by the mesoscale strain field and estimate the lateral eddy heat and salt fluxes. To calculate these fluxes, we will need the eddy diffusivities and a background gradient for temperature and salinity. Using ARGO floats, spanning a 20-year period, a preliminary background gradient was calculated through the leakage corridor as a first step towards understanding the dynamics in the region, particularly, how these two water masses are interacting on varying neutral density levels.

Will Downs (ATM)
Tropical Surface Analysis Using Deep Learning

Tropical waves and the intertropical convergence zone (ITCZ) / monsoon trough impact human interests directly through wind and rainfall and indirectly through their relationship to tropical cyclogenesis. The Tropical Analysis and Forecast Branch (TAFB) of the National Hurricane Center has released text analyses of current weather features in the tropical North Atlantic and Northeast Pacific every six hours since 2003. The archive of TAFB discussions is thus a trove of information on significant tropical weather, though its text-based format and limited temporal record are obstacles to easily using this information in comprehensive research. We use tropical wave and ITCZ / monsoon trough locations extracted from TAFB's discussions and ERA5 reanalysis data to train two neural networks based on the U-Net 3+ architecture. One network is trained to identify tropical wave axes, while the other is trained to identify and distinguish between the ITCZ and monsoon trough. Both of our networks demonstrate significant skill at identifying the locations of their respective target objects when verifying against TAFB's feature labels. We present a climatology of wave and ITCZ / monsoon trough locations across the North Atlantic and East Pacific based off of our network outputs. The effectiveness of both of our networks suggests that they will serve as useful tools for studying different regimes of tropical cyclogenesis in the reanalysis era and that they may be a helpful tool for operational tropical surface analysis.

Samantha Nebylitsa (ATM)
Identifying Model Sensitivities to Tropical Cyclone Intensification:
Are Slowly Intensifying Storms Even Possible?

Maybe. The environment around tropical cyclones (TCs), mainly the wind shear and relative humidity, are known to greatly influence the intensification rate of TCs. The majority of previous studies that have investigated differences between rapid intensification (RI; ≥30 kts 24hr–1) and slower intensifying (SI) or non-intensifying storms have done so with highly simplified wind profiles. To produce more realistic profiles, this study uses average storm-relative profiles of environmental wind and moisture at onset for all RI and SI events between 1980-2021, based on ERA5 reanalysis data. Using the time-varying point downscaling technique within the Weather Research and Forecasting (WRF) Model, the environment around the idealized TC is forced to the corresponding RI and SI wind and moisture profiles. The forcing rates, in conjunction with other model initialization parameters, are adjusted to create a set of storms that undergo the appropriate slow and rapid intensification as expected from the profiles. Results indicate that idealized TCs are very susceptible to RI, even in SI conditions. Shear analysis demonstrates that the far environment around the TC is behaving as expected, and that other factors, such as storm tilt, may be used to explain the differences in timing of RI.


Junfei Xia (MPO)
Applications of Machine Learning in Oceanography:
Understanding and Annotating Sea Surface Imagery

Recent advancements in machine learning and artificial neural networks have significantly enhanced data analysis capabilities across various scientific fields, notably in ocean science. Our study specifically explores the application of machine learning algorithms to address complex oceanic data sets, aiming to streamline data preprocessing and improve data labeling accuracy. We investigate common pitfalls in applying machine learning to ocean science and propose methodologies to overcome these challenges, thereby enhancing model performance. A notable contribution of our research is the introduction and evaluation of the Vision Transformer (ViT) for image classification tasks, comparing its efficacy against traditional Convolutional Neural Networks (CNNs) using the Labeled SAR imagery dataset (TenGeoP-SARwv), encompassing ten geophysical phenomena. Our findings demonstrate that ViT significantly outperforms commonly used CNNs across all categories, achieving precision, recall(accuracy), and F1 scores above 0.95. The superior performance of ViT can be attributed to our strategic approach to mitigating dataset imbalance and the model's self-attention mechanism, which captures the global context of images more effectively.

