All posters will be up for display in the lobby of the conference center for review and discussion throughout the conference. Poster Presenters will give a 1-minute-1-slide presentation at the end of Wednesday's sessions. The conference attendees will then converge in the lobby for the Welcome Reception and Poster Review. Posters are listed below by Presenting Author Last Name.
Precipitation Partitioning Across Grey Zone Scales Using Scale-Aware Cloud Formulations: Impacts of Aerosols
Summary:The Weather Research & Forecasting model with Aerosol Cloud Interactions has been developed to investigate the scale dependency of aerosol-cloud interactions (ACI) across the "grey zone" scales for grid scale clouds (Morrison Scheme) and subgrid scale clouds (Multi-Scale Kain-Fritsch scheme) that include cloud microphysical processes. The impacts of ACI are examined across regions in the eastern and western U. S. at 36, 12, 4, and 1 km grid spacing for short-term periods during the summer of 2006. ACI impacts are determined by comparing simulations with current climatological aerosol levels to simulations where aerosol levels have been reduced by 90%. The aerosol cloud lifetime effect is found to be the dominant ACI process leading to suppressed precipitation in regions of the eastern U.S., while regions in the western U. S. experience offsetting impacts on precipitation from the cloud lifetime and thermodynamic invigoration effects of aerosols. Subgrid scale ACI are dominant at 36 km while grid scale ACI are dominant at 4 and 1 km, however, grid scale and subgrid scale ACI are comparable at 12 km. The equivalent impacts of grid scale and subgrid scale processes at 12 km lead to peculiar behavior of ACI, indicating competition for available moisture by the grid-scale and subgrid-scale cloud formulations. The work implies including subgrid scale convective cloud microphysics and ice/mixed phase cloud ACI processes may be necessary in weather and climate models to study ACI.
A novel ensemble design for fine particulate matter probabilistic predictions and quantification of their uncertainty
Summary:The National AQ Forecasting Capability (NAQFC) at the National Oceanic and Atmospheric Administration (NOAA) produces forecasts of ozone, particulate matter, and other pollutants so that advance notice and warning can be issued to help individuals and communities limit the exposure and reduce health problems caused by air pollution. The current NAQFC, based on the U.S. Environmental Protection Agency (EPA) Community Multi-scale AQ (CMAQ) modeling system, provides only deterministic AQ forecasts and does not quantify the uncertainty associated with the predictions, which could be large given the chaotic nature of atmosphere and nonlinearity in atmospheric chemistry. We explore and evaluate an ensemble prediction system of PM2.5 with a novel design. The ensemble is generated by perturbing three key aspects of PM2.5 modeling namely the meteorology, emissions, and CMAQ secondary organic aerosol (SOA) formulation. The combination of different meteorological, emissions, and SOA formation perturbations results in a large number of ensemble members which could be computationally unfeasible for NAQFC operations. We have addressed this limitation by down selecting the combination perturbations which preserve the most part the skill and quality of the original ensemble. The proposed ensemble significantly improves the accuracy of operational PM2.5 predictions while providing a reliable quantification of the PM2.5 prediction uncertainty.
Ongoing improvements to surface-layer turbulence modeling in the Weather Research and Forecasting model
Summary:The accuracy of large-eddy simulations is highly dependent on the surface boundary condition. Here, we will present ongoing improvements to the surface boundary condition and surface-layer turbulence treatment in the Weather Research and Forecasting (WRF) model. First, we will discuss development related to immersed boundary methods (IBMs), which allow WRF to be used in regions of steep terrain. While IBMs have previously been tested and validated for very high resolution (~1-10 m), we will focus on their performance as grid resolution coarsens (~10-100 m). Through comparisons to similarity theory, observations, and native WRF results, we will reveal the sensitivity of IBMs to grid resolution and surface turbulence treatment. Second, we will discuss our implementation of a canopy model framework for improved representation of surface roughness. The model includes two options: a resolved canopy model for use when roughness elements are resolved by the vertical grid and a newly-developed "pseudo-canopy model" for use when roughness elements are unresolved. Both options are validated against idealized test cases and shown to improve surface layer velocity profiles relative to simulations employing a traditional boundary condition based on similarity theory. Together, these model improvements expand WRFâ€™s applicability to regions of steep, forested, or otherwise complex terrain and improve the accuracy of such simulations.
