Skip to content

GFDL Past Events & Seminars - 2019

Click a row to display that event's description/abstract.

Date Speaker Affiliation Title of Presentation
Feb. 6Lunchtime Seminar Series - Keith DIxonGFDLToward Evaluating & Bias Correcting CM4's Surface Climate for the Conterminous USA
This presentation's main thrust will be to compare the GFDL CM4 model's representation of several surface temperature characteristics over the 48 contiguous states to those of some common observation-based data products. The period of interest is 1981-2014. Several analyses make use of a subset of the ClimDEX indices (Climate Indices of Extremes; www.climdex.org) and adopt a regional perspective akin to that used in the Fourth National Climate Assessment (nca2018.globalchange.gov). Preliminary results show substantial regional and seasonal differences in diagnostics computed from CM4 surface temperatures, with some tendency towards cool biases and smaller than observed diurnal temperature ranges. Additionally, results from the application of one bias correction method illustrate the extent to which, for the historical period, such statistical methods can account for model biases and provide outputs at finer spatial resolution -- factors important to several types of climate impacts studies. Plans for upcoming analyses will be discussed, including potentially expanding analyses to CM4 and ESM4 historical and future projections, examining additional surface variables, applying different bias correction and statistical downscaling techniques, and making statistically refined output available for use in in-house impacts studies. As this is an ongoing project, we are open to ideas and potential internal collaborations regarding the expansion of these analyses of the modeled and observed CONUS surface climate.
Feb. 7Formal Seminar - Patricia QuinnNOAA-PMELThe Remote Marine Boundary Layer Cloud Condensation Nuclei Budget
Sea spray aerosol (SSA) consists of inorganic sea salt and organics that are scavenged from surface seawater during a wind-driven bubble bursting process that results in a flux of aerosol from the ocean to the atmosphere. Sea spray aerosols impact Earth's radiation balance by directly scattering solar radiation. They also act as cloud condensation nuclei, thereby altering cloud properties including reflectivity, lifetime and extent. The influence of sea spray aerosol on cloud properties is thought to be particularly strong over remote ocean regions devoid of continental particles. Yet the contribution of sea spray aerosol to the population of cloud condensation nuclei in the marine boundary layer remains poorly understood. This talk will assess what is known about the impact of surface ocean biology on SSA composition and CCN activity based on several research cruises in the sub-Arctic North Atlantic — home of the world's largest phytoplankton bloom. In addition, the relative contribution of SSA to the marine boundary layer CCN budget will be assessed based on measurements over the Pacific, Southern, Arctic, and Atlantic oceans made over the past 25 years.
Feb. 13Lunchtime Seminar Series - Hussein AluieUniversity of RochesterEnergy Pathways & Cascades in both Scale and Space: Ocean Applications
Large-scale currents and eddies pervade the ocean and play a prime role in the general circulation and climate. The coupling between scales ranging from $O(10^4)$ km down to $O(1)$ mm presents a major difficulty in understanding, modeling, and predicting oceanic circulation and mixing, where our constraints on the energy budget suffer from large uncertainties. To this end, we have developed a coarse-graining (or filtering) framework for analyzing the multi-scale dynamics on the sphere. This is made possible by ensuring that our filtering operators and spatial derivatives on the sphere commute, thereby allowing us to derive the PDEs governing any sets of scales. I will demonstrate the application of this framework to satellite altimetry data and to strongly eddying high-resolution simulations using General Circulation Models. Sponsor: Stephen Griffies
Feb. 21Formal Seminar - Richard SeagarLamont, Columbia UniversityRecent strengthening of the tropical Pacific zonal SST gradient is a dynamically consistent response to rising greenhouse gases
As exemplified by El Nino, the tropical Pacific Ocean strongly influences regional climates and their variability worldwide. It also regulates the rate of global temperature rise in response to rising greenhouse gases (GHGs). The tropical Pacific Ocean response to rising GHGs impacts all of the world's population. State-of-the-art climate models predict that positive radiative forcing reduces the west-to-east warm-to-cool sea surface temperature (SST) gradient across the equatorial Pacific. In nature, however, the gradient has strengthened in recent decades as GHG concentrations have risen sharply. This stark discrepancy between models and observations has troubled the climate research community for two decades. Here, by returning to the fundamental dynamics and thermodynamics of the tropical ocean-atmosphere system, and avoiding sources of model bias, we show that a parsimonious formulation of tropical Pacific dynamics yields a response that is consistent with observations and attributable to rising GHGs. We use the same dynamics to show that the erroneous warming in state-of-the-art models is a consequence of their cold bias in the equatorial cold tongue. The failure of state-of-the-art models to capture the correct response introduces critical error into their projections of climate change in the many regions sensitive to tropical Pacific SSTs.
Feb. 27Lunchtime Seminar Series - Liping ZhangCICS-PDecadal variability and predictability in the Southern Ocean - implications for interpreting recent observed trends
While decadal variability and predictability in the North Atlantic and North Pacific have received considerable attention, there has been less work on decadal variability and predictability in the Southern Ocean. As shown previously, a coherent mode of decadal to centennial variability exists in multiple climate models. The mechanism involves a multidecadal accumulation of heat in the subsurface of the Southern Ocean, which is then rapidly discharged through intense oceanic convection when the accumulation of subsurface heat reduces the stratification of the water column. The release of this accumulated subsurface heat can have considerable regional scale climatic impacts, along with substantial impacts on ocean heat uptake. Using a large suite of perfect predictability experiments, in concert with long control simulations, we show that this variability has a high degree of predictability. We present further results that show this type of variability may play an important role for interpreting recently observed trends of sea ice and temperature in the Southern Ocean. Specifically, observed trends over the last several decades resemble a particular phase of this variability in which reduced oceanic convection leads to subsurface warming and surface cooling, with associated increases in sea ice extent. This phase of natural variability may substantially contribute to observed decadal trends, working in concert with other factors.
Feb. 28Formal Seminar - Hugh MorrisonNCARAre we "deadlocked" on the problem of parameterizing microphysics?
Current microphysics schemes in cloud, weather, and climate models exhibit deficiencies that are due in part to their simplified representation of a complex natural state, and in part due to a fundamental lack of understanding of microphysical processes. Although there have been many advancements, there is little evidence that increasing sophistication of schemes results in convergence on the "truth". For example, model solution spread using different bin microphysics schemes (explicitly evolving the particle size distributions) is similar to that using different bulk schemes in recent model intercomparison studies. Two key challenges will be discussed: 1) numerical implementation, and 2) fundamental uncertainty of the underlying process rates and parameters. The recently-developed Lagrangian super-particle method helps to address the former, and I will advocate for its use in process-level microphysical modeling. In this approach, the cloud and precipitation particle populations are represented by a sub-sampling of point particles that follow Lagrangian trajectories in the modeled flow. However, like all approaches, the microphysical process rates and parameters using the super-particle method still have large uncertainties, especially for the ice phase. These uncertainties arise from the fact that there is no benchmark set of governing microphysical equations, combined with the inherent challenge of directly observing process rates. To address these inevitable uncertainties and constrain process rates rigorously with the wealth of indirect observations now available (e.g., satellite and ground-based remote sensing), the incorporation of Bayesian statistics directly into schemes is proposed. This idea will be discussed in the context of a novel probabilistic bulk parameterization framework called the Bayesian Observationally-constrained Statistical-physical Scheme (BOSS). BOSS combines existing, though inherently limited, process level microphysical knowledge with flexible process rate formulations and parameters trained to observations through Bayesian inference. Using a raindrop size distribution (DSD) normalization method that relates DSD moments to one another via generalized power series, generalized multivariate power expressions are derived for the microphysical process rates as functions of a set of prognostic DSD moments in BOSS. The approach is flexible and can utilize any number and combination of prognostic moments and any number of terms in the process rate formulations. This means that both uncertainty in parameter values and structural uncertainty associated with the process rate formulations can be investigated systematically, which is not possible using traditional schemes. I will discuss results using a Monte Carlo Markov Chain sampler within the BOSS framework to constrain microphysical process rates and parameters directly with synthetic observations.
