GFDL Past Events & Seminars - 2013

Representing topography with porous barriers

Physical mechanisms of cloud-aerosol interaction at different scales

In order to derive climate statistics, long and homogeneous time series are needed. For the western North Pacific different observational data sets (best track data) of tropical cyclone activity show strong discrepancies in derived trends for the last decades. As an alternative we therefore employ an atmospheric regional model (CCLM) to derive tropical cyclone activity changes by dynamically downscaling NCEP/NCAR re-analyses for the last decades. The reconstructed intense tropical cyclone's variability shows good agreement with the observed one in best track data, both on inter-annual and decadal time scales. Changes in intense tropical cyclones are associated with large-scale patterns of thermodynamic factors, like Maximum Potential Intensity, sea surface temperature (SST). Canonical correlation analysis showed that the relationship between SST and TC activity is very similar for both: simulated and observed (BTD provided by Japan Meteorological Agency) TC data sets. Trends for the annual number of tropical cyclone days show an increase for the period 1948-2011, with a short decrease in the last decade. Intense tropical cyclones increasingly influence the subtropical latitudes in the western North Pacific. The upward trend of tropical cyclone activity in the South China Sea region is mainly due to weaker storms. Decreasing tropical cyclone activity is found in the south-eastern part of the western North Pacific. Overall, the results indicate an increase and north-westward tendencies of track patterns for the period 1948-2011, which is consistent with more recent observations.

Water-soluble organic gases are ubiquitous in the atmosphere at the surface and throughout the atmospheric column. The chemical potential for these water-soluble organic gases to partition to particle phase liquid water (H2Optcl) and form secondary organic aerosol (SOA) is high in the Eastern U.S., higher than for partitioning pathways whereby semi-volatile organic gases partition to dry organic matter (OM). The higher potential of the wet pathway is driven by the predominance of water-soluble gases and higher H2Optcl mass concentrations relative to particle phase OM. The spatial extent, in particular dominance in the Eastern U.S. is driven by patterns in H2Optcl. The temporal and spatial co-location of water-soluble organic gas phase compounds and particle phase liquid water may resolve a key discrepancy whereby biogenic SOA mass concentrations in the southeast (SE) U.S. are high, but low in the Amazon. In the SE U.S., anthropogenic pollution and biogenic emissions routinely mix and H2Optcl concentrations are predicted to be high, largely as a consequence of sulfur pollution. The Amazon is not impacted by anthropogenic pollution (e.g., NOx, SOx) to the same degree and H2Optcl concentrations are predicted to be low. The ability of anthropogenic pollution to increase H2Optcl is well-established and routinely included in atmospheric models, however the potential for gas phase organic species to partition to this liquid water and form SOA is not. These findings suggest the current estimate that roughly half of biogenic SOA in the eastern U.S. forms only when anthropogenic pollution is sufficient to facilitate formation is too low.

Visualizing the Ocean's Overturning with Surface 14C Measurements

The history of GFDL computing: a user perspective

Estimation of Surface Fluxes of Carbon from Atmospheric Data Assimilation Experiment

Labor capacity reduction from heat stress under climate warming

Orographic convective precipitation in the tropics: Results from the Dominica Experiment.

Constraining transient climate sensitivity using coupled climate model simulations of volcanic eruptions

Mixing in Estuaries and River Plumes

Revisiting the Rectifier Using Millions of LIDAR Soundings from CALIPSO

Deep convection triggering by boundary layer thermals: Stochastic formulation and parametrization for climate models