Emissions, Transport, and Chemistry of Smoke from Western U.S. Wildfires
Summary:Air quality forecasts using regional chemical models provide key information for affected communities and smoke management efforts, yet many models fail to accurately predict ozone (O3) and particulate matter levels during fire events. The satellite-based emissions and plume rise are large sources of model uncertainty. To improve emissions and plume rise parameterizations, we utilize aircraft and ground-based data from recent field campaigns, such as the 2018 NSF/CU Biomass Burning Fluxes of Trace Gases and Aerosols using SOF on the Wyoming King Air (BB-FLUX) and NSF/CSU Western wildfire Experiment for Cloud chemistry, Aerosol absorption and Nitrogen (WE-CAN) field campaigns, and the 2019 NOAA/NASA Fire Influence on Regional and Global Environments Experiment - Air Quality (FIREX-AQ) field campaign. Hourly fire emissions based on Geostationary Operational Environmental Satellite (GOES)-16/17 fire radiative power are implemented in WRF-Chem. Emission factors (EFs) are updated from estimates from FIREX Fire Lab and BB-FLUX/WE-CAN/FIREX-AQ field observations, and separate EFs are implemented for flaming and smoldering combustion. Uncertainties in emissions and plume injection heights in the model are quantified by comparison with aircraft- and satellite-based estimates. WRF-Chem simulations are also compared with satellite retrievals of trace gases and aerosols, and are used to quantify fire air quality impacts and examine formation/aging mechanisms for O3 and SOA.
Effect of biomass burning on Light-Absorbing Particles (LAP) vs. snow albedo reduction on Central Andes: the analysis of WRF-Chem modeling
Summary:The Central Andes (CA) of Argentina and Chile (30°S and 40°S), includes the largest glaciated areas in South America (SA) outside southern Patagonia. This water reserve is relevant because it is the main water source of the largest metropolitan area in Chile, Santiago and the fourth largest city in Argentina, Mendoza. Recent studies conducted by our research group show that the presence of atmospheric aerosols would be related to the negative trend variations of snow albedo during the last almost 20 years, which leads to an increasing the snowmelt. Besides, in CA there is a high degree of uncertainty about the identification and characterization of the light absorbing particles (LAP) as Black carbon (BC) and organic carbon (CO) that contribute the most to the decrease of the albedo. In this work, inventories of regional anthropogenic emissions of own elaboration and inventory of burning of high resolution are incorporated to the simulations carried out with the WRF-Chem model to study the relationships: LAP spatial distributions - snow albedo. The partial results show that air masses from Brazil and northern Argentina are reaching the snowy areas of the CA and at the same time there is a snow albedo decrease. The results suggest that the LAP from the open field biomass burning are being deposited on the snow and decreasing it albedo, which produces a negative impact on snow and hydrological resources generated in the CA.
Evaluating the impact of assimilating aerosol optical depth observations on dust forecasts over North Africa and the East Atlantic using different data assimilation methods
Summary:This study evaluates the impact of assimilating Moderate Resolution Imaging Spectroradiometer (MODIS) aerosol optical depth (AOD) data on dust analyses and forecasts over North Africa and the North Atlantic. In addition, the use of different data assimilation (DA) methods and the impact of deep-blue-algorithm-retrieved AOD data on AOD analyses and forecasts are also evaluated. Seven DA cycling experiments with 72-h forecasts for June 2015 are conducted using the Weather Research and Forecasting (WRF) dust model and the Gridpoint Statistical Interpolation (GSI) analysis system. These experiments differ in whether all or only dark targeting AOD data are assimilated and which DA method, among the three-dimensional variational (3D-Var), ensemble square root filter (EnSRF), and hybrid methods, is used. The analyses and forecasts from all experiments are verified against MODIS AOD, Aerosol Robotic Network (AERONET) AOD, and Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) backscatter profiles. The assimilation of MODIS AOD data clearly improves AOD analyses and forecasts, and the benefit on AOD forecasts can last for 48 hours. Results also show that the assimilation of deep-blue data improves AOD analyses and forecasts over most of source and surrounding regions but also degrades some areas. On average, the deep-blue AOD data have a positive impact on AOD analyses and forecasts and the exclusion of the data reduces the benefit of AOD assimilation on AOD forecasts for 18 hours. Among the three examined DA methods, the hybrid and EnSRF methods, which benefit from flow-dependent background error covariance, produce better AOD analyses and forecasts than the 3D-Var method. However, while the assimilation of AOD data improves AOD analyses and forecasts, its impact on the vertical profiles of backscatter analyses and forecasts is marginal due to the lack of the vertical distribution information in AOD observations. Backscatter analyses and forecasts are slightly better when the 3D-Var and hybrid methods are used, and this is in part because an adjoint AOD operator is used in these two methods.