Mar. 5Intern Informal Seminar - Dan GableGFDLArgo netCDF Improvements and New FRE
During my coop here at the GFDL, I have worked on two major projects, updating the Argo netCDF to ASCII converter from MATLAB to python, and helping design and implement a new workflow to replace FRE. The Argo converter takes netCDF files created by Argo floats and converts them into ASCII files using MATLAB. MATLAB was used because of the many matrix operations used in the converter. Some of the built-in MATLAB functions had to created in python. Python is a better option for cost and speed, because python is open source and much faster. First, the netCDF file is unpacked, and the potential temperature is calculated. Next, many quality control checks fix missing and incorrect data. Finally, the data is formatted and written to an ASCII file. The biggest improvement to the converter is speed. The MATLAB converter can process a file in 3 to 4 minutes, while the python converter takes around 30 seconds. The current workflow uses XML files to hold the information need to run different experiments. The newly proposed workflow uses SQL tables with a python wrapper. The main advantages to SQL tables is readability and inheritance. The information connected to a single experiment lends itself well to the SQL table format. Originally the XML worked great for experiments, because a single XML contained all the information needed to run an experiment. If a scientist wanted to replicate a coworker's experiment, the scientist could run the exact experiment with the same XML. Unfortunately, XML files are very long and no longer hold all of the information of one experiment, so replicating an experiment can be more complicated. Inheritance allows for one, small SQL table to hold all the relevant information of an experiment. Information specific to an experiment will be held in the scientist lowest level SQL which will inherit from higher level tables. POC: Jessica Liptak
Mar. 6Lunchtime Seminar Series - Alex ZhangCICS-PTropospheric ozone inter-annual variability and extreme events: Multi-model assessments and observational constraints
Better knowledge on how well current global climate-chemistry models represent the inter-annual variabilities (IAVs) and extremes of tropospheric ozone (O3) as well as their driving factors is with great importance for understanding and predicting long-term O3 trends with models. Here we explore the IAVs of tropospheric O3 in the northern hemisphere simulated with multiple chemistry-climate models participating in the Chemistry-Climate Model Initiative phase 1 (CCMI-1) and assess the models' performances of simulating the O3 extremes during extreme weather conditions in the U.S. We find that the examined models have diverse performances on simulating the tropospheric O3 IAVs with or without observational constraints and all models underestimate O3 anomalies during heat waves over the past three decades. Using intensive field measurements and the latest AM4 and GEOS-Chem at high resolutions, we specifically investigate the IAVs and sources of springtime O3 in the western U.S. and find that background O3 is the key driver of O3 IAV there with STT O3 dominating the background O3. Our findings have advanced the knowledge of the abilities and deficiencies of current climate-chemistry models in representing the IAVs and extremes of tropospheric O3.
Mar. 7Formal Seminar - Paul NewmanNASA/GSFCThe Major Stratospheric Sudden Warming of February 2018
A major stratospheric sudden warming (SSW) occurred in February 2018. This was a wave-2 SSW that split the stratospheric polar vortex. The magnitude of the eddy forcing of the event was possibly the largest observed in the satellite date era (1980-present). This SSW warmed the Arctic, reversed the normal westerly winds to easterlies, and dramatically increased total column ozone levels in the Arctic. In addition to the polar effects, this SSW had a major impact on trace gas distributions in the mid-latitudes and tropics. In this presentation we characterize the 2017-18 winter state, show the SSW evolution, the context of this SSW against the SSW climate record, and show its dynamical impact across the northern hemisphere and tropics. In particular, we also analyze the SSW's impact on trace gas distributions.
Mar. 12Visiting Scientist Informal Seminar - Matthew Wazniak Airborne pollen: the potential for a coupled vegetation-atmosphere system via aerosolcloud-precipitation interaction
We know today that the atmosphere and the land surface are coupled in a numerous ways, exchanging mass, momentum and energy in a systematic fashion, an example being precipitationevaporation coupling. Terrestrial vegetation adds nuance to these interactions by mediating them (e.g. transpiration mediating the hydrologic budget) and also by causing a number of unique coupled interactions (e.g. CO2 fertilization feedbacks on photosynthetic carbon uptake). Here, I show the potential for a unique vegetation-atmosphere coupled system involving airborne pollen. Pollen grains are adapted in particular vegetation to be emitted by wind and transmitted in the atmosphere as an aerosol. Temperature, wind, humidity, rain - the specific meteorological environment of the vegetation - determine when and how much pollen is emitted to the atmosphere. In particularly humid or rainy conditions, pollen grains rupture from the osmotic pressure exerted by surface wetting, thereby releasing subpollen particles (SPPs) to the atmosphere. SPPs are small (50-1,000 nm), hygroscopic and have the potential to activate as nuclei for cloud droplet growth, or cloud condensation nuclei (CCN), thereby altering cloud and precipitation behavior via aerosol-cloud-precipitation interactions. I present the results of a regional climate model experiment in which simulations of meteorologically-forced pollen emissions, pollen rupture and the release of SPPs, and the activation of SPPs as CCN are used to demonstrate the potential for SPPs to affect rain through a warm cloud aerosol-cloud-precipitation effect. Simulations over the continental United States were run during the spring pollen season for 3 cases: 1) a control run with a uniform CCN background of 100 CCN cm-3, representing clean continental conditions; 2) an experimental run with SPP active as CCN using a literature-based estimate of 10-3 SPPs released per pollen grain; and 3) same as (2) but with an upperbound estimate of 106 SPP grain-1. Results show negligible changes to precipitation over land using 10-3 SPPs grain-1. However, using the upper-limit 106 SPPs grain-1, accumulated precipitation over land is reduced by a domain-average of 32% during the spring pollen season, and daily rain rate probabilities are shifted from 1-25 mm day-1 to
Mar. 13Hyung-Gyu LimDivision of Environmental Science and Engineering, POSTECH, Pohang, KoreaClimate feedback of bio-geochemical processes in the Arctic
The Arctic warming affects the decreasing marine phytoplankton mass in future climate. The Arctic warming is also affected by changing the marine phytoplankton via absorbing more shortwave radiation and in turn radiative redistribution in the upper ocean layer, so-called bio-geophysical feedback. This modulating the shortwave absorption rate by marine phytoplankton leads to conceive importance investigating bio-climate interaction. However, previous works were majorly focusing on annual mean chlorophyll change and its linear impact of bio-geophysical feedback and assuming the closed system of marine biogeochemical cycle. In this study, the evolutions, responses of Arctic phytoplankton activity in sub-seasonal to seasonal timescales, and both linear and nonlinear bio-geophysical feedback processes in the present-day and future climates were investigated by model simulations using a Geophysical Fluid Dynamics Laboratory (GFDL) CM2.1 earth system model (ESM). Results of historical run and Representative Concentration Pathway 4.5 (RCP4.5) scenario in the fifth's Coupled Model Intercomparison Project (CMIP5) models are discussed to support the results of single model experiments. Part I (Lim et al. 2018) suggested that two nonlinear rectifications of chlorophyll variability play cooling role in present day climate. Part II (Lim et al. 2019) suggest that the decreased interannual chlorophyll variability may amplify Arctic surface warming (+10% in both regions) and sea ice melting (-13% and -10%) in Kara-Barents Seas and East Siberian-Chukchi Seas in boreal winter, respectively. Projections of earth system models show a future decrease in chlorophyll both mean concentration and interannual variability via sea ice melting and intensified surface-water stratification in summer. We found that suggested two nonlinear processes in Part I will be reduced by about 31% and 20% in the future, respectively, because the sea ice and chlorophyll variabilities, which control the amplitudes of nonlinear rectifications, are projected to decrease in the future climate. The Arctic warming is consequently enhanced by the weakening of the cooling effects of the nonlinear rectifications. Thus, this additional biological warming will contribute to future Arctic warming. This study suggests that effects of the mean chlorophyll and its variability should be considered to the sensitivity of Arctic warming via biogeophysical feedback processes in future projections using earth system models.
Mar. 14Formal Seminar - Hai LinEnvironment CanadaMJO Teleconnections and Subseasonal Prediction
The Madden-Julian Oscillation (MJO) is the dominant mode of subseasonal variability in the tropics, which has a significant impact on the global atmospheric circulation through teleconnections, and thus represents an important source of predictability on the subseasonal to seasonal (S2S) time scale. To understand the dynamical processes of the MJO teleconnection, a primitive equation global atmospheric circulation model is used to study the extratropical atmospheric response to tropical diabatic thermal forcing of the MJO. Several aspects are examined, including sensitivity to the forcing location, nonlinearity of the response, and sensitivity to the initial state of the subtropical westerly jet. Most of the operational S2S models are able to capture the main features of the MJO teleconnection. The contribution of the MJO to S2S prediction of surface air temperature in North America and the North Polar region is also discussed.
Mar. 18Visiting Scientist Lecture - Meng ZhaoUniversity of California - IrvineUsing satellite remote sensing data for eco-hydrological process understanding and evaluating Earth system models
Vegetation connects terrestrial water, carbon flux, and energy cycles. Earth system models (ESMs) that simulate terrestrial ecosystems and biogeochemical cycles, are important tools to predict the potential outcome of climate change impact on vegetation. Such predictions are essential to inform sound policies that mitigate climate change impacts on ecological functions and services. However, current ESMs have large uncertainties in simulating land eco-hydrological processes. This is in part due to the limited observations for process understanding as well as for model diagnosis and assessment. Satellite observations are useful for these purposes because they have the advantage of global coverage and high revisit cycle. Here we use terrestrial water storage estimates from the Gravity Recovery and Climate Experiment (GRACE) in combination with a multitude of state-of-the-art satellite remote sensing observations to improve our understanding of land hydrology, ecosystems, and their interactions. We then use these observations to evaluate the counterparts simulated in CMIP5 ESMs including ESM2M from GFDL. We find large inter-model discrepancies and all models fail to reproduce the observed plant-water relations. Most of the CMIP5 models underestimate the impact of available land water supply on ecosystem productivity under dry atmospheric moisture conditions. Given the large uncertainty in predicting land water supply patterns, it is challenging to accurately project the impact of future warming-exacerbated atmospheric dryness on terrestrial ecosystems.