The global evaporation (E) and precipitation (P) surface fluxes set the spatial pattern of surface salinity in the ocean. Collectively, these E-P ocean surface fluxes comprise 75-85% of the climatological annual-mean global water cycle. For this reason long-term changes to the ocean salinity field provide an insight into water cycle change expressed by E-P changes. When considered in a simplified framework, ocean salinity responds to an amplification and redistribution of E-P and not E or P changes individually, with the advantage that salinity integrates these noisy E-P changes (providing a low-pass filter) over oceanic circulation timescales. Additionally, as the ocean contains 97% of the Earth's free water it's clearly a good place to look for small temporal perturbations in the observed record. Using the CMIP (Coupled Model Intercomparison Project phase 3 & 5) simulations, we diagnose the relationship between the simulated ocean surface salinity and the simulated E-P (water cycle)changes. Using the 20C3M/historical (20th century) and SRES/RCP (future 21st century)simulations, explicitly dealing with model drift, and a technique to extract the broad-scale, zonalchange patterns, a strong linear relationship is found between changes in the global surface water flux (E-P) and surface salinity over the global oceans. The CMIP simulations indicate that spatial salinity patterns amplify around twice the rate of the corresponding E-P patterns. This result suggests that fresh ocean regions become fresher, and salty regions saltier in response to E-P changes driven by climate change - an ocean proxy for the 'rich get richer' (wet gets wetter, dry gets drier) mechanism. Observed near-surface salinity estimates suggest a salinity spatial pattern amplification of 8% (16°C -1) has occurred 1950-2000. Using the CMIP relationship between salinity and E-P change, we infer a 3-4% amplification of observed E-P for 1950-2000. Considering the observed global surface warming of 0.5°C over this period, this suggests a 6-8% °C -1 rate of change, closely following Clausius-Clapeyron (7% °C -1). Importantly, the CMIP ensemble mean 20th century values (CMIP3 4% °C -1 and CMIP5 ~2% °C -1), a frequently used metric to express projected future changes,greatly underestimates the observed 1950-2000 rate of ocean salinity change (8% °C -1). Further, if the CMIP ensemble mean warming of 2-3°C by 2100 is realised; this would imply a water cycle amplification of 16-24% will occur. Additional investigations into the full three-dimensional pattern of salinity- and temperature-driven steric changes in the global oceans will also be discussed. Basin-scale depth-integrated salinity changes strongly supporting the concept of an enhanced water cycle (E-P) over the observed record. Some discussion of recent detection and attribution studies which considered observed and CMIP5 changes and recent modifications to the global surface temperature anomaly timeseries (HadCRUT3 & HadCRUT4) and implications for rates of water cycle changes will also be briefly covered.

Assessing ENSO risks for the coming decades

Bi-weekly meeting

CO2 First: The case for deferring action on short lived climate pollution for at least 50 years

Diagnosis of Seasonally Dependent Predictability in Observations and CM2.5

Early investigations by Lorenz suggested that atmospheric circulations with characteristic scales of 20-40 km would have very limited predictability. Later researchers arrived at much more optimistic estimates, noting that many mesoscale phenomena are produced i) by interactions between known small-scale forcing agents, like topography or land-sea contrasts, and the relatively predictable large scale flow, or ii) by the large-scale features themselves through processes such as frontogenesis. We revisit the question of whether mesoscale features forced by the larger-scale flow enjoy enhanced predictability relative to what might be expected based on turbulence models. The predictability of lowland snow in the Puget Sound region of the Pacific Northwest is explored by analyzing the spread in 100-member ensemble simulations for two events from December 2008. Sensitivities to the microphysical and boundary-layer parameterizations in these simulations are minimized by estimating the likely precipitation type from the forecast 850-hPa temperatures and the established rain-snow climatology. Our results suggest the ensemble spread in events such as these, which were triggered by amplifying short waves, may include a significant fraction of both rain-likely members and snow-likely members at forecast lead times as short as 36 hours. The relation of these results to those from classical turbulence models is explored by examining the perturbation kinetic energy of the ensemble members about the ensemble mean . The initial spectral power in is not maximized at small scales. Instead, the initial spectrum of produced by our EnKF data assimilation cycle increases with increasing horizontal scale. The power in subsequently grows with time, while maintaining approximately the same spectral shape. There is no evidence of small-scale perturbations developing rapidly and transferring their influence upscale. Instead, the large-scale perturbations appear to grow more rapidly during the first 12 hours than those at the smallest resolved scales.

Discerning Environmental Dependencies on Mixed-Phase Cloud Lifetime with a Focus on Ice Particle Habit Evolution

TBA

Food production is forecasted to double in the next 40 years to meet growing food demand. The resulting increase in agricultural emissions will likely have adverse effects on human and ecosystem health. Here I focus on ammonia (NH3) emissions, to which agriculture contributes 80%. I first present a new "top-down"estimate of NH3 emissions derived from adjoint inversion of observed ammonium wet deposition fluxes. Then I show that a newly developed detailed "bottom-up"inventory of agricultural NH3 emissions can successfully reproduce many of the seasonal and geographical features of the "top-down"NH3 emissions. Finally, I use this inventory to estimate the effect of US agricultural exports on air quality.