Eric Mischell (ATM)
Two Approaches for Reducing Uncertainty in Projections of Regional Climate Change

For large-scale measures of climate change, like the globally-averaged response of precipitation, there is, in general, consistency between predictions from basic theory, sophisticated Earth system models, and observations. However, the response of precipitation in any particular region is subject to much greater uncertainty, with credible models disagreeing, for example, on the sign of the change over much of North America. This uncertainty arises not from differences in climate sensitivity between models, or from differences in the pattern of warming; rather it arises from differences in the climatological base state. In this presentation we outline two approaches, one dynamical and one statistical, for reducing the uncertainty of projections of regional climate change. The first exploits the relationship between short-term forecast errors in a model and the long-term climatological bias to correct the model's base state. This method assumes that model biases arise from "fast" processes in the model's parametrization, and that by introducing a bias correction, the base state error can be reduced. The second approach involves weighting models in an ensemble average according to their ability to simulate the present-day pattern of precipitation. Rather than assuming all models are independent and equally reliable, this method assigns optimal weights to models based on their degree of similarity to the observational record. Overall we show two complimentary approaches for reducing the uncertainty of projections of regional climate change, which we intend to explore further in the coming months.

Snigdha Samantaray (MPO)
Understanding the Convection in Water Vapor "Lakes"
in the Western Equatorial Indian Ocean

Over the Western Equatorial Indian Ocean (WEIO), isolated high-column Water Vapor "Lakes" stand out in stark contrast to the typically drier zones and often bring rain to the African coast. Our study explores the genesis and persistence of these vapor lakes. Central to this research is the hypothesis that vapor lake dynamics are maintained by a combination of horizontal advection, secondary atmospheric circulations, and Precipitation-Evaporation (P-E). We define these lakes by defining closed contours of 55mm column water vapor threshold in MERRA-2 data to examine the moisture budget as a function of distance from these contours. Preliminary findings indicate distinct seasonal landfalling patterns of vapor lakes following the warmest SST, but they are stronger in the southern hemisphere. Tracking studies of some vapor lakes show lifespans exceeding a week with behaviors like merging and splitting, and an east-to-west drift. The study's composite analysis indicates a distinct peak for clouds and precipitation within the confines of vapor lakes. Inside these lakes, we observed a marked moisture convergence corresponding to the P-E sink. These lakes also have vorticity signatures, raising the question of whether moisture or momentum fields are more important to their dynamics. Also, many other variables exhibit structure across the vapor lake boundary. Future work includes evaluating model forecasts and simulations with the aim of gaining deeper insights into longevity and propagation mechanisms of these systems, and the ability of the models to capture these.

Mar 15: NO SEMINAR (Spring Recess)


Luke Rosamond (ATM)
The Stratospheric Gravity Wave Field and Momentum Fluxes
Produced by Isolated Supercells

Luke Rosamond, David S. Nolan, Yi Dai, and Chris Heale

Alongside topographic forcing, deep moist convection makes a significant contribution to the global budget of upward momentum transport by gravity waves. Long-lived thunderstorms with rotating updrafts, known as supercells, produce strong and highly variable vertical motions over several hours. This study uses an idealized modeling framework in WRF to simulate supercells and their associated gravity waves up to 60 km altitude for multiple different wind profiles and convective modes. In contrast to many previous studies, the supercell is brought to an end and the simulations continue until most of the wave energy has dissipated. Thus, upward momentum transport can be computed over the entire life cycle of the storm and its associated waves, providing a more complete picture of the total impact of the event. The shapes of the wind profiles in the upper troposphere and lower stratosphere are found to strongly control the total momentum and energy transported into the upper stratosphere, so varying the stratospheric wind profile illuminates the behavior of the gravity waves in the stratosphere, particularly their vertical propagation. We also investigate the extent to which different modes of supercell structure, such as high-precipitation, low-precipitation, and classic supercells, lead to different intensities and spectra of the resulting gravity waves. In addition, the WRF model diabatic heating and vertical motions will be used as forcing conditions for stratospheric models such as MAGIC and CGCAM for the purposes of 1) comparison to WRF results between 20 and 60 km, and 2) so that wave propagation, momentum transport, wave breaking, and momentum deposition can be evaluated to altitudes above 80 km.

Hope Elliott (OCE)
Iron in Volcanic Ash: Iron-Specific Mineralogy Explains Solubility

Hope Elliott1, Cassandra J. Gaston1, Edmund Blades1, Haley M. Royer1, Amanda M. Oehlert1, Ravi Kukkadapu2, Swarup China2, Zezhen Cheng2, Nurun Nahar Lata2, Clifton Buck3, Charlotte Kollman3, Esteban Gazel4, Adrian Hornby4, Andrew Ault5

1Rosenstiel School of Marine, Atmospheric & Earth Science, University of Miami, FL
2Pacific Northwest National Laboratory, Richland, WA
3Skidaway Institute of Oceanography, University of Georgia, GA
4Cornell University, Ithaca, NY
5University of Michigan, Ann Arbor, MI