Mar. 19Visiting Scientist Lecture - Dillon AmayaScripps Institution of Oceanography, University of California - San DiegoThe North Pacific pacemaker effect on historical ENSO and its mechanisms
Recent studies have indicated that North Pacific sea surface temperature (SST) variability can significantly modulate the evolution of the El Niño-Southern Oscillation (ENSO), but there has been little effort to put these extratropical-tropical interactions into the context of historical ENSO events. To quantify the role of the North Pacific in pacing the timing and magnitude of observed ENSO events, we use a fully-coupled climate model to produce the first ensemble of North Pacific Ocean-Global Atmosphere (nPOGA) SST pacemaker simulations. In nPOGA, SST anomalies in the North Pacific (>15ËšN) are restored back to observations, but are free to evolve throughout the rest of the globe. We find that North Pacific SST variability has significantly influenced the observed trajectory of historical ENSO, accounting for approximately 15% of the total variance in boreal fall and winter. The interaction between the North and tropical Pacific is the result of two physical pathways: 1. A Wind-Evaporation-SST propagating mechanism in boreal spring, and 2. A convective response associated with the Seasonal Footprinting Mechanism in boreal fall. The latter accounts for 25% of the observed zonal wind variability around the equatorial dateline. On an event-by-event basis, nPOGA most closely reproduces the 2014-2015 El Niño and the 2015-2016 El Niño. In particular, we show that the 2015 Pacific Meridional Mode event increased wind forcing along the equator by 20%, potentially contributing to the extreme nature of the 2015-2016 El Niño. Our results illustrate the significant role of extratropical noise in pacing the initiation and magnitude of ENSO events and may improve the predictability of ENSO on seasonal timescales. Host is Tom Delworth
Mar. 20Visiting Scientist Seminar - Dr. Hemant KhatriJet drift over topography and jet-topography interactions
Alternating jet patterns have been observed in altimetry and float observations. The jets play an important role in the heat and tracer transport in the oceans. Oceanic jets are seen to possess spatio-temporal variability in the presence of topography. For example, over a zonally sloped topography, the jets tend to tilt from the zonal direction and drift in the meridional direction. In this talk, I will discuss why the jets tilt and drift over topography, and what implications these jets can have on the large-scale circulation. In a baroclinic quasi-geostrophic model, the jets drift to compensate for the potential vorticity (PV) advection across PV isolines by the mean flow. In addition, the drifting jets are directly forced by the imposed vertical shear via coupling through topography. On the other hand, eddies oppose the jets, which is opposite to the case of eddy-driven zonal jets. Also, eddy fluxes are significantly enhanced over topography. Furthermore, in addition to the tilted jets, other large-scale spatial patterns are observed, which efficiently interact and exchange energy. This indicates that oceanic zonal flow patterns can appear due to interactions among various large-scale modes and eddies can transfer energy to meridional scales larger than the jet-width scale set by Rossby waves and eddy energy. POC: Steve Griffies
Mar. 20Lunchtime Seminar Series - Yujin ZengPrinceton CICS-PImpact of Asian Irrigation on Water Availability in the Sahel
Asian irrigation accounts for about three-quarters of global irrigation water withdrawal. This irrigation plays an important role in reshaping terrestrial water resources. From an Earth-system perspective, irrigation can also be considered an important anthropogenic forcing on climate. However, previous research focused mostly on local or regional impacts of irrigation on hydrological or climatic systems, while the remote effects that may be caused through atmospheric connections are ignored. Here we estimate that large-scale irrigation in Asia could significantly increase the precipitation and runoff in the Sahel region of Africa by 8% and 12%, respectively. Summertime irrigation in Asia causes a large-scale cooling and induces air sinking. The sinking in Asia then causes the air over Europe and northern Africa to move east, and thus a low pressure anomaly forms over Europe and northern Africa. This creates a cyclonic anomaly that transports vapor from the Atlantic Ocean to the Sahel in Africa. As a consequence, water-vapor flux converges in the Sahel, and precipitation and runoff then increase. Moreover, the irrigation-induced water-vapor convergence releases latent heat and expands the ITCZ northward with a magnitude that rivals the impacts of CO2 emissions. Our results demonstrate that irrigation deserves more attention when studying water resources and climate change in some key regions.
Mar. 21Formal Seminar - Annica EkmanStockholm UniversitySulfate aerosol emission changes and Arctic amplification: what roles do the ocean and atmosphere play?
Anthropogenic emissions of aerosols have both an immediate effect on local air quality as well as regional and global effects on climate in terms of e.g. changes in the temperature and precipitation distribution. Recent studies have shown that regional reductions in sulfur emissions induce a substantial warming of the Arctic. However, the mechanisms that drive this response are still unclear. It is also not clear if emission reductions in different regions or by different aerosol components induce a different pattern of temperature response. We have used the Norwegian Earth system model (NorESM) to examine the respective contribution of changes in ocean heat convergence and contributions driven by changes through the atmosphere to the Arctic warming. We have also examined whether the temperature response is dependent on emission region, aerosol component and different levels of aerosol pollution. To distinguish the role played by the atmosphere and the ocean, we coupled NorESM to a mixed-layer ocean model and performed a set of idealized simulations where we modified the atmosphere and ocean simultaneously as well as separately. The simulations show that atmospheric heat transport drive the Arctic warming while the ocean tends to dampen the temperature response. Fully coupled ocean-atmosphere simulations show that emission changes in different regions, and by different aerosol components, generate a very similar pattern of temperature response. However, the response is non-linear; the response becomes stronger as the level of pollution decreases.
Mar. 25Visiting Scientist Lecture - Xinsheng QInUniversity of Washington (Seattle)Efficient Multi-scale Tsunami Modeling
Tsunami modeling generally requires resolving waves at a wide range of spatial scales, including (from large to small scale) transoceanic wave propagation, wave run-up near the coast line, inland inundation, and interaction between fluid and individual coastal structures. The numerical modeling of tsunami inundation that incorporates the built environment of coastal communities is challenging for both 2-D depth-integrated models and 3-D Navier-Stokes models. For 2-D models, inundation and flooding in this region can be very complex with variation in the vertical direction caused by wave breaking on shore and interactions with the built environment, and the model may not be able to produce enough detail. For 3-D models, a very fine mesh is required to properly capture the physics, dramatically increasing the computational cost and rendering impractical the modeling of some problems. In this talk, I will first give comparisons between a depth-integrated 2-D model and a 3-D model for tsunami inundation modeling. I will then present an efficient CUDA implementation of a 2-D tsunami model. Numerical experiments on realistic tsunamis show the validity and efficiency of the code. The GPU implementation, when running on a single GPU, is 3 to 6 times faster than the original model running in parallel on a 16-core CPU. With the code, the Japan 2011 Tohoku tsunami can be fully simulated for 13 hours in 3.5 minutes wall-clock time, using a single NVIDIA P100 GPU. This is done with AMR that adequately resolves the waves propagating across the ocean, and with 60-meter resolution near Crescent City, CA, producing good agreement of simulation results with observed tide gauge records in this harbor.
Mar. 27Lunchtime Seminar Series - Olivier PauluisThe global atmospheric overturning as seen in the ERA-5 data.
TBD
Mar. 28Formal Seminar - Christina PatricolaLawrence Berkeley National LaboratoryNatural and Anthropogenic Influences on Tropical Cyclones
Tropical cyclones (TCs) are among the costliest and deadliest natural hazards. There are numerous influences on TCs that can originate from the atmosphere and ocean, act constructively or compensate, operate on timescales spanning weather to climate, and arise from natural variability or anthropogenic forcings. Due to this complex set of factors, as well as the limited period of consistent observations, I use ensembles of convection-permitting regional climate model experiments to uncover causal relationships. Starting with anthropogenic change, I will (1) discuss how the intensity and rainfall of 15 recent destructive TCs could be different if similar events were to occur in pre-industrial and future climates. On the topic of TC variability, I will (2) demonstrate that Atlantic TCs are not limited by their typical precursor (African easterly waves) on the climate timescale, and (3) discuss the influence of the diverse spatial patterns of El Niño's sea-surface temperature (SST) warming on TCs. Finally, I will (4) present a new index for the El Niño-Southern Oscillation (ENSO) that, for the first time, uniquely captures ENSO's diversity and extremes and accounts for both the nonlinear response of deep convection to SST and background SST changes associated with the seasonal cycle and climate change. Altogether this research sheds light on the ongoing debate as to whether climate change has yet affected TCs and identifies the utility of potential atmospheric and oceanic sources of seasonal-centennial TC predictability.
