FALL 2024
Fridays at 11:00 am, Rosenstiel School Auditorium / Virtual Auditorium
Aug 23: NO SEMINAR
Aug 30: NO SEMINAR
Sep 06 (11:15 am, Zoom): Dr. Chandra Venkataraman
Department of Chemical Engineering, Indian Institute of Technology Bombay
Guest of the Department of Chemical, Environmental, and Materials Engineering, College of Engineering,
and the Department of Atmospheric Sciences, Rosenstiel School, University of Miami
Rapid Climate and Clean Air Action in India:
Crucial Emission Sectors and Interventions
Presented at the Law School Auditorium, Room E352
Click to Request Link to Zoom Recording
There is wide agreement on the abatement of greenhouse gas emissions as key to tackling long-term climate change. Less known is the growing scientific evidence for action on short lived climate pollutants (SLCPs), which are responsible for about 45% of present-day global warming. Addressing SLCPs can deliver rapid climate and clean air benefits. This talk will present research evidence linking human activity to the emissions and impacts of carbonaceous aerosols, a key class of SLCPs, on climate and air quality in India. These findings are from the multi-institutional pan-India COALESCE network. Intrinsic aerosol chemical and optical properties will be discussed, vis-à-vis their net cooling or warming nature. A multi-scale emission inventory management system, the Speciated Multi-pollutant Generator (SMoG)-India-COALESCE, with new calculators underpinned by the largest yet primary datasets from pan-India field surveys and field emission factor measurements will be presented. Measurements and modeling evidence will be used to identify source-sectors significantly influencing levels of PM2.5 levels and its climate relevant carbonaceous constituents. The control of air pollution relies heavily on the phase-out of solid fuel combustion. A key question relates to prioritizing interventions, between formal or regulated economic sectors (coal for energy generation and industry) and informal sectors (biomass for residential energy and waste disposal). The talk will emphasize equity and sustainable development at the core of key interventions for delivering both climate and clean-air benefits in India.
Chandra Venkataraman is the Shobha Dixit Chair Professor of Chemical Engineering and Climate Studies at the Indian Institute of Technology Bombay. Her research focuses on the intersection of technology and emissions, aerosol science, climate science, and air pollution. Prof. Venkataraman has built academic and research programs to further climate and air quality studies in South Asia. She was the Founding Head of the Centre for Climate Studies, IIT Bombay, and National Coordinator for the multi-institutional COALESCE network of the Indian Ministry of Environment. Prof. Venkataraman has contributed to policy networks globally and is a member of the Science Advisory Panel of the Climate and Clean Air Coalition of the UNEP. She has authored over 150 widely cited publications, a book, and four patents. She is a fellow of the Indian National Academies of Science and Engineering and recipient of a Fulbright-Nehru Fellowship, the PRL Vikram Sarabhai Award, and teaching and research excellence awards at IIT Bombay. Prof. Venkataraman has strongly mentored students to academic and research positions worldwide. She has worked to address institutional systems to support the advancement of women in STEM.
Sep 13: NO SEMINAR
Sep 20: NO SEMINAR
Sep 27: Dr. Yair Cohen
NVIDIA Research, Santa Clara, CA
Guest of the Department of Atmospheric Sciences, Rosenstiel School
Recent Advances in Applying Machine Learning to Weather Modeling
at NVIDIA and Beyond
Recent years have witnessed a paradigm shift in weather and climate prediction, driven by rapid advancements in Machine Learning (ML). The application of cutting-edge ML architectures such as transformers, Fourier neural operators, and diffusion models to train on state-of-the-art, data-assimilating weather products has shown remarkable promise. As major meteorological institutions like NOAA and ECMWF incorporate ML-based forecasts into their daily operations, these methods are poised to revolutionize weather modeling in several ways. On the global scale, ML models trained on ERA5 reanalysis data (~25 km resolution) have demonstrated forecast skills rivaling those of operational numerical models, while offering an unprecedented four-orders-of-magnitude speedup. This breakthrough unlocks new possibilities for large ensemble forecasts, potentially enhancing our ability to quantify uncertainties and predict extreme events with greater accuracy. Recent research has explored coupling these weather models with parsimonious ocean models, aiming to extend forecast skill beyond the medium range and bridge the gap between weather and seasonal prediction. Concurrently, on regional scales, ML-based downscaling approaches have shown remarkable promise in super-resolving and forecasting kilometer-scale cloud properties, such as radar reflectivity. These downscaling techniques represent a crucial step towards the operational deployment of global ML forecasts. The inherent stochasticity of atmospheric processes at these fine scales necessitates the use of advanced generative ML methods, capable of capturing the probabilistic nature of local weather phenomena. Ultimately, the potential of better and more adaptive, ML-based data assimilation techniques highlights the potential of such models to surpass the skill of their numerical counterparts.