The interplay between external forcing and internally generated decadal timescale variability is explored through analysis of case studies of multi-decadal climate shifts, focusing particularly on the Pacific. The Interdecadal Pacific Oscillation (IPO) in its positive phase adds to warming from external forcing to contribute to accelerated warming decades like the mid-1970s shift. The IPO in its negative phase counteracts warming from external forcing to contribute to decades with little warming such as the early-2000s hiatus. In the CCSM4 in future climate simulations, hiatus periods with zero global warming trend can last for 15 years due to this internal variability. Initialization with observations produces improvement over uninitialized free-running 20th century simulations for the mid-1970s shift and early-2000s hiatus. A CMIP5 multi-model data set of 30 year predictions shows about 16% less global warming for the period 2016-2035 partly due to initialization with observations during the cooler hiatus, and partly due to a reduced trend from bias adjustment. Initialization also improves predictions of area-averaged Pacific-region precipitation compared to the uninitialized projections for the mid-1970s shift and early-2000s hiatus.

Climate Impact Assessments 2.0.

Precessional Forcing, Monsoons and Isotopic Composition of Precipitation

TBA

Past and Future Hurricane Activity

Over the past thirty years, Arctic sea ice has rapidly declined. This has led to a greater interest in marine access to the region and reliable predictions of ice conditions. However, little is known about the predictability characteristics of sea ice on interannual to decadal timescales. Here I present results that diagnose mechanisms giving rise to potential predictability in Arctic sea ice in Community Climate System Model (CCSM) integrations. This includes an analysis of statistical relationships from control and 20th-21st century integrations and how these change with long-term sea ice loss. It also includes analysis of perfect model studies designed to investigate the initial-value predictability of sea ice. These perfect model ensemble integrations are initialized with identical ice-ocean-terrestrial conditions allowing us to diagnose the potential predictability that reside with those initial conditions on seasonal to interannual timescales. Finally, a brief comparison to Antarctic sea ice predictability characteristics and mechanisms is provided.

Global modeling of oceanic internal tides within an eddying general circulation model

Challenges in measuring tropical carbon dynamics in the Anthropocene

Guyot Hall Room 220 - Main Campus

Discover what Allinea MAP can do for your application performance Topic: Introducing performance profiling with Allinea MAP and revealing how complex parallel performance issues are easily identified in your source code.

Neoproterozoic Snowball Earth Initiation in CCSM

Ensemble Forecasts Using the GFDL Hurricane Model

Direct estimates of eddy mixing across the Antarctic Circumpolar Current

Skillful predictions of North Atlantic decadal variability in the GFDL forecast system

Systematic errors associated with tropical oceanic rainfall have challenged the global modeling community for decades. Among these errors are tendencies toward a double ITCZ and the over-prediction of regional rainfall. Also related to systematic errors is low predictive skill associated with the Madden-Julian Oscillation (MJO). This seminar addresses three related topics: 1) the influence of SST gradients on the excitation of deep moist convection, as evidenced by the onset of rainfall; 2) common regimes of organized convection and propagating event lifecycles; 3) correlation of rainfall with lower boundary forcing at short-range and MJO frequencies, as well as with mesoscale SST structure over the four year period of record. This research is observations based together with application of established theory. We employ four-year timeseries of satellite estimated rainfall and daily SST. A strong statistical association is revealed between mesoscale SST gradients and the location of rainfall onset. Preferred locations for rainfall onset are in the mid-range of the background SST distribution. The warmest SST locations are not favored (i.e. neutral) and the coolest are disfavored. The lifecycles of especially long-lived rainfall events are characterized statistically. These events live longest and rain strongest in tropospheric shear > 10-3 s-1, and dissipate in environments with reduced shear and cooler SST. Events of three days duration produce disproportionately large rainfall compared to events of one-day duration, perhaps causally related to specific environmental differences. The timeseries of rain and SST have been filtered at both short-range weather and MJO scales. There are phase relationships at short-range, fully consistent with coupled ocean-atmosphere responses. For the MJO, it appears that an SST positive anomaly leads MJO rainfall, essentially irrespective of the MJO phase, whereas the -Laplacian of SST appears to lead MJO rainfall by ~15-20 days specifically at its most active phase (5) in the western Pacific.