Since volcanic material is rich in iron (Fe), deposition of volcanic ash into the ocean is thought to stimulate marine productivity and promote carbon sequestration though relieving Fe limitation specifically. However, whether or not ash-derived Fe is soluble and available for uptake by marine microbes remains unclear, and ash from different eruptions varies extensively in composition and presumably solubility. We measured Fe solubility in volcanic ash collected after the 2021 eruptions of the La Soufrière volcano and the Cumbre Vieja volcanic ridge. We also used single-particle and bulk techniques to gather information about morphology and composition of these two types of ash. Our analyses showed that the Cumbre Vieja ash contains more soluble Fe than the La Soufrière ash, and Fe-specific mineralogy can explain the observed differences in solubility. The Cumbre Vieja ash contains more total Fe, and a larger percentage of that Fe is contained in silicate minerals where it is likely more soluble than when locked up in other Fe-bearing minerals like oxides. For both ash types, Fe is concentrated in specific regions on particle surfaces rather than uniformly coating them. However, Cumbre Vieja ash particles also have elevated levels of surficial fluorine (F), which may indicate that acid processing has further enhanced Fe solubility. Overall, we estimate that Fe in both types of ash is <5% soluble. Therefore, volcanic ash does not appear to release significantly more soluble Fe per unit weight than other types of natural aerosols like mineral dust.

Paloma Cartwright (OCE)
Understanding Variations of Water Masses in the Florida Current

The Florida Current marks the beginning of the Gulf Stream, flowing in the narrow channel between Florida and The Bahamas. It constitutes the major western boundary current of the North Atlantic Ocean and acts as a vital limb of the Atlantic Meridional Overturning Circulation. It also closes the wind-driven interior gyre of the subtropical North Atlantic. Because of the interconnectedness of the Florida Current, its water masses are not all of local origin. This analysis uses Conductivity, Temperature, Depth (CTD) and Lowered Acoustic Doppler Current Profiler (ADCP) data from the NOAA Western Boundary Time Series to observe the variations in water mass properties of the Florida Current. Preliminary findings indicate the Florida Current can be split into three different water masses using temperature, salinity and potential density categorizations. Surface waters in the Florida Straits are above the 24.4 kg/m3 isopycnal and highly susceptible to surface fording. The waters between the 24.4 and 26.8 kg/m3 isopycnals are subtropical and below the 26.8 isopycnal are Antarctic Intermediate waters. Further analysis highlights the seasonal variability of these water masses.

Mar 29: Jamie Meacham
Imperial College London, UK

Clustering of Floating Tracer in Quasigeostrophic Coherent Structures
Recording Available at COMPASS ON DEMAND

Observations of dense aggregations of marine flotsam, such as microplastics, motivates investigation of the dynamics of buoyant material in ocean currents. We focus on microscopic objects which resist vertical motion due to positive buoyancy but are otherwise passively advected. In this talk, a weakly ageostrophic flow with high spatiotemporal resolution is derived from a two-layer quasigeostrophic forced-dissipative turbulence model. We introduce surface Lagrangian tracer particles to this flow, then track the formation of clusters and voids.

Even in this low Rossby number context, there is fast and persistent clustering. The distribution of floating tracer is found to be almost entirely determined by the strongest coherent structures of the flow, as detected through Lagrangian Coherent Structure metrics. The coherent vortices are observed to transport as much as 100 times more floating tracer than passive tracer, due to the formation of dense clusters on the vortex cores. We find that LAVD (Lagrangian Averaged Vorticity Deviation) is particularly effective at identifying clusters and argue that this is a natural consequence of the dynamics. This result has implications for inertial tracer dynamics and biogeochemical models, suggesting that proper representation of the clustering process will be significant in many ocean modelling problems of urgent interest.