This talk will present a comprehensive overview of recent advancements in ML weather modeling, focusing on three key areas. First, I will present recent works on Ensemble Forecasting using global ML models, discussing their successes challenges and potential pitfalls, biases, and strategies for mitigation. Second, I will discuss two recent weather downscaling works that used generative ML models for regional forecasts. Third, I will discuss the possibility of using generative ML methods (diffusion models) for data assimilation. The presentation will highlight key findings from several groundbreaking studies conducted at NVIDIA and other leading institutions in the field. By synthesizing these recent developments, we aim to provide a comprehensive view of the current state of ML applications in weather and climate science, and offer insights into how these technologies are poised to transform our forecasting capabilities in the near future.
Some relevant publications can be found here:
https://arxiv.org/pdf/2401.15305
https://arxiv.org/abs/2408.03100
https://www.arxiv.org/abs/2408.01581
https://arxiv.org/abs/2309.15214
https://d1qx31qr3h6wln.cloudfront.net/publications/StormCast.pdf
https://arxiv.org/abs/2406.16947
Oct 04: NO SEMINAR
Oct 11: Dr. Claire Paris-Limouzy
Department of Ocean Sciences, Rosenstiel School
A Coupled ESM-Lagrangian Modeling Framework for Quantifying the Exposure
to Solar Radiation in Biotic and Abiotic Particles Moving in the Ocean
Recording Available at COMPASS ON DEMAND
Sunlight irradiance is a source of energy for life on Earth but it can also be a stressor at high doses of UV radiation. Solar radiations also affect chemical pollutants by enhancing the photooxidation of oil and increasing the fragmentation and photodegradation of plastics in the ocean. Irradiance varies spatially, temporally, and with water bio-optical characteristics and is expected to increase with the climate crisis. To better manage solar radiation and predict its effects with climate change, we need to understand how light-related processes affect the ocean ecosystems. The first step is quantifying radiation exposure on living and non-living particles. So far, current irradiance exposure estimates have neglected the advection by the ocean and the movement of particles due to their biological or physicochemical traits. Here we present a novel Irradiance Module developed to quantify the cumulative UV radiation exposure on active particles. The module coupled the open-source Connectivity Modeling System (CMS) to the net surface shortwave radiations from the NAVy Global Environment Model (NAVGEM). Case scenario simulations of zooplankton and microplastics in the global ocean show that the cumulative doses of irradiance received by particles are primarily determined by their 3D pathways rather than by the irradiance reaching the ocean’s surface. With the particles' motion behavior overlooked, the fine scale light-related processes affecting them remains invisible. Several biological, chemical, and physical applications may emerge from this solar irradiance modeling framework.
Claire Paris, professor in the Department of Ocean Sciences of the University of Miami's Rosenstiel School of Marine, Atmospheric, and Earth Science, is a biological oceanographer with interdisciplinary expertise in marine ecology and Lagrangian dynamics. She specializes in the dispersion and fate of larvae and marine pollutants and has developed innovative Lagrangian approaches designed to track inertial particles at sea. Paris has expertise in modeling deep-sea oil spills and served as Lead PI for oil-spill modeling research for the Gulf of Mexico Research Initiative (GOMRI) C-IMAGE Consortium. She served as President of the Early Life History section of the American Fishery Society (AFS), is the 2018 recipient of the prestigious American Geophysical Union (AGU) Rachel Carson Lecturer Award, and is appointed to the Ocean Research Advisory Panel. Paris is regarded as an authority on marine dispersal, as evidenced in her 36 keynotes and invited talks, 138 peer-reviewed publications, and 34 book chapters, and Editor of two Deepwater Horizon Legacy Springer books. Claire Paris has a B.Sc. in Biochemistry from the University of Bordeaux, France, an M.Sc. in Biology and Living Resources from the University of Miami, and a Ph.D. in Coastal Oceanography from the State University of New York (SUNY) at Stony Brook, USA.