Apr. 4Formal Seminar - Nils WediECMWFAun Aprendo - Earth System modelling at the European Centre for Medium-Range Weather Forecasts
The gradual progress in global numerical weather prediction includes a systematic approach to assess and quantify the associated forecast uncertainty by means of high-resolution ensembles of assimilation and forecasts. This involves simulations with billions of gridpoints, the continuous assimilation of billions of observations, rigorous verification, validation and uncertainty quantification, and it involves increasing model complexity through completing the descriptions of the global water and carbon cycles. The coupling of atmosphere, land-surface, ocean, sea-ice, and waves in ECMWF's Earth-system model requires a careful consideration of the highly varying temporal and spatial scales of the individual processes at the interfaces and their consistent initialisation. All this is necessary to increase the fidelity of daily forecasts and of European Copernicus Services, e.g. through the provision of state-of-the-art atmospheric monitoring services, warning systems for flood and fires, and providing reanalyses. A particular challenge arises from ensuring energy efficiency for these extreme-scale applications. This talk will comprehensively describe the current state-of-the-art in model development at ECMWF, discuss challenges, and sketch the next steps towards accelerating and improving storm-scale Earth-System modelling and assimilation in the coming decade. Sponsor: Lucas Harris
Apr. 10Lunchtime Seminar Series - Xianan JiangAOS - Visiting ScientistKey model processes for realistic simulations of the Madden-Julian oscillation
Despite tremendous influences on global weather extremes and a critical role in the seasonal to subseasonal (S2S) climate prediction of the Madden-Julian oscillation (MJO), representation of the MJO remains a grand challenge for present-day climate models. In this talk, I will start with an introduction of our recent analyses based on observations and multi-model simulations using the moisture mode framework, which suggest that realistic simulations of large-scale seasonal mean lower-tropospheric moisture pattern is crucial for faithful representation of MJO eastward propagation in climate models. Additional evidences on the importance of low-level mean moisture for MJO propagation will be further presented by examining the observed seasonal and interannual variations in MJO propagation characteristics, as well as based on idealized GCM simulations. These results provide important guidance for improvement of MJO representation in climate models.
Apr. 11Formal Seminar - Isaac GinisUniversity of Rhode Island - OceanographyAdvancing modeling capabilities to improve prediction of extreme weather events in the Northeastern United States
The Northeast U.S. coast experiences infrequent, high-impact landfalling hurricanes and nor'easters with complex storm characteristics. We are developing a modeling system to predict the consequences of coastal and inland hazards associated with these extreme storms in order to better prepare coastal communities for future risks. Our modeling approach adds new capabilities to the real-time ADCIRC-Surge Guidance System (ASGS), such as a highly refined computational mesh in order to properly resolve the complicated coastal geometry of the New England coast including narrow inlets and salt ponds, improved surface wind modeling near the coast and over land and coupling of storm surge and wave models. The Precipitation-Runoff Modeling System is applied to simulate rainfall runoff for all major rivers. The New England states are especially vulnerable to inland flooding since the rivers are relatively short and high-river discharge can coincide with coastal storm surge during extreme wind and rain events. The prediction system includes 3D visualization of hazard impacts on critical infrastructure such as buildings, bridges and wastewater treatment plants. Geographic points representing specific vulnerabilities are indexed directly into the computational grids of the hazard models to provide actionable outputs that are relevant to the users in real time. Host: Morris Bender
Apr. 16Special Lunchtime Seminar - Alistair HobdayCSIRODreaming of Ocean Predictability for extreme events, seasonal, and decadal scales
Alistair is a Research Director at CSIRO Oceans and Atmosphere and an Adjunct Professor at the University of Tasmania. His research focuses on investigating the impacts of climate change on marine biodiversity and fishery resources, and developing, prioritising and testing adaptation options to underpin sustainable use and conservation into the future. His projects involve multi-disciplinary teams that seek to support management and policy uptake of research, via co-production with stakeholders. In addition to more than 250 publications, he has contributed to the IPCC 4th and 5th assessment Australasia chapters, covering fisheries, oceanic and coastal systems, and is an editor for Fisheries Oceanography, Marine Ecology Progress Series, and Global Change Biology. In this interactive presentation, he will discuss the research progress and motivation for predictability of extreme events, such as marine heatwaves, and in seasonal and decadal forecasting.
Apr. 18Formal Seminar - Adina PaytanInstitute of Marine Sciences, UC Santa CruzAerosol impacts on marine biogeochemistry
Atmospheric deposition of trace elements and nutrients to the ocean can significantly modify seawater chemistry and influence oceanic productivity. However, mounting evidence suggests that the response of phytoplankton to atmospheric deposition depends on the chemical composition of the aerosols and varies across different phytoplankton species. Responses are also different depending on oceanographic setting and season. To determine if and how nutrients and metals from atmospheric deposition influence phytoplankton community structure in the Ocean we analyzed nutrient (nitrogen and phosphorous) and metal (Fe, Cu, Zn, Ni) concentrations in marine aerosols and tested how these constituents impact phytoplankton. This is done using incubation experiments with natural phytoplankton assemblages with different sources and amounts of aerosol or pure nutrients and metal additions. Laboratory-based culture experiments with phytoplankton from different taxonomic groups helped identify species that were most sensitive to aerosol additions. Variance in utilization of nutrients and susceptibility to metal toxicity was identified among different taxa, suggesting that aerosol deposition could potentially alter patterns of marine primary production and phytoplankton community structure. In addition, input of bioaerosols can also affect phytoplankton communities and should be considered. Changes in atmospheric deposition and aerosol composition that are impacted from natural and anthropogenic change could therefore have effects on ocean chemistry and productivity with potential feedbacks to the carbon cycle. Host: Lori Sentman
Apr. 24J. R. ToggweilerGFDL'Southern Ocean Winds' Comes to the Tropics
TBD
May. 1Lunchtime Seminar Series - Gaurav GovardhanGFDLSimulations of aerosol over the Indian region: Evaluation-Improvements and Climatic Implications
Aerosols affect climate by scattering and absorbing atmospheric radiation. In this study, upon evaluation of aerosol simulating regional and global chemistry transport models (WRF-Chem, SPRINTARS and HadGEM) over the Indian region, we find that the models underestimate mass concentration of Black Carbon (BC) aerosol by a large margin and the corresponding BC emissions appear to be the primary cause for it. The Aerosol Optical Depths (AOD) also appear to be unrealistic. The possible role of state of mixing of BC on such AOD simulations has been examined. It is found with the help of an offline model constrained by observations of aerosol particles (using a scanning electron microscope) that, the effects of realistic state of mixing of BC on AOD and related forcing are lesser than previously thought. Lastly, we examine the vertical profile of BC from model simulations and find that the model captures the observed sharp elevated layers of BC only after the prescription of BC emissions from aircraft. The analysis of model simulations and space-based LIDAR data shows that such BC layers can get vertically transported further up upon their interaction with the underlying strong monsoonal convection. Such lifted BC layers can potentially have severe climatic implications. Host: Dr. Ram
May. 2Informal Seminar -Shaoqing Ocean University of ChinaHow CESM have been ported onto newer architectures
To discuss how CESM have been ported onto newer architectures.
May. 2Formal Seminar - Ethan GutmannResearch Applications Lab, NCARMaking Climate Model Outputs Useful for the Water Resources Community
The water resource sector is one of those most likely to experience significant effects from climate change; however, water resource managers are unable to directly use the output from climate models to inform their planning. Such models do not adequately represent important hydroclimate phenomena; most notably, the effect of regional mountain ranges on precipitation and temperature is not resolved. As a result, water managers have turned to simplistic statistical downscaling methods, though these methods may not represent the physical processes any better than the global model. Even regional climate models (RCMs) have too great a computational cost at the spatial resolutions desired. Here we present a new modeling approach, the Intermediate Complexity Atmospheric Research model (ICAR) developed to represent processes important for water resources, particularly orographic effects. ICAR uses an analytical solution for flow over topography to adjust the large scale wind field; this is used to advect heat and moisture through the domain while calculating the same physical processes that a traditional RCM would (e.g. cloud microphysics and land surface feedbacks.) Because of its simplifying assumptions, ICAR is capable of running 100-1000 times faster than a traditional RCM, while still reproducing key facets of the regional climate. In addition, we can use ICAR to explore some of the, often ignored, uncertainties within regional and global climate modeling including the strength of the snow-albedo feedback effect and the variability in cloud microphysical parameters. Host: Keith Dixon.
May. 8GFDL Poster ExpoGFDL Poster Expo
GFDL Poster Expo For the program and list of posters, please visit: https://www.gfdl.noaa.gov/poster-expo-program-list/
May. 15Lunchtime Seminar Series - V. BalajiNOAA/GFDL and Princeton UniversityMetamodels and supermodels: Ideas and challenges from machine leaning in Earth System Science.
In this talk, we examine approaches to Earth system modeling in the post-Dennard era, inspired by the industry trend toward machine learning (ML). ML presents a number of promising pathways, but there remain challenges specific to introducing ML into multi-phase multi-physics modeling. A particular aspect of such 'multi-scale multi-physics' models that is under-appreciated is that they are built using a combination of local process-level and global system-level observational constraints, for which the calibration process itself remains a substantial computational challenge. These include, among others: the non-stationary and chaotic nature of climate time series; the presence of climate subsystems where the underlying physical laws are not completely known; and the imperfect calibration process alluded to above. The talk will present ideas and challenges and the future of Earth system models as we prepare for a post-Dennard future, where learning methods are poised to play an increasingly important role.