Steven Akin (MPO)
Incorporating the Coastal Ocean Into Hurricane Forecast Models

Hurricane Michael (2018) experienced two distinct rapid intensification (RI) events in the Gulf of Mexico before slamming into the Florida panhandle, making it the first landfalling CAT 5 (maximum sustained winds of 140 knots) in the North Atlantic since Andrew (1992). Michael's second intensification event is particularly curious because it occurred over the West Florida Shelf (WFS), where the water was less than 100 meters deep and OHC levels were relatively low. The Upper Ocean Dynamics Laboratory deployed three EM-APEX floats (accounting for 668 vertical profiles) and 137 airborne expendable ocean probes into Michael. Our dataset is rounded out with observations from Wide Swath Radar Altimeters (WSRA), Stepped Frequency Microwave Radiometers (SFMR), ARGO floats, and a suite of drifters. With such a robust dataset, we can identify 3-D velocity, temperature, and salinity as a function of pressure across the full cycle of the storm. The EM-APEX floats captured the ocean response during the first RI event, while the expendable probes captured air-sea interactions during the second event at landfall. There are serious gaps in our understanding of how hurricanes and the coastal ocean respond to each other. This dataset will enable us to make the first step toward incorporating the coastal ocean into hurricane forecast models.

Elizabeth Yanuskiewicz (OCE)
Isotopic Evidence for Alteration Pathways of Particulate Organic Matter
Under Contrasting Productivity Regimes

Particulate organic matter (POM) that is exported from the ocean surface is a critical food source for mesopelagic food webs and has the potential to contribute to carbon sequestration in the deep ocean. Predictive understanding of oceanographic processes that influence carbon export and sequestration is a main goal of NASA's EXport Processes in the Ocean from RemoTe Sensing (EXPORTS) program. As part of the EXPORTS program, I used nitrogen isotopic indices from amino acids extracted from size-fractionated and sinking POM at surface to mesopelagic depths to identify how POM is altered and its flux is attenuated in distinct ocean environments. Working on material collected from the northeast Atlantic during the spring phytoplankton bloom, I found that a greater proportion of fresh, phytoplankton biomass was found to be exported into the mesopelagic, when compared to results from the northeast Pacific (Wojtal et al., 2023). Significant microbial alteration of POM is evident in northeast Atlantic sinking particles, along with temporal changes to POM composition most likely related to bloom dissipation. These findings prompt my next work, using additional organic characterization techniques to examine the degradation of POM. I will also describe how my results here have motivated me to assemble a global dataset of chemical classification of POM to better understand export efficiencies in different oceanic conditions.

Tyler Tatro (ATM)
More Biomass Burning Aerosol is Advecting Westward
Over the Southern Tropical Atlantic in the Past 20 Years

Each year, agricultural fires in southern continental Africa emit approximately one third of the world's biomass burning aerosol, which is advected westward by free tropospheric winds over a semipermanent, subtropical stratocumulus cloud deck. The radiative effects from aerosol-cloud interactions impact regional circulations and hydrology. Here we examine how changes in the coupled southern African earth system over the past 20 years impact the southeast Atlantic. We combine satellite-derived burned area datasets with ECMWF carbon monoxide, black carbon, and meteorological reanalysis data to analyze recent trends in the biomass burning season (May-October) in southern Africa. The burning season begins in May in woody savannas in the northwest and shifts to open savanna and grassland fires in the southeast, with small fires (less than 1 km2) contributing significantly to total burned area. More small fires are occurring in the middle of the biomass burning season and the overall season is shorter, corroborated by carbon monoxide fields from reanalysis. Significantly increased free tropospheric winds, shifted southward, transport smoke aerosol further southwest over the southeast Atlantic. The increased aerosol advection is coupled with a southern shift in the south Atlantic subtropical high and an increase in the low cloud fraction on the southern edge of the stratocumulus cloud deck. While smoke emissions sources have not changed significantly, changes in the smoke transport pathway, attributed to increasing surface temperatures in southern Africa and tropical expansion, combined with more low cloud, are also altering the regional radiation balance.


Aidan Mahoney (ATM)
Tropical Cyclone Potential Intensity in Historical and Future Climates

Potential Intensity (PI) is the theoretical maximum sustained wind speed and minimum central pressure for a tropical cyclone (TC), assuming idealized and perfect environmental conditions. PI is computed from the sea surface temperature (SST), mean sea level pressure (MSLP), and profiles of temperature and moisture through the troposphere. It is best interpreted as a climatological parameter, with higher PIs indicating favorable thermodynamics capable of supporting stronger TCs. In the real world, observed TCs rarely attain their respective PI due to chronically imperfect atmospheric dynamics. The exceptions are intense TCs (Saffir-Simpson Category 4+), which exist in highly favorable environments and more closely approach their PI. These historically strongest TCs are projected to be stronger in a warmed climate due to more favorable thermodynamic environments; these are captured by positive trends in PI in both the historical record and future projections. The goal of this work is to assess PI across reanalysis datasets, model runs forced by the historical climate, and model runs of future climates. The AM2.5C360 is an atmosphere-only model with a control simulation forced by historical SSTs (CTRL) and a simulation with uniform warming of 2 K (plus2K). HighResMIP is a high-resolution protocol of the CMIP6 family of climate models, and includes historical simulations forced by observed SSTs. We find an inter-model spread in PI within these historical HighResMIP simulations that is larger than the anthropogenically forced change in PI from AM2.5C360 CTRL to plus2K. The spread in PI is a result of the spread in MSLP; all other inputs to PI are in strong agreement across all datasets.