Conference

To view the slideshow: 1. Click the link below or copy and paste the link to a browser: https://cpomapp.webex.com/cpomapp/onstage/g.php?t=a&d=628988255 2. Enter your name and e-mail address, and click Join Now. If necessary, enter the event passcode: 20910 3. To hear the audio, utilize the on-screen dial-in instructions visible after logging into webex Webex and the teleconference line can accommodate only 100 attendees on a first-come, first-served basis. Please try to share a connection with colleagues at your institution to preserve space for others.

Informal Seminar followed by Round Table discussion; Abstract: Underestimating rainfall, especially during the dry season, over Amazonia is a common problem of the Coupled Model Intercomparison Project phase 3 (CMIP3) models. Such a dry bias is an important source of uncertainty in seasonal rainfall assessment and projecting global carbon-climate feedbacks. Our evaluation of the CMIP5 historical simulations shows that some models still tend to underestimate rainfall over Amazonia. During the dry season, GFDL-ESM2M and IPSL show notably more pentads with no rain. In the dry and transition seasons, models with more realistic moisture convergence and surface evapotranspiration generally have more realistic rainfall amounts. In some models, overestimates of rainfall all associated with the adjacent tropical and eastern Pacific ITCZs, whereas in other models, too much surface net radiation and a resultant high Bowen ratio, appears to cause underestimates of rainfall. During the transition season, low pre-seasonal latent heat and high sensible flux and a weaker influence of cold air incursions contribute to the dry bias. About half the models can capture, but overestimate, the influences of teleconnection. HadGEM2-ES outperforms other models, whereas GFDL-ESM2M has the strongest dry bias presumably due to its overestimated moisture divergence induced by overestimated ITCZs in adjacent oceans and high Bowen ratio from the surface. One common way to tackle with the projection uncertainty is using multi-model ensemble mean (MMEM). However, it might cause a shifted distribution of rainfall when a considerable number of the models tend to have dry bias on regional scale. Our study suggests the understanding of individual model performance using a Bayesian framework can help us to put more weights on good models, and the weighted MMEM can largely reduce the present biases and increase our confidence in future projection of the climate mean state. Another important issue lies in the change of rainfall extremes, largely depending on the assessment of rainfall distribution. In this case, we look at the underlying distribution of daily precipitation and reconstruct it based on all available models using parametric statistical method, i.e. estimate shape and scale parameters from Gamma distribution fitted for models. We will compare the results with observations and discuss the applicable regions of this method.

Land-atmosphere interactions occur when the response of the land surface (soil and vegetation) to atmospheric variations feeds back, positively or negatively, on these atmospheric variations. These feedbacks mainly originate from soil moisture anomalies, conveyed to the atmosphere primarily through their impact on surface heat fluxes. The resulting physical coupling between the land and the atmosphere has the potential to affect regional mean climate, climate variability and extremes. Diagnosing these interactions is thus a crucial element of a process-based evaluation of climate models. In this presentation we show results from ongoing activities at GFDL aimed at investigating land-atmosphere coupling in some of GFDL's climate models. In a first part, we evaluate the feedback between surface turbulent energy fluxes and precipitation, which is one of the key issues associated with land/atmosphere interactions. Using the NARR reanalysis data, Findell et al. (2011) recently provided an assessment of the impacts of morning surface latent and sensible heat fluxes on the frequency and intensity of afternoon convective rainfall over North America. Here, we apply the same metrics to AMIP-like simulations from GFDL's atmospheric models AM2.1 and AM3. We show that these models perform differently in representing the impact of surface fluxes on convection. The sources of these differences are investigated. In a second part, we analyze results from GLACE-CMIP5 simulations performed at GFDL: i.e., historical and climate change transient AMIP-like simulations performed with ESM2M, with prescribed or interactive soil moisture. Here we focus on the impact of soil moisture dynamics and associated feedbacks to the atmosphere on the distribution of daily near-surface temperature. We show that soil moisture dynamics strongly enhance temperature variability over apparent regional 'hotspots', at different time scales over different regions. Analysis of the daily distribution of the different variables involved (soil moisture, surface fluxes, temperature) also highlights changes in higher-order moments of variability, including skewness and bimodality: soil moisture dynamics disproportionately impact the high side and tail of the temperature distribution. In addition, we show that soil moisture-atmosphere interactions appear responsible for the interannual coupling between temperature and precipitation variability over land.