May. 16Formal Seminar - David RandallColorado State UniversityInteractions of convection and radiation under homogeneous boundary conditions on the sphere, with and without rotation
We study the interactions among convection, radiation, and large-scale circulations in idealized frameworks. Results will be shown from simulations of both rotating and non-rotating radiative-convective equilibrium with several versions of the Community Atmosphere Model. In all cases the sea surface temperature and the incident solar radiation are globally uniform. Both rotating and non-rotating cases are considered. The model versions differ in both their dynamical cores and their physical parameterizations. The dynamical cores tested are FV, SE, and MPAS. The parameterizations are CAM6 physics and super-parameterization. The latter is tested only with the MPAS dynamical core. The results are broadly consistent with earlier studies, but nevertheless differ strongly across the model versions. Analysis focuses on the organization of convection and the associated circulation systems, with an emphasis on the role of cloud-radiative effects. We also analyze how measures of the strength of the circulation vary with surface temperature. Host: Shiv Priyam
May. 20Informal Seminar Series - Maike SonnewaldEarth, Atmospheric and Planetary Sciences, MITOcean exploration with machine learning: An Antidote to Chaos?
Machine learning has the potential to widely influence oceanography, if applied with care. Three case studies highlight the potential for greatly accelerating the efficiency of ocean exploration using supervised (neural networks) and unsupervised (clustering) machine learning. First, two decades of data from the realistic ECCO state estimate 3D physical fields are used to objectively determine global physical regimes using k-means clustering. The identified regions correspond closely to those predicted by canonical theory from physical oceanography and the method can be scaled to analyze vast amounts of data from e.g. CMIP. Second, the high-dimensional dataset from the biogoechemical DARWIN model reveals the existence of ecological niches using t-SNE and DBSCAN clustering. Constraining ocean biomes, individual niches can be collated into larger socioeconomically relevant regions, and examined to understand how sensitive they are to climate forcing which is crucial to protecting the base of the ocean food chain. Finally, a multilayer perceptron (MLP) is trained to predict which global physical regime is present on the basis of local sea surface height. Using the results from the k-means clustering as labels, we achieve a recognition rate of >80%, with good performance across the physical regions. These case studies demonstrate that algorithms can be developed to explore the ocean that have vast potential for understanding complex problems. Host: V Balaji
May. 21Informal Seminar Series - Melanie Bieli Columbia University Extratropical Transition of Tropical Cyclones: From Climatology to Prediction
Tropical cyclones (TCs) undergoing extra tropical transition (ET) pose a serious threat to coastal regions in the mid-latitudes. In this talk, I will present the first climatology of ET that encompasses all major global TC basins and is based on a consistent set of data, time period, and method. Using best-track data from 1979-2017 to define the tracks of the storm centers, we identify storms that undergo ET by means of their paths in the cyclone phase space (CPS), calculated from geopotential height fields in reanalysis datasets. The results are used to study the seasonal and geographical distributions of storms undergoing ET, inter-basin differences in the statistics of ET occurrence, and the differences between the ETs defined by the CPS and those defined by the 'extratropical' labels (determined subjectively by human forecasters using a wider range of data) in the best-track archives. In the second part, I will introduce a logistic regression model that predicts ET in the North Atlantic and the Western North Pacific. The model uses elastic net regularization to select predictors and estimate coefficients. Predictors are chosen from the 1979-2017 best-track and reanalysis datasets, and verification is done against the tropical/extra-tropical labels in the best-track data. In an independent test set, the model skillfully predicts ET at lead times up to two days, with latitude and sea surface temperature as its most important predictors. The model can be integrated into statistical TC risk models and may provide baseline guidance for operational forecasts.
May. 22Lunchtime Seminar Series - Linjiong ZhouPrinceton University-GFDLGlobal to Convective-scale Prediction in the FV3-based Global Atmosphere Prediction System
GFDL has developed the FV3-based global atmosphere prediction system that serves as the prototype of the Next Generational Global Prediction System (NGGPS). This talk will cover the origin of the NGGPS project, the dynamical core selection process, and the model development progress at GFDL. The model being continually developed at GFDL is named SHiELDS (System for High-resolution modeling for Earth-to-Local Domains). SHiELDS is being used for multiple applications ranged from seasonal scale to hour scale, from global scale to convective scale. This talk will demonstrate the performance of SHiELDS in global to convective-scale prediction through 500 hPa height statistics, zonal mean temperature, cloud hydrometeors/fraction, and energy balance. The uniform-resolution SHiELDS significantly outperforms the operational Global Forecast System (GFS) in predicting 500 hPa height. Some biases to observation are noticed and being addressed. The variable-resolution SHiELDS improves the prediction of convective-scale features over the Contiguous United States while maintaining skillful global forecasts.
May. 23Formal Seminar - Mark ZelinkaLLNLCloud Feedbacks: What we knew then, what we've learned since, and what keeps us up at night
Predictions of global warming in response to a doubling of atmospheric carbon dioxide concentration vary widely across global climate models. Most of this range is attributable to model diversity in the strength of cloud feedbacks - the change in the planet's energy budget due to clouds per degree of global warming. Models whose clouds become less effective at cooling as the planet warms tend to warm more in the future, and vice versa. In this talk, I will provide an overview of the dominant cloud feedbacks in the climate system. For each feedback, I will discuss the state of knowledge at the time of the 5th IPCC Assessment Report roughly 6 years ago. I will then discuss key areas of progress made since that time. Finally, I will highlight areas of active research at the frontiers of knowledge in this challenging field. Host: David P. or Leo D.
May. 24Informal Seminar - Yan YuDepartment of Geography, University of California, Los AngelesInteractions between vegetation, dust, and climate in North Africa assessed from observation and modeling
North Africa is characterized by pronounced ecological and moisture gradients. The land surface varies remarkably from the tropical Congo rain forest, to steppe vegetation across the Horn of Africa, to savanna and croplands in the Sahel, to the Sahara Desert, therefore facilitating comprehensive investigations of vegetation-dust-climate interactions. The current understanding of land surface feedbacks has largely come from running and analyzing coupled vegetation-climate model simulations that have several key limitations. Observational studies are critically needed for testing these model-based findings, but require a powerful statistical tool for extracting the observed land surface feedbacks. In this talk, I will first introduce a recently developed multivariate statistical method, the Stepwise Generalized Equilibrium Feedback Assessment (SGEFA), that is designed for extracting the role of terrestrial forcings on the regional climate. Second, I will present our recent examinations of the observed impacts of vegetation and soil moisture anomalies on the regional climate of North Africa through the application of SGEFA to an array of observational, reanalysis, and remote sensing data products. The observational SGEFA analysis identifies positive soil moisture-vegetation-rainfall feedbacks in the semi-arid Sahel under a moisture-recycling mechanism, and promotes the hypothesis about positive soil moisture-vegetation-dust-rainfall feedback. This hypothesis partly motivates extensive examinations of satellite-based dust emission and transport from North Africa, which will be presented in the final part of this talk.
May. 24Mark ZelinkaRound-table Discussion of clouds and climate
Round-table discussion of clouds and climate with speaker - Mark Zelinka
May. 28Informal Seminar - Mingjing TongSAIC @ GFDLAll-sky radiance assimilation using individual hydrometeors as cloud control variables for FV3GFS with GFDL microphysics scheme
The all-sky radiance assimilation framework of FV3GFS has been upgraded to use five individual hydrometeors as cloud control variables. This development effort was motivated by the model microphysics upgrade to the GFDL microphysics scheme. With this model upgrade, cloud liquid water, cloud ice, rain, snow and graupel become model prognostic variables. Adding precipitation hydrometeors allows the assimilation of precipitation-affected radiance in addition to cloudy radiance in the all-sky framework. The Community Radiative Transfer Model (CRTM) was carefully configured for the assimilation of precipitation-affected radiance. More stringent quality control was used to handle larger first guess departures when precipitation hydrometeors were included. Radiance observations from the Advanced Microwave Sounding Unit-A (AMSU-A) and the Advanced Technology Microwave Sounder (ATMS) over water were assimilated in all-sky approach. This new constructed all-sky framework works reasonably well. Improvement was found in 500 hPa geopotential height forecast and temperature forecast in the lower troposphere.
May. 29Lunchtime Seminar Series - Lei WangHarvard UniversityAtmospheric Blocking as an Evolution of Rossby Wave Packets
Atmospheric blocking is an important process for both weather and climate, yet its first-order dynamics is still not well understood. The eddy straining mechanism of Shutts (1983) has been considered as the foundation to understanding the maintenance of blocks, which is consistent with the observation that strong wave breaking concurs with blocks. Using a large-ensemble of a two-layer quasigeostrophic (QG) model, we find that, however, the generic effect from straining eddies is insignificant to the maintenance of the blocking pattern. We find that the evolution of coherent and propagating Rossby wave packets plays a vital role in block lifecycles. Evidences of such Rossby wave packets evolution, including its growing and decaying phases, have been identified in a hierarchy of climate models including a two-layer QG model, a dry GCM, and climate model simulations. Due to the dispersive nature of Rossby waves, this wave packets theory of atmospheric blocking predicts that energy dispersion should dictate blocking frequency. Among the non-dimensional parameters in the QG model, we find that the latitude exerts strongest influence on blocking frequency. At higher latitudes, energy dispersion is weaker and nonlinearity of eddies is stronger, both of which offer a conducive environment for block-like wave packet to remain its shape beyond synoptic timescale. As a result, an order-of-magnitude increase of blocking frequency is identified when a two-layer QG model is placed at higher latitudes. Consistently, we also find that the blocking frequency in idealized GCMs enhances significantly when the jet shifts poleward, regardless the change of the meridional temperature gradient. Theoretical considerations, including a conceptual model of energy dispersion and block lifecycles, will be discussed.