Madeleine Dawson (OCE)
Deep Learning for Island Wake Feature Extraction and Segmentation
in Sentinel-1 Imagery

With the rapid growth of remote sensing platforms and increased data availability, Synthetic Aperture Radar (SAR) has become a premier remote sensing system to advance understanding of ocean and atmospheric phenomena. Our study focused on the usage of SAR Sentinel-1 imagery, sourced from NASA Earthdata and ESA repositories, to advance island wake parameterization through machine learning. Island wakes exhibit significant dynamic variability, and traditional methods of SAR image analysis require substantial manual inspection to accurately classify their structure. Therefore, deep learning models are appropriate to streamline evaluation. Deep learning models can efficiently handle SAR imagery, often spanning multiple gigabytes per image. The models we utilized are U-Net and a modified U-Net architecture. The U-Net model, known for its effectiveness in noisy biomedical image segmentation, is adapted here to SAR imagery, also known for inherent speckle noise. These models operate through an encoder-decoder architecture with skip connections, allowing it to capture both local and global features. Both models achieved average F1 score and accuracy above 0.89 and an average IoU (Intersection over Union) score of 0.68, thus demonstrating successful deployment. Future work includes further hyperparameter tuning, and the addition of multi-resolution training into the model architecture for improved feature extraction and segmentation accuracy.

Samantha Medina (OCE)
Intercomparison of In-Situ Sensors for Wave Spectra Parameters in Coastal Zones

Current operational forecast models for air-sea exchanges rely on drag coefficient parameterizations derived from open-water observations and Monin-Obukhov similarity theory (MOST). In coastal zones, the transition between the marine and land boundary layers leads to the development of spatial heterogeneity of the wind and wave fields. Previous parameterizations fail to account for nearshore spatial variability as well as the breakdown of MOST. The Coastal Land-Air-Sea Interaction (CLASI) project's goal is to develop parameterizations of air-sea fluxes through direct measurements of near-shore and onshore conditions in Monterey Bay, California, using Air-Sea Interaction Spar (ASIS) buoys, Inner-shelf Spar (I-Spar) buoys, and Spotter buoys. We will use sea surface height observations from wave staffs mounted on ASIS as well as Spotter buoys along key transects in Monterey Bay in various wind and wave conditions. We will then use these measurements to examine the relationship between spatial variability within the bay and directional wave spectra. Once better understood, this can be used to improve parameterizations for air-sea momentum fluxes in coastal zones.


Chanyoung Park (ATM)
Constraining Effective Radiative Forcing From Aerosol-Cloud Interactions
Using Activation Rate From Satellite Observations

The effective radiative forcing due to aerosol-cloud interactions (ERFaci) is difficult to quantify, leading to large uncertainties in model projections of historical forcing and climate sensitivity. In this study, satellite observations are used to examine the low-level cloud radiative responses to aerosols. While it is commonly assumed that the activation rate of cloud droplet number concentration (Nd) in response to variations in sulfate aerosols (SO4) or the aerosol index (AI) has a one-to-one relationship in the ERFaci estimation, we find this assumption to be incorrect, and explicitly accounting for the activation rate is crucial for accurate ERFaci estimation. Our research suggests a smaller and less uncertain value of the global ERFaci than previous studies (–0.39±0.28 W m–2 for SO4 and –0.24±0.18 W m–2 for AI, 90% confidence), indicating that ERFaci may be less impactful than previously thought. This finding is supported by three different methods to estimate ERFaci from CMIP6 model simulations. Our results are also consistent with observationally constrained estimates of total cloud feedback and 'top-down' estimates that models with weaker ERFaci better match the observed hemispheric warming asymmetry over the historical period.