Numerical models used to forecast daily weather and those used to make projections for longer-term climate share much in common - after all, "physics is physics". However, the relative importance of different factors associated with the global atmosphere-ocean system varies depending upon the time scale, region, and application of interest, which in turn, leads to there being some important differences in the way weather and climate models are developed. In this talk, we will discuss how global climate models serve as tools used by researchers to investigate and gain a better understanding of the way the global climate system "works". Topics to be covered include the contrast between an initial value problem and a boundary value problem, the role of the three-dimensional ocean circulation in climate, sources of uncertainty associated with multi-decadal climate projections, and how climate models are used to study whether observed climate changes are due to natural variability or are human-induced. Like all models, climate models have their strengths and weaknesses. We will conclude by pondering the question of what measure(s) one might use to assess how "good" a climate model truly is - a more philosophical question than one might think.

Over recent years, work on several climate change assessments focused on the Northeast and New York has identified several key stakeholder concerns regarding future climate change impacts and adaptation strategies in the agriculture, natural resources, and flood management and preparedness sectors. In almost all cases, high spatial (and in some cases temporal) resolution downscaled climate projections have emerged as a key stakeholder data need. Three specific example are discussed, highlighting differences in downscaling methodology and identifying weakness and advantages of different approaches. For snow cover, a key determinant for species survival, the establishment of invasive pests and rural economic viability, projections of snow cover change cited in the Northeast Climate Impacts Assessment (NECIA) are contrasted with those using a variant of the Statistical Downscaling Methodology. To quantify agricultural frost risk, a stochastic approach in conjunction with climate projections from the NARCCAP is described. Preliminary work to provide estimates of extreme rainfall return frequencies for engineering design standards in New York is outlined in some detail. Convenient access to downscaled climate data and products is also an often-heard desire from stakeholders. In this regard the Northeast Regional Center has developed a web services application, GridData. The design and benefits of this software is discussed in terms of its convenience in both research and climate service applications.

Remote sensing data and information are shown great potential in supplying relevant spatial data and parameters at the appropriate scale for use in distributed hydrological models for water resource applications. In contrast with many conventional data normally represented by point measurements, remote sensing based measurements are spatially averages over the pixels can appropriate for distributed hydrological model. Furthermore, remote sensing enables data access from remote areas, where data are typically sparse. Remote sensing technology used electromagnetic spectrum in the range of wavelengths of different radiations reflected or emitted by objects. There are two main types of remote sensing: passive remote sensing and active remote sensing. The passive systems are based on the measurement of the natural thermal emission in the form of brightness temperature from the earth surface. On the other hand, the active microwave systems generate their own radiation, which is transmitted toward the earth surface, and measures the reflected energy. Three main elements are necessary for a reliable forecast and monitoring of hydrological processes, namely, good-quality precipitation data, accurate characterization of surface conditions and robust, accurate and detailed hydrological models. Here, the focus is placed on satellite-based land surface and precipitation products and their integration into operational hydrological models.

Flooding over the Central US

The basic roles of tropical deep convection and extratropical baroclinic eddies in the atmospheric general circulation have been studied for a very long time, but these studies have been limited by the paucity of truly global observational detail to just a few basic atmospheric parameters (temperature, humidity, horizontal winds) and to more detailed, but infrequent case studies over tropical oceans and northern hemisphere land areas. The advent of global synoptic observations from satellites makes possible a more general investigation, especially with the relatively recent increase in the number of quantities measured. Much more detail is now available that can be used to study, in particular, cloud processes - precipitation and radiation - in association with the atmospheric disturbances that produce clouds. Early progress in this work has focused on the role of "storm"cloud processes in the energy part of the atmospheric circulation; these results will be summarized. More remains to be done as will be discussed. The momentum part of the problem is more difficult because observations of the wind field are not sufficient for direct diagnosis of storm effects. This part of the problem will likely require careful model investigations using the newly diagnosed storm energetics as constraints. Some ideas along these lines will be outlined.