Jun. 5Lunchtime Seminar Series - Ben BronselaerGFDL Visiting Scientist - U of AZCurrent and future impact of Antarctic meltwater and Southern Hemisphere westerly winds
Meltwater from the Antarctic Ice Sheet is projected to cause up to one meter of sea-level rise by 2100 under the highest greenhouse gas concentration trajectory (RCP8.5). However, the effects of meltwater from the ice sheets and ice shelves of Antarctica are not included in CMIP5 models. CMIP5 models also underestimate historical trends in Southern Hemisphere westerly wind strength. To assess the current impact of wind and meltwater trends, we use pre-2005 ocean shipboard measurements alongside recent observations from autonomous floats with biogeochemical sensors to calculate changes in Southern Ocean temperature, salinity, nitrate, dissolved inorganic carbon, oxygen and pH over two decades. The forced response of GFDL ESM2M does not reproduce the observed patterns. Accounting for meltwater and poleward-intensifying winds in ESM2M qualitatively improves reproduction of the observed large-scale changes. We thus attribute recent warming, salinification and biogeochemical trends to these processes typically not represented in climate models. To assess the future impact of Antarctic meltwater, we assess a large ensemble simulation of ESM2M that accounts for RCP8.5-projected Antarctic Ice Sheet meltwater. Relative to the standard RCP8.5 scenario, accounting for meltwater delays global-mean atmospheric warming, enhances drying of the Southern Hemisphere and reduces drying of the Northern Hemisphere, increases the formation of Antarctic sea ice and warms the subsurface ocean around the Antarctic coast. The meltwater-induced subsurface ocean warming could lead to further ice-sheet and ice-shelf melting through a positive feedback mechanism, highlighting the importance of including meltwater effects in simulations of future climate."
Jun. 6Formal Seminar - Gilbert CompoCIRES, University of ColoradoNOAA-CIRES-DOE 20th Century reanalysis version "3" (1836-2015) and Prospects for 200 years of reanalysis
The new historical reanalysis dataset generated by the Physical Sciences Division of NOAA's Earth System Research Laboratory and the University of Colorado CIRES, the Twentieth Century Reanalysis version 3 (20CRv3), is a comprehensive global atmospheric circulation dataset spanning 1836 to present, assimilating only surface pressure and using monthly Hadley Centre sea ice distributions (HadISST2.3) and an ensemble of daily Simple Ocean Data Assimilation with Sparse Input (SODAsi.3) sea surface temperatures. SODAsi.3 was forced with a previous version of 20CR that itself was forced with a previous SODAsi, allowing these "iteratively-coupled" boundary conditions to be more consistent with the atmospheric reanalysis. 20CRv3 has been made possible with supercomputing resources of the U.S. Department of Energy and a collaboration with GCOS, WCRP, and the ACRE initiative. It is chiefly motivated by a need to provide an observational validation dataset, with quantified uncertainties, for assessments of climate model simulations of the 19th to 21st centuries, with emphasis on the statistics of daily weather. It uses, together with the NCEP global forecast system (GFS) numerical weather prediction (NWP) land/atmosphere model to provide background "first guess" fields, an Ensemble Kalman Filter (EnKF) data assimilation method. This yields a global analysis every 3 hours as the most likely state of the atmosphere, and also yields the uncertainty of that analysis. Host: Xiaosong.
Jun. 7Informal Seminar -Aditi SheshadriAssistant Prof., Dept. of Earth System Science, Stanford Univ.Mid-latitude jet variability: forcing from above and below
I will discuss aspects of midlatitude eddy-driven jet variability and its response to forcing from the stratosphere and by orography. The atmospheric jet stream over the North Atlantic exhibits three 'preferred positions', latitudes where the jet maximum occurs more frequently than others. Using a state-of-the-art atmosphere model, the forcing of these preferred positions by upper-atmosphere circulation and northern hemisphere mountain ranges are explored. Changes in the latitude of the time-mean jet manifest as changes in the relative probabilities of the preferred positions, not changes in the preferred latitudes. I will present strong evidence that the Greenland ice sheet is responsible for the northern preferred latitude. This is a cautionary tale of the interpretation of results from reducing the dimensionality of data - the preferred positions seem to imply complex jet dynamics; however, the northern peak can be explained by the existence of an orographically forced jet. I will also discuss stratospheric impacts on midlatitude jet variability in terms of Principal Oscillation Pattern analysis, which reveals separate tropospheric and stratospheric modes in idealized model integrations that are set up to resemble midwinter variability of the troposphere and stratosphere in both hemispheres. Host: V. Balaji
Jun. 12Lunchtime Seminar Series - Andrew RossGFDLForecasting estuarine temperature, salinity, and oxygen: data-driven and numerical modeling approaches
The ability to skillfully forecast conditions in estuaries and other coastal ocean regions is a challenge with a range of potential benefits including improved management of fisheries and water quality, protection of public health, and better preparedness for storm surge and flooding. In this talk, I present two different models for forecasting conditions in Chesapeake Bay, a large estuary along the U.S. East Coast. First, I develop a machine learning model that uses information about temperature, density, mean sea level, wind speed, and nutrient loading to predict column minimum dissolved oxygen (DO) concentration. This model shows that DO is primarily controlled by density stratification and temperature, and in most regions of the bay, DO concentrations can be skillfully predicted if the current states of these variables are known. However, forecasting DO in advance, when stratification and temperature are not known, is challenging. Experiments with the model show that skillful forecasts of stratification, with an RMSE of less than about 1 kg m-3, are necessary to skillfully forecast oxygen. Given the importance of forecasting stratification and the difficulty of using a data-driven approach to do so, I instead test whether skillful forecasts can be produced using a dynamical model of Chesapeake Bay. An extensive suite of reforecast simulations, in which the dynamical ocean model is driven by 35-day atmospheric forecasts from the GEFS weather model, shows substantial skill at forecasting sea surface temperatures up to about 15 days in advance, beyond the scale of weather predictability. Salinity and stratification can also be skillfully forecasted for the majority of the 35-day model simulations, likely as a result of the predictability of tides and the slow response of the estuary to river discharge forcing. Bottom oxygen forecasts do not compare well against observations, but hypoxic volume, a spatially integrated measure of oxygen, can be forecasted skillfully in some cases. The prediction skill is further examined in two case studies representing a summer heat wave and a landfalling tropical cyclone, two events with promising subseasonal scale predictability. Overall, these studies demonstrate the potential for substantially improved and useful forecasts of coastal and estuarine regions.
Jun. 19Lunchtime Seminar Series - Chris DupuisGFDLEDGI: Simple Machine Learning for Geoscience at the GFDL
Machine learning is often mystified in science fields, even when the basic scientific needs are relatively straightforward. We present an aggressively simple prototype command-line tool developed at the GFDL that combines principal component analysis with single-node parallelism, automatic NetCDF I/O, and a variety of basic and specialized flavors. Standard PCA with real- and complex-valued data, PCA of real-valued circular data, PCA of analytic singals, and PCA with multiple variables will be demonstrated. Further, new developments in C++ permit automatic parallelization of array-based statistics, which will inform first full-fledged version of EDGI PCA and a planned ensemble consistency test for the GFDL model.
Jun. 20Formal Seminar - Helen HewittUK Met OfficeOcean modelling for seamless prediction: resolution and process representation
Ocean model configurations developed by the Met Office (and partners) are utilised for applications from short-range ocean forecasting to climate projections ('seamless prediction'). The opportunities as well as the challenges and conflicting requirements of the seamless approach are described. The choice of ocean model resolution is discussed including the need to provide evidence to support the additional computational and storage costs of higher resolution. Alongside enhanced resolution, improved process representation in the model configurations is also a target and current efforts to improve overflows and convection are described.
Jun. 26Lunchtime Seminar Series -Tom KnutsonTropical Cylones and Climate Change: An Assessment Report
An assessment was made of whether detectable changes in tropical cyclone (TC) activity are identifiable in observations and whether any changes can be attributed to anthropogenic climate change. Overall, historical data suggest detectable TC activity changes in some regions associated with TC track changes, while data quality and quantity issues create greater challenges for analyses based on TC intensity and frequency. A number of specific published conclusions (case studies) about possible detectable anthropogenic influence on TCs were assessed using the conventional approach of preferentially avoiding Type I errors (i.e., overstating anthropogenic influence or detection). We conclude there is at least low-to-medium confidence that the observed poleward migration of the latitude of maximum intensity in the western North Pacific is detectable, or highly unusual compared to expected natural variability. Opinion on the author team was divided on whether any observed TC changes demonstrate discernible anthropogenic influence, or whether any other observed changes represent detectable changes. We also assessed published detection and attribution findings from a Type II error avoidance perspective, and presented summaries of future TC projections for a 2 degree Celsius global warming.