Evaluation of PBLH simulated by WRF using a new LiDAR network in California
Summary:The planetary boundary layer height (PBLH) is a key parameter for air pollution and greenhouse gas (GHG) modeling. In California, most of the long-term PBLH measurement sites are located in coastal regions. However, a majority of the emission sources are concentrated in the inland area, especially in the San Joaquin Valley (SJV). This highlights the need for additional PBLH measurements in the inland areas. To meet this demand, the California Air Resources Board has installed a new LiDAR network comprised of high-fidelity laser-based ceilometers at five ground-based sites across California to monitor atmospheric vertical layers based on the aerosol backscatter measurements. This is a first-of-its-kind statewide network developed to collect long-term, high-resolution, atmospheric vertical measurements. The network also complements the statewide PBLH retrievals at inland locations. In this study, we evaluate PBLHs simulated by the Weather Research and Forecasting (WRF) model using three types of PBLH retrievals, including ceilometers, radiosondes, and wind profilers, to improve the estimates of GHG emissions in California. PBLH retrievals from the three datasets will be used to evaluate PBLH information in the WRF simulations by combining different PBL schemes and land-surface schemes. The presentation will discuss results comparing the WRF PBLH and observed PBLH retrievals, and their impacts on the estimation of top-down GHG emissions budgets.
Exploring future climate effects on northwestern US air quality
Summary:Air quality regulations have reduced emissions of pollutants in the US, but many studies suggest that the future air quality still can be degraded by global change factors. Those studies were typically based on computationally expensive 3D Eulerian CTMs. To study how future air quality at a local scale will be influenced by global factors in an efficient way, we have developed a Lagrangian air quality modeling framework, called HYSPLIT-MOSAIC. It consists of HYSPLIT, an air trajectory model developed by NOAA, and MOSAIC, a gas and aerosol chemistry and dynamics model developed at PNNL. To apply HM, we employ 4-km gridded statistically downscaled meteorology data from 20 global climate models for two RCPs from the MACA compilation. The future anthropogenic emissions are from EPA projections, and biogenic emissions are generated using the MEGAN model. The initial and boundary conditions are from the ModelE2-TOMAS global model output. We apply HYSPLIT cluster analysis to generate the representative back trajectories using historical NAM meteorology data. To evaluate our approach, we have applied HM to simulate air quality at 55 AIRNOW sites along with the CMAQ-based AIRPACT5 air quality forecasts over the Pacific Northwest since 2018. Overall our model results are comparable to the observations and 3D model results. We will run simulations for these sites for 2030s and 2050s and estimate the climate change effects on future O3 and PM2.5 concentrations under various scenarios.
Effects of urban land use on meteorology and atmospheric chemistry in Pacific Northwest urban areas
Summary: Pacific Northwest (PNW) region has experienced significant urban growth in recent years. Urbanization modifies the distribution and structure of the atmospheric boundary layer, causing variations in the thermodynamic, radiative, dynamic, and physical mechanism, and consequently affecting the concentration and transport of atmospheric pollutants and thus the air quality in a city and its downwind areas. The goal of this project is to investigate the impact of urban land use on regional and urban weather and air quality over the main cities of the PNW. This work primarily focuses on the Spokane urban area. The WRF-Chem modeling system is used coupled with Urban canopy models under different weather conditions. World urban database and access portal tool (WUDAPT) is an open access project based on Local Climate Zones (LCZ) classification framework and includes up to 10 urban classes that represent better the diversity of urban structures, in contrast to the limited number of urban classes in the land use classifications available in WRF. We adopted the WUDAPT approach to introduce those urban classes and urban canopy parameters into the WRF-Chem simulations. Atmospheric turbulence, wind and temperature fields, surface-level ozone and particulate matter concentrations will be analyzed and compared to observations. We expect to find a better comparison between simulations and observations when urban canopy models are included in the WRF-Chem system. Additional Authors: Alex Schmies , Amit Sharma , Brian Lamb , Yunha Lee 
 University of Minnesota
 Laboratory for Atmospheric Research, Washington State University
On the evaluation of air quality forecasts during the ORACLES 2018 campaign
By: Gonzalo Ferrada, The University of Iowa
Assessing the Goddard Earth Observing System model in non-resolved to convection-permitting regimes