A convective quasi-equilibrium view of observed monsoons

Interactions between natural and anthropogenic activities

I will discuss recent research linking extreme stratospheric wave propagation (upward and downward) to North Atlantic weather and climate. The extreme events significantly impact the residual circulation in the stratosphere and are connected to North Atlantic Oscillation-like anomalies in the troposphere. The representation of these links in Coupled Model Intercomparison Project Phase 5 (CMIP5) models, including different versions of the GFDL model, and their connection to biases in the Atlantic jet stream will be discussed.

Preindustrial to present-day changes in tropospheric hydroxyl radical and methane lifetime from the Atmospheric Chemistry and Climate Model Intercomparison Project (ACCMIP)

Howard University has lead a series of maritime field experiments supporting research and observations of the microphysical characteristics (optical properties, surface characteristics, chemical composition, deposition rates, and size- & mass distributions) of mineral dust aerosols and their impact on the marine troposphere annually since 2004. These are known as the AERosols and Ocean Science Expeditions or AEROSE. The AEROSE project is a comprehensive interdisciplinary mission that returned multiple, high quality, and unique data sets. The majority of the missions (seven of the past eight) and the upcoming 2013 cruise) are conducted aboard the NOAA Ship Ronald H. Brown. These activities involve comprehensive trace gas and aerosol monitoring and aerosol collection including observations of mineral dust microphysics, optical properties, and chemical impacts on the regional and global atmosphere. Howard University teams also participate in continuous monitoring stations (El Paso, Texas and Isla Magueyes, Puerto Rico), at-sea investigations of short and long duration and execute field intensives in collaboration with NOAA and other partners. These field programs are excellent student training opportunities and generate data sets that can be utilized for improvement of model parameterizations in weather, climate, & air quality forecast models, for satellite validation, and for the improvement of retrieval algorithms (Morris et al 2006, Nalli et al 2011). A survey of the results and analysis from the AEROSE campaign will be presented. In particular, insights pertaining to the mass distribution dynamics and surface composition will be highlighted

Decadal, or near-term, climate prediction, whereby climate models are initialized from the observed climate state, has recently gained much attention. Results tend to show that such initialization improves predictions of past surface temperature, especially over the North Atlantic Ocean. Such skill in predicting North Atlantic temperature is encouraging given its important impact on the wider climate. However, the reasons for the improvement in sea surface temperatures are not fully understood, and there is currently much less evidence of useful predictions over land. In this talk I'll highlight how examining predictions of case studies of decadal change are an important route to build confidence in decadal predictions. By examining two prominent case studies of North Atlantic decadal change - the cooling in the 1960s, and the warming in the 1990s - it is found that predictions can successfully capture these two events. Interestingly, the initialization of the ocean circulation was found to be key for both events. Finally, by focusing on these large decadal-change events, we also show that predictions also capture many of the surface climate changes that were observed, particularly over North America and Europe.

Parametrizations for surface fluxes and convection: Who is to blame for the tropical precipitation biases in the ECHAM climate model?

From climate change to marine ecosystems with CMIP5 models: multiple stressors, end-to-end modelling, decadal predictability

Eddy memory and the persistence of atmospheric annular modes

Symposium

Influence of Decadal Variability of Global Oceans on South Asian Monsoon and ENSO-Monsoon Relation

This work explores the predictability of El Niño flavors in a 4000-year pre-industrial run of the GFDL CM2.1 coupled GCM, which has a reasonably realistic ENSO simulation. Central Pacific (CP) and Eastern Pacific (EP) flavors are defined in the phase space of the two leading principal components (PCs) of tropical Pacific sea surface temperature anomalies. The predictability of the different El Niño flavors is quite limited due to the intrinsic chaotic property of the climate system. The interference of two leading transient growing modes is shown to contribute to El Niño diversity. The precursors (i.e. optimal initial patterns) of these modes in the simulation are diagnosed using linear inverse modeling and singular vector analysis, which are then applied in a statistical model to forecast the probability, given any initial state, of evolution into each El Niño type. We find that the horizon to distinguish the precursors of flavors in the PC space is ~3 months before a CP El Niño peak or 6 months before an EP El Niño peak. The approach in this work is potentially useful for evaluating coupled GCMs, both as a dynamical diagnostic and as a better baseline for forecast skill than persistence.