Jul. 3Lunchtime Seminar Series - Pu LinGFDLUnderstand the interaction between stratospheric ozone and climate
Changes in the stratospheric ozone play important roles in shaping the climate over the historical periods as well as in the warming future. While the direct radiative effects of ozone changes is straightforward to understand and is well simulated in most climate models, the full interaction between ozone and climate involves much more processes and lacks a complete understanding. Here, we compare AM4 simulations with prescribed ozone versus interactive ozone. With similar amount of ozone depletion over the historical period, the model with interactive ozone simulates stronger stratospheric cooling over the Antarctica. This is because the interactive ozone leads to a weaker damping to the waves in the stratosphere and hence a weaker dynamical heating over the polar region. In the SST warming simulations, models with interactive ozone simulate ozone changes as a result of a stronger overturning circulation in the stratosphere, which further leads to colder tropical tropopause and drier stratosphere. However, these changes in ozone and the associated stratospheric water vapor do not translate into any significant changes in the Cess sensitivity or tropospheric circulation patterns.
Jul. 8Informal Seminar - Yushi MoriokaJAMSTEC/VAiG/APL, Yokohama, JapanInterannual-decadal variability and predictability over the South Atlantic and southern Indian Oceans
Interannual-decadal sea-surface temperature (SST) variability in the South Atlantic and southern Indian Oceans has great influences on regional rainfall variability over the southern Africa through modulation of moisture transport. The Interannual SST variability, so called “Subtropical Dipole” which is characterized by a meridional dipole pattern of SST anomalies in each basin, is induced by ocean mixed-layer thickness variability associated with subtropical high variability. A series of climate model experiments reveals that besides remote forcing such El-Niño Southern Oscillation (ENSO) and the Southern Annular Mode (SAM), sea-ice extent variability over the Weddell Sea has a potential to influence the atmospheric variability over the South Atlantic. On a decadal timescale, the SST variability over the South Atlantic tends to propagate eastward as quasi-stationary oceanic Rossby waves under the influence of the Antarctic Circumpolar Current (ACC) and induce the decadal SST variability over the southern Indian Ocean. Decadal reforecast experiments with different ocean initial conditions show that the decadal SST variability over the South Atlantic (southern Indian Ocean) is predictable when the model's SST, subsurface ocean temperature and salinity are initialized (only the model's SST is initialized).
Jul. 18Lunchtime Seminar - Jong-Yeon ParkChonbuk National UniversitySeasonal to Multi Annual Marine Ecosystem Prediction with a Global Earth System Model
Climate variations have a profound impact on marine ecosystems and the communities that depend upon them. Anticipating ecosystem shifts using global Earth system models (ESMs) could enable communities to adapt to climate fluctuations and contribute to long-term ecosystem resilience.We show that newly developed ESM-based marine biogeochemical predictions can skillfully predict satellite-derived seasonal to multiannual chlorophyll fluctuations in many regions. Prediction skill arises primarily from successfully simulating the chlorophyll response to the El Ni?o?Southern Oscillation and capturing the winter reemergence of subsurface nutrient anomalies in the extratropics, which subsequently affect spring and summer chlorophyll concentrations. Further investigations suggest that interannual fish-catch variations in selected large marine ecosystems can be anticipated from predicted chlorophyll and sea surface temperature anomalies. This result, together with high predictability for other marine-resource?relevant biogeochemical properties (e.g., oxygen, primary production), suggests a role for ESM-based marine biogeochemical predictions in dynamic marine resource management efforts.
Jul. 19Informal Seminar - Casimiar de LavergneCNRS, LOCEAN Laboratory, Sorbonne University, ParisA parameterization of local and remote tidal mixing
Internal tides power much of the observed small-scale turbulence in the ocean interior. To represent mixing induced by this turbulence in climate-scale ocean models, energy routes from the generation to the dissipation of internal tides must be understood and mapped. Here we present a mixing scheme which accounts for the local breaking of high-mode internal tides and the distant dissipation of low-mode internal tides. The scheme relies on four static 2D maps of internal tide dissipation, constructed using mode-by-mode Lagrangian tracking of energy beams from sources to sinks. Each map is associated with a distinct dissipative process and a corresponding vertical structure. Applied to an observational climatology of stratification, the scheme produces a global 3D map of dissipation and mixing which compares well with available microstructure observations and with upper-ocean fine-structure mixing estimates. Implemented in the NEMO global ocean model, the scheme improves the representation of deep water-mass transformation and obviates the need for a fixed background diffusivity.
Jul. 23Intern Seminar - Drew PendergrassHarvard UniversityModeling the impact of near-term climate forcers on precipitation and air quality in India and China with the GFDL-ESM4 coupled model
Near-term climate forcers (NTCFs) are a major source of uncertainty in estimates of changes to the Earth's energy balance. Moreover, NTCFs including tropospheric ozone and fine particulate matter (PM2.5) cause serious public health burdens. We use the Geophysical Fluid Dynamics Laboratory Earth System Model version 4 (GFDL-ESM4) to investigate two key intersections between NTCFs and the Earth system: (1) mechanisms behind the apparent historical decline of March-April-May (MAM) mineral dust aerosol in northern India, and (2) the mid-21st century impact of a Chinese low-NTCF emission policy on air quality and climate, relative to a business-as-usual scenario. During the pre-monsoon months (MAM) in India, desert dust strengthens the monsoon circulation via the elevated heat-pump mechanism. However, recent observational studies note that pre-monsoon dust loading has declined. We link this decline to weakened emissions from the Thar desert in southeastern Pakistan and northwestern India and find that this shift cannot be explained by greenhouse gas induced warming alone. In China, we simulate future climate and air quality using the Shared Socioeconomic Pathway 3 scenario (SSP3) assuming a 7.0 W m2 increase in radiative forcing by 2100. We compare these to an alternate scenario in which emissions of NTCFs from China decline rapidly. We find that these NTCF emission reductions would, on balance, decrease surface PM2.5 but increase surface ozone by midcentury in China, with some seasonal variations. We also describe the impact of these NTCF emission changes on climate in China.
Jul. 24Lunchtime Seminar Series - Jun-Ichi YanoMeteo France, ToulouseTropical Atmospheric Madden-Julian Oscillation: Strongly-Nonlinear Free Solitary Rossby Wave?
The Madden-Julian oscillation (MJO), a planetary-scale eastward propagating coherent structure with periods of 30-60 days, is a prominent manifestation of intraseasonal variability in the tropical atmosphere. It is widely presumed that small-scale moist cumulus convection is a critical part of its dynamics. However, the recent results from high-resolution modeling as well as data analysis suggest that the MJO may be understood by dry dynamics to a leading-order approximation. Simple, further theoretical considerations presented herein suggest that if it is to be understood by dry dynamics, the MJO is most likely a strongly nonlinear solitary Rossby wave. Under a global quasi--geostrophic equivalent-barotropic formulation, modon theory provides such analytic solutions. Stability and the longevity of the modon solutions are investigated with a global shallow water model. The preferred modon solutions with the greatest longevities compare overall well with the observed MJO in scale and phase velocity within the factors. Host: Leo Donner
Jul. 29AOS Workshop Plenary Talk - SonnewaldElucidating complexity: providing robust system insight using unsupervised learning
The geosciences are becoming data rich both from modeling and observational studies. The available data can be overwhelmingly complicated when treated naively, but understanding the interactions in the data is important for gaining insight into the complex system. Unsupervised learning is a way to leverage data science tools to classify key interactions even in high-dimensional spatial datasets. The use of dimensionality reduction techniques and statistics are important to arrive at robust and reproducible classifications, where a naive approach would be misleading. In this lecture, the uses and misuses of unsupervised learning will be discussed with the aim of encouraging a careful and critical approach.
Jul. 30AOS Workshop Plenary Talk - Noah BrenowitzCoupling neural networks to GCMs
One of the most obvious ways to improve climate models, is to reduce the errors made by sub-grid-scale parameterizations. While there have been steady improvements to traditional sub-grid-scale parametrizations over the past decades, machine learning presents a much more direct path to reducing these errors based on realistic datasets. However, ML parameterization is not a straightforward prediction problem (e.g image classification) because any scheme must interact well when coupled to the dynamical core of a GCM. For instance, the ML schemes should avoid numerical instability. In this talk I will discuss our efforts to couple a neural network parmetrization of moist physics to a GCM with a focus on the challenges we overcame to produce stable and accurate weather forecasts.