Caitlin Collins (ATM)
Winds of Change: A Review of the ENSO Predictability Debate

The El Niño-Southern Oscillation (ENSO) is the dominant mode of climate variability in the tropical Pacific but plays an essential role in modulating the global climate through its numerous teleconnections. As a coupled phenomenon, ENSO is implicitly multivariate. Despite extensive research, debate persists regarding which event initiation mechanism is dominant and hence there remains a lack of consensus in its predictability. The challenge is that the unpredictable component of ENSO stems from its sensitivity to stochastic processes. Moreover, ENSO predictability estimates are influenced by, and rooted in, different theoretical frameworks. Through systematic experimental design, state-of-the-art coupled models provide an avenue to elucidate the relative role of the subsurface preconditioning, unstable coupled instabilities, and stochastic processes that shape ENSO dynamics and predictability. This presentation aims to synthesize the current literature, highlighting open questions that motivates the proposed research focusing on how uncoupled atmospheric noise affects predictability estimates with and without subsurface preconditioning.

Gabrielle Ricche (OCE)
Detection and Monitoring of Iceberg Drift Velocity Utilizing SAR Imagery

As the climate continues to warm, the Arctic region has experienced an increase in the formation and dissolution of icebergs, posing significant hazards to maritime navigation. Additionally, Arctic icebergs significantly influence ocean dynamics and have biogeochemical implications on the surrounding environments. Because they can persist for several years and travel large distances, it is important to better understand how iceberg drift velocity and trajectory are impacted by external forcing conditions, such as wind, currents, and waves. While recent studies have explored this issue through the use of dynamic models of iceberg drift coupled with direct observations of Arctic icebergs, our study employs high-resolution Synthetic Aperture Radar (SAR) and other remote sensing techniques to detect, quantify, and track different types of icebergs in the Arctic region. This will be accomplished, in part, through the utilization of the AI4Arctic / ASIP Sea Ice Dataset - version 2 (ASID-v2), which contains Sentinel-1 SAR scenes matched with ice charts, produced manually by the Danish Meteorological Institute in 2018-2019. Marine surface winds and currents will also be extracted from SAR, along with iceberg volumetric and geometric parameters. This information will then be used to compute iceberg drift velocity vectors.

Apr 26: Sisam Shrestha
Department of Atmospheric Sciences, Rosenstiel School
(1-hour ATM student seminar)

Analysis of Changes in Large-Scale Atmospheric Circulation
During the Satellite Era
Recording Available at COMPASS ON DEMAND

Large-scale atmospheric circulations play a crucial role in determining the spatial distribution of clouds and precipitation. Thermodynamical constraint on the hydrological cycle predicts a weakening of the atmospheric circulation with warming. In coupled climate models, this primarily manifests as a weakening of the Walker Circulation (WC), a robust response across the CMIP6 coupled models for the 21st century. However, observational datasets over the past several decades indicate a strengthening of the Pacific WC. This discrepancy between observations and coupled climate models raise questions about the models' ability to represent critical energetic and hydrologic constraints responsible for the predicted weakening. Additionally, there is discrepancy among studies regarding the response of the Intertropical Convergence Zone (ITCZ), an intense band of precipitation along the rising branch of the Hadley circulation, to a warming climate. Past works have shown a southward shift of the ITCZ during the second half of the 20th century attributed to indirect effects of anthropogenic aerosols. With a decrease in anthropogenic aerosols following the Clean Air Act, the ITCZ would be expected to migrate north. However, studies analyzing the past few decades find insignificant shift in the ITCZ location. Thus, in this work, we leverage a growing record of satellite observations and climate models to ascertain the validity of the thermodynamical constraint in a warming climate and analyze the trend in ITCZ location during the satellite era. We find that model simulations with either observed or model-projected warming patterns predict a robust weakening of the atmospheric circulation, despite having opposing changes in the Pacific WC strength. More importantly, weakening inferred from satellite observations is reproduced in coupled models only when anthropogenic forcing is included, suggesting that a human-induced weakening of the global atmospheric circulation is already detectable in observations. Regarding zonal-mean ITCZ, CMIP6 coupled models show a northward migration in recent years in agreement with the ITCZ shift inferred from the Global Precipitation Climatology Project dataset. Using single-forcing experiments, we find this northward ITCZ shift to be similar in magnitude for both aerosol-only and GHG-only experiments, consistent with a similar magnitude of increase in northward cross-equatorial atmospheric energy transport.


Sep 06 (11:15 am, Zoom): Dr. Chandra Venkataraman
Guest of the Department of Chemical, Environmental, and Materials Engineering
Department of Chemical Engineering, Indian Institute of Technology Bombay
(Zoom broadcast from Frost Institute for Chemical and Molecular Science, UM main campus)