Oct 18: Dr. Abhishek Savita
Department of Atmospheric Sciences, Rosenstiel School
Impact of Multidecadal Climate Modes' Variability on the European Climate
in the Recent Decades
Recording Available at COMPASS ON DEMAND
Pronounced negative trends in wintertime surface temperature have been observed over Eurasia during 1993-2013. However, it is unclear whether the negative trends were due to internal atmospheric variability or forced externally from either the ocean surface, solar forcing, or atmospheric composition. In this study, we use ensembles of atmosphere-model integrations for the period 1993-2013 to investigate the mechanisms of the wintertime cooling over Eurasia. We find that the cooling was mostly driven by intrinsic atmospheric variability, specifically the North Atlantic Oscillation (NAO), and to a lesser extent by multidecadal ocean variability such as the Interdecadal Pacific Variability (IPV) and Atlantic Multidecadal Variability (AMV). A dominant role for atmospheric variability as shown by our results implies limited predictability of decadal climate variability over Eurasia.
Oct 25: NO SEMINAR (Rosenstiel School Faculty Meeting)
Nov 01: SPECIAL ATM & OCE FACULTY PRESENTATION SERIES
Dr. Tamay Özgökmen
Department of Ocean Sciences, Rosenstiel School
The Motion of the Ocean at Submesoscales
Recording Available at COMPASS ON DEMAND
I will review the main scientific basis, approach, and results from 10-year project CARTHE, which was funded in the aftermath of the Deepwater Horizon oil spill in the Gulf of Mexico. In addition, recent research projects that span out of this project will also be introduced.
Nov 08: STUDENT SEMINARS
Victoria Pizzini (ATM)
Intercomparison of Hurricane Wind Fields Over the Ocean
From Satellite Sensors and Dropsondes
The NOPP Hurricane Coastal Impacts (NHCI) project, under the National Oceanographic Partnership Program (NOPP), brings together over 10 institutions from the U.S. and the Netherlands to improve predictions and assessments of damage from landfalling hurricanes, including advancements in satellite remote sensing. Our team at the Rosenstiel School and its Center for Southeastern Tropical Advanced Remote Sensing (CSTARS) focuses on rapidly mapping topographic changes and identifying flooded areas on land, and refining wind and wave field products for ocean monitoring. In collaboration with NOAA's 2024 Hurricane Field Program through the Advancing the Prediction of Hurricanes Experiment (APHEX), we compare high-resolution wind fields from multiple C-band synthetic aperture radar (SAR) satellites with data from the Advanced Scatterometer (ASCAT), the Cyclone Global Navigation Satellite System (CYGNSS), and in-situ data from dropsondes. This enables us to evaluate each instrument's strengths and limitations and identify discrepancies between wind data products and algorithms. Dropsonde data from NOAA P-3 flights timed with SAR satellite overpasses allow us to study SAR-derived hurricane wind fields in detail. ASCAT and CYGNSS data, however, average over larger areas and can be offset by ±2 hours relative to SAR overpasses, with further delays introduced by dropsonde deployment. To address these challenges, we discard distant data points and account for spatial and temporal proximity in our regression analysis, as well as for the motion of the hurricane eye. In one particularly interesting case, SAR detected a storm that went unnoticed by other sensors, highlighting its unique potential and the need for further investigation.