The world's major Eastern Boundary Currents (EBC) such as the California Current Large Marine Ecosystem (CCLME) are critically important areas for global fisheries. Computational limitations have divided past EBC modeling into two types: high resolution regional approaches that resolve the strong meso-scale structures involved, and coarse global approaches that represent the large scale context for EBCs, but only crudely resolve only the largest scales of their manifestation. These latter global studies have illustrated the complex mechanisms involved in the climate change and acidification response in these regions, with the CCLME response dominated not by local adjustments but large scale reorganization of ocean circulation through remote forcing of water-mass supply pathways. While qualitatively illustrating the limitations of regional high resolution studies in long term projection, these studies lack the ability to robustly quantify change because of the inability of these models to represent the baseline meso-scale structures of EBCs. In the present work, we compare current generation coarse resolution (one degree) and a prototype next generation high resolution (1/10 degree) Earth System Models (ESMs) from NOAA's Geophysical Fluid Dynamics Laboratory in representing the four major EBCs. We review the long-known temperature biases that the coarse models suffer in being unable to represent the timing and intensity of upwelling-favorable winds, along with lack of representation of the observed high chlorophyll and biological productivity resulting from this upwelling. In promising contrast, we show that the high resolution prototype is capable of representing not only the overall meso-scale structure in physical and biogeochemical fields, but also the appropriate offshore extent of temperature anomalies and other EBC characteristics. Results for chlorophyll were mixed; while high resolution chlorophyll in EBCs were strongly enhanced over the coarse resolution ESM, they were still considerably lower than observed values. In terms of representation of large scale circulation and biogeochemistry, results were also mixed, with the high resolution prototype addressing some, but not all, of the biases in the coarse resolution ESM. While considerable work remains to understand the current strengths and weaknesses of the high resolution ESM and continue to improve fidelity, this work is a major step forward in demonstrating the added value of high resolution in global ESMs and represents a fundamental leap forward towards both ecological forecasting and long term projection of climate, ecosystem, and acidification baselines and sensitivity.

Cumulative carbon emissions and irreversible climate change: New insights

Important, rich, complex and highly uncertain are aerosol indirect effects (AIEs) on clouds. To help models to obtain better estimate of impacts of AIEs, observations are indispensable in constraining model physics and results. Here we use ideal, natural experiments to probe aerosol indirect effects in two cloud regimes with satellite data. In one experiment, we show strong evidence of aerosols invigorating maritime tropical convection at a large scale. The invigoration effect manifests in characters of precipitation radar reflectivity vertical profiles, cloud top ice particle size and cloud glaciation temperature. Furthermore, lightning, as a hallmark of strong convection, increases at a rate of 20-40 times per unit increase of aerosol optical depth. Aerosol-induced lightning changes also have interesting implications for ozone chemistry and wildfire activity. In the other experiment, we show that strong modifications of trade cumulus cloud fields are associated with increased sulfate aerosols at a large scale. They include decreased droplet size, decreased precipitation efficiency and increased cloud amount. In addition we find significantly higher cloud tops for polluted clouds. The observationally inferred "total shortwave aerosol forcing" is almost an order of magnitude higher than aerosol direct forcing alone. It highlights the strong leverage of AIE in this cloud regime. Furthermore, the precipitation reduction associated with enhanced aerosol leads to large changes in the energetics of air-sea exchange within trade wind boundary layer. I will briefly discuss how such observations might be used to constrain model results.