Jul. 30AOS Workshop Panel Discussion Machine Learning and Climate ModelingAOS Workshop Panel Discussion Machine Learning and Climate Modeling
Panelists: Noah Brenowitz, Mike Pritchard, Maike Sonnewald, V. Balaji
Jul. 31AOS Workshop Plenary Talk - Mike PritchardAchieving conversation properties and probing inter-pretability using a refined deep learning emulator of global cloud super-parameterization
I will begin with an overview of preliminary work that in 2018 demonstrated surprising skill for a crude feed-forward deep neural network (DNN) to emulate the essence of O(10k) cloud-resolving models’ multiscale information flows within a superparameterized global climate model run on an aquaplanet. New tests are assessed that go beyond the aquaplanet and include realistic geography and seasons for the first time. These expose a potential challenge in learning highly bimodal rainfall relationships such as those that occur over arid continents – that is, a new “drizzle problem” is exposed, and a case for stochastic expansions to the method is made to resolve it. I will then introduce a new way to reformulate this DNN to achieve column conservation properties of mass, enthalpy and water, as well as self-consistency between vertical radiative heating profiles and their corresponding fluxes at top and bottom of atmosphere – all necessary features for operational use in ML-assisted climate prediction. I will conclude with preliminary results and current outlook regarding physical interpretability – for instance by exploiting the DNN’s Jacobian as a convenient summary of moist convection’s column energetics across multiple distinct basic states and the transitions between. Throughout the talk, I will emphasize the interesting opportunities and fas4 cinating philosophical tensions being raised by this disruptive new technology, and my personal hope that it might help achieve the potential of a true turbulence-permitting superparameterization ahead of schedule.
Jul. 31Intern Informal PresentationsTo be provided
Intern Informal Seminar by: Maurizia De Palma - Anthropogenic carbon uptake and ocean acidification in GFDL's CM4 and ESM4 models Alex Chang - An Evaluation of the GFDL C192-AM4 Model Simulations of Atmospheric Rivers
Aug. 5Xin Rong Chua-FPOThe Effects of Greenhouse Gases and Absorbing Aerosols on Tropical Precipitation
Xin Rong Chua-FPO
Aug. 8Intern Informal Presentations - Intern Informal Presentation
Intern Informal Presentations by: Alexandra Matthews - Where Phytoplankton Live and Die: A Global Perspective of the Biological Carbon Pump in Temperature Space. Nana Yaa Takyia Afrh - Exploring the trend and variability of Carbon Monoxide in ESM4
Aug. 9Justin Ng's FPOJustin Ng's FPO
Justin Ng's FPO POC: Anna Valerio
Aug. 12Spencer Clark - FPOControls on Tropical Mean State and Intraseasonal Precipitation Variability in an Idealized Moist Atmospheric Model
Spencer Clark - FPO
Aug. 14Three Intern Informal PresentationThree Intern Informal Presentation
Intern Informal Presentation by: Ana Bolivar - Simulated changes of North Atlantic air-sea heat flux feedback in a warmer climate. Mariela Arceo Madriz - Marine Organic & Sulfate Aerosols in CM4 and ESM4. Nkeh Perry Boh - Evaluating Parallel Computing with Dask on GFDL’s PP/AN.
Aug. 15Jenny Chang FPOEddy Equilibration in Idealized Models of the Extratropical Troposphere
Jenny Chang FPO POC: Anna Valerio (AOS)
Aug. 21Lunchtime Chat Series - Gabriel LauDirector, Institute of Environment, Energy an Sustainability, Univ of Hong KongRiverside chat with GFDL Personalities from the past
Chat Series I - discussion with Professor-cum-Director, Gabriel Lau, on his scientific career and experiences at GFDL, and his impressions of how GFDL science evolved through the decades.
Sep. 4Lunchtime Seminar Series - Monica MorrisonIndiana University, BloomingtonLocal Epistemologies, Model Purpose, and Pluralism of Methods in Climate Model Development
Climate modeling, like many areas of scientific practice, is composed of local research communities with distinct histories, whose practices are governed by their own local epistemologies. The local epistemology of a research community is constituted by the scientific aims, goals, assumptions, values, methodologies, and standards that operate in the research community and determine the course of research. In climate modeling, these local research communities and their associated epistemologies influence the course of model development by way of decisions about scientific goals, model purpose, and cognitive values, which are reflected in the subsequent decisions about representational priorities, implementation protocols, models of analysis, investigative strategies, methodological rules, and metrics and standards for determining model skill. Different local communities will have variations in the types of scientific problems they investigate, and differences in their other epistemic commitments, which ultimately results in a distinct perspective being taken by each local community on the complex system. For models this means that each model, which is the product of a local community, will partition the causal space according to their specified interests and related representational priorities—each model will thus pick out slightly different features of the complex causal system and represent them in slightly different ways consistent with their epistemology. Given this picture of modeling practice there are many models that are perspectival in nature, but no model that is a complete representation of the complex system. The perspectival nature of models might seem problematic because only certain features of the causal space are being represented and surveyed in a given model. However, I demonstrate that in historical and contemporary modeling practices there exists a plurality of communities, epistemologies, and associated models that exhibit representational pluralism, thus providing a multitude of different representational perspectives on the complex system. To maximize knowledge and increase degrees of objectivity, a plurality of models is desirable, and this desirable pluralism is a historical and contemporary feature of modeling practice.
Sep. 11Lunchtime Seminar Series - Dr. Yun QianPacific Northwest National Laboratory (PNNL)Uncertainty Quantification in Climate Modeling
DOE - Energy Exascale Earth System Model (E3SM) is a new model and has included many new features in the physics parameterizations compared to its predecessors. Potential complex nonlinear interactions among the new features create a significant challenge for understanding the model behaviors and parameter tuning. Using the one-at-a-time method, the benefit of tuning one parameter may offset the benefit of tuning another parameter, or the improvement in one target variable may lead to degradation in another target variable. To better understand the model behaviors and physics, we conducted a large number of short simulations in which 18 parameters carefully selected from parameterizations of deep convection, shallow convection and cloud macrophysics and microphysics were perturbed simultaneously using the Latin Hypercube sampling method. From the Perturbed Parameters Ensemble (PPE) simulations and use of different skill score functions, we identified the most sensitive parameters and their global spatial distribution, quantified how the model responds to changes of the parameters, and estimated the maximum likelihood of model parameter space for a number of important fidelity metrics. Results from this analysis provide a more comprehensive picture of the E3SM model behavior at global scale as well as at process level in different cloud regimes. The difficulty in reducing biases in multiple variables simultaneously highlights the need of characterizing model structural uncertainty (embedded errors) to inform future development efforts. In this talk, besides presenting above parametric sensitivity results in E3SM atmosphere model, I will also highlight a few applications using Uncertainty Quantification technics in model calibration and optimization, wind energy and aerosol study conducted at PNNL. Finally, I will discuss a few challenges and paths forward in this area. POC: Sarah Kapnick, and Paul Ginoux.
Sep. 18Lunchtime Seminar Series - Gan ZhangAOSReconciling Predictability and Uncertainties in Seasonal Predictions and Future Projections of Tropical Cyclone Activity
Improvements of dynamic models have transformed long-range predictions of tropical cyclone (TC) activity, and these improvements also open up new paths for exploring the predictability and uncertainties for such predictions. The first part of this talk focuses on the future projection of TC activity on the regional scale. The regional projection is currently considered highly uncertain because of uncertainties related to the large-scale environmental changes. Here we focus on some future environmental changes that are relatively robustly simulated by CMIP5 models but received little attention among TC researchers. Such environmental changes include a weakening of extratropical eddy activity and a poleward shift of midlatitude westerlies. Using idealized regional simulations, we show that a dramatic suppression of extratropical weather variability may profoundly affect TC activity in the subtropics, including the storm frequency and their movement. Using more realistic global large-ensemble simulations, we show that potential future changes in extratropical circulation with global warming could affect TC propagation. The findings may have important implications for populated coastal regions outside the tropics, where TC-related risks are relatively low in the current climate. The second part of this talk investigates the seasonal predictability of TC activity and pathways to improve dynamic prediction systems. Using the ensemble hindcasts by GFDL's FLOR prediction system, we show that TC activity in coastal regions and/or at higher latitudes is sensitive to uncertainties of initial conditions. Our analysis also suggests that the seasonal predictability of regional and basin-wide TC activity might be higher than the skill that has been realized by pre-existing FLOR prediction systems. Using idealized prediction experiments, we show that the gap might be narrowed by reducing ocean biases or improving the initialization of land-atmosphere components. With improved simulations of the large-scale atmospheric environment in the tropics and/or the extratropics, the skill gains of seasonal TC prediction are statistically significant. The promising findings suggest new research opportunities and will help with the design of next-generation prediction systems of TC activity.
Sep. 19Formal Seminar - Joao TeixeiraJPLTurbulence, Convection and Clouds in Global Atmospheric Models: The Unified EDMF Approach
The parameterization of turbulent and convective mixing in global atmospheric models has been a major challenge in weather and climate research for several decades. In particular, different parameterizations are used, and patched together often artificially, for different types of convection: dry or moist, in the boundary layer or in the full troposphere. The Eddy-Diffusivity (ED) approach has been relatively successful in representing the properties of neutral and stable boundary layers and surface layers in general. The Mass-Flux (MF) approach, on the other hand, has been used for the parameterization of shallow and deep moist convection. In this presentation, an approach based on an optimal combination of the ED and MF parameterizations (EDMF) is discussed in detail as a solution for the full unification of the parameterizations of turbulent and convective mixing in atmospheric models. In particular, we will present results from a new multi-plume stochastic EDMF parameterization that fully unifies the representation of convection in weather and climate models: One single parameterization that represents the effects of dry, shallow and deep moist convection in the atmosphere.