Tyler Tatro (ATM)
The Impact of Biomass Burning Aerosol on Climate Model Diurnal Cycle Cloudiness
in the Southeast Atlantic
Every June to October, the biomass burning season in southern Africa produces biomass burning aerosol (smoke), which is transported westward over a semipermanent stratocumulus cloud deck. Smoke particles absorb solar radiation, which modifies the radiative heating profile and local circulations, but the effects vary based on the altitude of the smoke and the low cloud cover. Smoke particles can also act as cloud condensation nuclei, which can brighten clouds, suppress precipitation, and increase cloud lifetimes. Therefore, the net impact of biomass burning aerosol on the regional radiative budget remains uncertain, but recent works find that smoke reduces cloudiness in August by preventing nighttime recoupling between the stratiform cloud layer and sub-cloud moisture. In this work, we use sub-hourly data available from the Cloud Feedback Model Intercomparison Project (CFMIP) to explore the diurnal cycle of August cloudiness in fixed sea surface temperature and sea-ice simulations (AMIP) of three climate models. Model responses to smoky conditions scale with mean biases in cloud-convection coupling and the single scattering albedo of the smoke. Higher-order simulations from the UK Met Office Unified Model and observations available from the Department of Energy Layered Atlantic Smoke Interactions with Clouds (LASIC) field campaign provide constraints on model physics.
James Christie (ATM)
Halogen Production From Saline Playa Dust Emitted From the Great Salt Lake:
Implications of the Shrinking Great Salt Lake on Regional Oxidant Budgets
Atomic chlorine (Cl•) is a powerful atmospheric oxidant which affects air quality and climate. One important source of Cl• is the photolysis of nitryl chloride (ClNO2), formed from the reaction of dinitrogen pentoxide (N2O5) and chloride-containing aerosol. However, sources of chloride-containing aerosol in inland regions, and their subsequent reaction kinetics, remain poorly constrained. To better understand inland sources of Cl•, we analyzed playa (i.e., dried saline lakebed) samples collected from dust emitting regions along the Northern and Southern areas of the shrinking Great Salt Lake to investigate their mineralogy, reactivity, and ClNO2 forming potential. The reactive uptake coefficients of N2O5 (γN2O5) for all samples ranged from 0.005 to 0.064, with the average γN2O5 of the Northern Area samples approximately double the average γN2O5 of the Southern Area samples. We attribute the increased γN2O5 of Northern playas to increased particulate chloride and silicate, while the reduced γN2O5 in Southern playas is due to particulate organics and high quantities of gypsum, a non-reactive mineral. The yield of ClNO2 is >50% for all playas tested, with one exception. Using our kinetic data during an ambient wintertime case study, we estimate playa dust contributes up to 5% of observed ClNO2, a lower estimate which likely increases during the spring, when dust emissions are higher. Our work highlights the importance of including playa dust as an oxidant source in current air quality models, especially as reductions of anthropogenic halogen sources are implemented in the United States, and ephemeral lakes continue to shrink globally.
Nov 15: STUDENT SEMINARS
Aidan Mahoney (ATM)
Tropical Cyclones and Upper Tropospheric Temperature Uncertainty
in Global Climate Models
Aidan D. Mahoney1, Brian J. Soden1, and Bosong Zhang2
1Rosenstiel School, 2Princeton University, Princeton, NJ
The potential intensity (PI) of tropical cyclones (TCs) is expected to increase in a warming climate. This provides more favorable thermodynamic environments for TCs and allows for the strongest TCs to become stronger in a warmed climate. HighResMIP is a high-resolution version of CMIP6 that can natively resolve TCs. The range in PI among the HighResMIP historical simulations is larger than coarser CMIP6 models. HighResMIP historical simulations use common surface boundary conditions based on observed SSTs. We therefore propose that upper tropospheric temperature differences are responsible for the large range in PI in the HighResMIP historical simulations. Globally and in all TC Basins, we find a strong relationship between the spread in PI and upper tropospheric temperatures, with higher PI corresponding to cooler upper tropospheric temperatures. Limitations of the resolved TCs in HighResMIP prevent us from concluding how the spread in PI may impact TC activity. To elucidate this, we run simulations using the Geophysical Fluid Dynamics Laboratory High Resolution Atmospheric Model (HiRAM) with the upper ten model levels nudged by ± 4 K and ± 8 K from the control simulation. TC activity is evaluated in the four nudged simulations and the control, and we find strong relationships consistent with prior literature between metrics of TC activity and upper tropospheric temperatures.
Caitlin Martinez (ATM)
When Is Atmospheric Noise Most Important for Limiting ENSO Predictability?