Biological processes such as photosynthesis and respiration are now widely represented in modern climate models because their influence on global carbon cycling is a potentially important climate feedback. However, ecological mechanisms - those governing the adaptation, assembly and acclimation of biological communities - are much less common. Does the neglect of these mechanisms matter for the accuracy of climate change predictions? I will focus on two case studies, one from the tropical rainforests of South America, and one from arctic permafrost peatlands, biomes rich in carbon (and hence, key to understanding carbon cycle feedbacks to climate), but representing very different modeling challenges: in the Amazon I ask whether ecological scaling theories that connect forest size structure to canopy dynamics and carbon cycling are relevant at large scales, using remote sensing observations from airborne LiDAR forest surveys to estimate forest size structures across Amazonian landscapes. In the Arctic case study, integrating ecosystem-scale measurements of methane isotopes with next generation high-throughput DNA sequencing of methanogenic communities, I investigate whether it is necessary for climate models to represent microbial community ecology to accurately predict methane cycle feedbacks during the decomposition of thawing permafrost.

The humidity of the atmosphere plays a large role in the radiative balance of the planet. Therefore understanding the processes that control the response of the tropical humidity to climate forcing is critical for determining climate sensitivity. A correct account of the complex set of processes involved requires consideration of large-scale transport, turbulence and cloud microphysical processes. While these are all included in climate models, many aspects associated with them are treated in a very simple parameterized form and are poorly observed. These limitations reduce the confidence one may place in the fidelity of simulated responses. Modern measurement capabilities, including in situ and satellite remote sensing, allow isotope ratio information to help constrain some aspects of processes acting in and near clouds, but required estimation methods that better account for uncertainties and biases in both the models and observations. We discuss both the importance of these constraints and some of the limitations associated with using stable isotope tracers in assessing the balance the microphysical controls against influences of large-scale moisture transport.

CMIP5 Model Assessments of 20th Century Regional Surface Temperature Trends and of Two Extreme 2012 Climate Events

Land use impacts on chemistry-climate interactions

Publishing with Nature: a climate science perspective

IPCC AR5 WGI

Internal-wave driven mixing and climate models

Resolution-dependent Eddy Parameterizations for Large-Scale Ocean Models

Climate driven variability of phytoplankton biomass in the United States Northeast Shelf

Climate predictability and prediction on seasonal to multi-year scales

Ocean mixing by internal tide breaking at topography

The changing terrestrial water cycle: Climate dynamics from biogeochemical clues

Statistical Downscaling: Is past performance an indicator of future results?

Atmospheric signals of changing global biogeochemistry

The AMOC and the Garden of Eden

The effect of human activities on climate is one of the most important issues facing society. Humans influence climate in many ways. Emissions of greenhouse gases (GHGs) tend to warm climate, by reducing the amount of infrared radiation that is emitted to space. Increased levels of suspended atmospheric particles ("aerosols") exert a net cooling effect by directly scattering and absorption of solar radiation (the "aerosol direct climatic effect"). Aerosols also affect clouds by acting as the seed for droplet (or ice crystal) formation; polluted clouds tend to have more droplets than their pristine counterparts, and affect their reflectivity, size, lifetime and ability to precipitate. It is thought that aerosol impacts on clouds (known as "aerosol indirect climatic effects") have a net cooling effect on climate. Despite their importance, aerosol impacts on clouds constitute one of the most uncertain components of climate, significantly affecting predictions of climate sensitivity to GHG levels. This uncertainty originates largely from the complex and multi-scale nature of aerosol-cloud interactions, which truly challenges the description of all processes involving aerosols and clouds. A major focus of our research is to improve the description of aerosol-cloud interactions in climate model frameworks through the combination of observations, theory and modeling. We will present key findings and demonstrate how modeling can be combined with in-situ observations to constrain process uncertainty. We will also demonstrate new approaches (based on adjoint sensitivity analysis) to quantitatively understand sources of variability in model simulations, and the spatiotemporal sensitivity of cloud parameters & processes to aerosol.

The global urban population has been increasing over the past few decades. Currently, over 50% of the people live in cities and this percentage is expected to keep increasing in the near future. As such, understanding the impact of urbanization on local and regional climate is important. What is equally important is correctly forecasting the impact of future climate change in cities, despite of their small area fraction globally. In this presentation, I will present some of my Ph.D. work on urban heat island modeling with the Weather Research and Forecasting (WRF) model, illustrating the importance of urban surface parameterization. Then I will focus on my plan of introducing an urban tile to the land model at GFDL and using the urbanized land model, in together with the high-resolution atmospheric model, to examine urban environmental issues that are centered on water sustainability such as extreme rainfall and heat waves.