The El Niño-Southern Oscillation (ENSO) is the dominant mode of climate variability in the tropical Pacific and plays an essential role in modulating the global climate through its numerous teleconnections. As a coupled phenomenon, ENSO is inherently multivariate, and coupled feedbacks are central to generating ENSO events. Debate remains, however, over which initiation mechanisms dominate, and consequently, there is a lack of consensus on ENSO predictability estimates. A key challenge in predicting ENSO lies in its sensitivity to stochastic processes. Coupled feedbacks are critically important in the tropical Pacific but the role of atmospheric noise in shaping ENSO growth remains unclear. The seasonal cycle of ENSO is closely tied to the annual variations in the coupled tropical Pacific system, thus affecting ENSO development and the influence of atmospheric noise. Moreover, ENSO predictability and forecast skill are seasonally dependent, with each season presenting unique challenges. Previous work demonstrates that suppressing ENSO variability, thereby isolating coupled instabilities, offers a baseline for understanding the tropical system's natural variability in the absence of ENSO. Initializing ensemble forecasts seasonally offers insights into the dynamics of ENSO evolution and when atmospheric noise most critically limits predictability. This presentation synthesizes the results of two 60-member CESM2 ensemble forecasts, initialized from an ENSO-neutral reference climatology, to examine when atmospheric noise most critically limits ENSO predictability.
Eric Mischell (ATM)
Observed Changes in Convective Mass Flux and Precipitation Extremes
Under Global Warming
Satellite observations and general circulation models have demonstrated that the atmospheric overturning circulation weakens in response to global warming. This is rooted in thermodynamic and energetic constraints on near-surface specific humidity and precipitation. While the response of the globally-averaged convective mass flux, specific humidity, and precipitation is well established, the response of extremes is more uncertain. We show, using satellite observations from the SSM/I and SSMIS instruments, the trends in tropical precipitation and convective mass flux by intensity from 1987 to 2024. We find that heavy rainfall events have increased in frequency by ~8%/decade, while moderate rainfall events have decreased in frequency by ~1.7%/decade. Meanwhile we find that the distribution of convective mass flux has shifted from higher to lower intensities. This implies that the increase of extreme precipitation is driven by an increase of available moisture rather than a strengthening of convective updrafts.
Nov 22: Dr. John van Leer
Department of Ocean Sciences, Rosenstiel School (retired)
Oceanographic Founders, Remembered
– How Ocean Sciences Took Root at University of Miami
Recording Available at COMPASS ON DEMAND
Our earliest oceanographic laboratory in the U.S. was established on the West Coast – Scripps Institution of Oceanography, in 1903. But none were located on the East Coast, despite numerous distinguished academic institutions. So, a study group was formed by the U.S. Academy of Sciences to study whether an East Coast laboratory should be established.
I was privileged to discuss these events with Henry Briant Bigelow (1879-1967), a key individual and the founding director of Woods Hole Oceanographic Institution (1930), in the last year of his consequential life. I had similar discussions with Walton Smith (1909-1989) after his retirement, about origin stories surrounding the Marine Lab at the University of Miami (1943). How did these foundings happen, and why did they choose their respective locations?
John van Leer was a professor at the Rosenstiel School for 51 years, from 1971 to 2022.
Nov 29: NO SEMINAR (Thanksgiving Recess)
Dec 06: Brian McNoldy
Department of Atmospheric Sciences, Rosenstiel School
A History of Naming and Retiring Atlantic Tropical Cyclones
Recording Available at COMPASS ON DEMAND
The 2024 Atlantic hurricane season ends on November 30, and some storms' names are probably etched in your memory while others are easily forgotten. Accounts of destructive hurricanes go back centuries, but the practice of giving them human-sounding names only goes back about seven decades. I will take you through the winding path of why and how they are given the names we use today. Then we'll explore the much more nuanced ritual of retiring storm names. A name can be retired from being reused if the storm is determined to have been especially deadly or destructive, and statistics of those retired names reveal some expected and some unexpected patterns. Looking at a history of retired storm names tells us less about nature and more about human psychology and socio-economic vulnerability.
SPRING 2025 PREVIEW
Mar 21: Dr. Shafer Smith
Department of Mathematics, New York University