Summary:We evaluate the Goddard Earth Observing System (GEOS) modeling and assimilation through a cascade of simulations with increasing horizontal resolution. The GEOS model is driven by the finite-volume cubed-sphere (FV3) non-hydrostatic dynamical core, a set of scale-aware physics package and data assimilation capability. The GEOS model is run for 30-days beginning on 01 August 2016 at seven uniform global resolutions of approximately 200, 100, 50, 25, 12, 6, and 3 km with 72 vertical levels up to 0.01mb. The model physics uses the Grell-Freitas (GF) scale-aware convection scheme to dynamically reduce the role of parameterized deep convection as resolved scale processes in the model take over at higher resolutions. Shallow convection is parameterized following the UW scheme, and cloud microphysics applies a single-moment formulation for rain, liquid and ice condensates. We provide an overview of the scale dependence of precipitation across the spatial scales, geometric characteristics of the simulated cloud fields, and comparison of IR brightness temperature with observations. We also include a look at the capability of GEOS to realistically represent the diurnal cycle of precipitation over global and land only regions across those scales
A Comparison of MPAS and WRF Meteorological Models in California: 2013 Winter and 2016 Summer Case Studies
Summary:The Modeling Platform Across Scales - Atmosphere (MPAS-A) and Weather Research and Forecast - Advanced Research (WRF-ARW) models were applied to simulate meteorological conditions over California for wintertime (January 10-24, 2013) and summertime (July 18-30, 2016) episodes during the Discover-AQ and CABOTS field campaigns, respectively. WRF was configured with three, two-way nested domains of 36, 12, and 4 km horizontal grid resolution, while MPAS was configured with a variable horizontal grid resolution ranging from 3 km over California to 48 km over the entire globe. Both models were configured with similar vertical structure and physics parameterizations as consistent as possible. Model results were evaluated against surface measurements throughout California using the METSTAT and AMET modeling analysis packages, and were compared using horizontal and vertical cross sections of various meteorological parameters, temporal evolution at surface stations, as well as 3D Lagrangian transport analysis. While the results are preliminary, the initial evaluation shows that MPAS is able to reasonably reproduce the observed meteorology and its characteristics, and may even exhibit superior performance to WRF over some regions. Future work will include additional simulations with higher vertical resolution and differing horizontal grid structures.
Using WRF-STILT to Determine the Relative Contributions of US and Mexican Emissions to High Ozone Events in El Paso, Texas
Summary:Over the past few years, the U.S. Environmental Protection Agency (EPA) has set increasingly stringent National Ambient Air Quality Standards for ozone (O3), and the design value for El Paso in recent years has been close to the latest ozone standard. As El Paso is on the border between the US and Mexico and directly borders Ciudad Juarez, the air quality in El Paso is influenced by emissions from both the US and Mexico. Therefore, it is useful for regulatory purposes to separate days where El Paso had high O3 primarily due to Mexican emissions of O3 precursors from days where the high O3 is primarily due to US emissions. We used the Stochastic Time-Inverted Lagrangian Transport (STILT) model driven by high resolution (1km) Weather Research and Forecasting (WRF) simulations to identify upwind source regions O3 precursors. The WRF-STILT framework calculates emissions "footprints" for each measurement that can be combined with bottom-up emission inventories to estimate the relative impacts of sources from different geographic regions. The WRF-STILT footprints were calculated for the 10 highest MDA8 O3 values in the El Paso area for each year from 2012 -2018. These WRF-STILT footprints were then applied to the bottom-up estimates of NOx and VOC emissions from the Emissions Database for Global Atmospheric Research (EDGAR) to determine the relative influence of US and Mexican sources of NOx and VOCs on high O3 days in El Paso.
Atmospheric chemistry modeling using machine learning
Summary:The simulation of chemical kinetics - central to air quality modeling - is computationally expensive due to the stiffness of the system of ordinary differential equations that describes atmospheric chemistry. Here we present an alternate approach to the computation of chemical kinetics based on machine learning. We find that tree methods such as random forest and XGBoost offer good predictability, and further improvements can be achieved through mass balance considerations and by accounting for error correlations. The machine learning approach shows many of the characteristics of the full chemistry reference simulation and has the potential to be orders of magnitude faster. For instance, the evaluation of a single tree in the random forest is more than 1000 times faster than the reference chemical solver. The machine learning methods are also much more parallelizable than conventional numerical solvers and naturally lend themselves to new soft/hardware environments such as GPUs and cloud-based computing. There is a wide range of applications for an atmospheric chemistry module based on machine learning, e.g. for use in air quality forecasts, sub-grid parameterization, chemical data assimilation, or inverse modeling.
What causes the observed surface ozone-temperature relationship? Effect of the eddy-driven jet on surface-level transport
Summary:The eddy-driven jet (EDJ) and the transient high- and low-pressure disturbances developing in its vicinity are important components of synoptic-scale circulation and variability in the mid-latitudes and have been shown to impact surface-level air quality. Here we investigate the role of these large-scale circulation features on surface-level ozone (O3) and its covariance with temperature in the Northern Hemisphere mid-latitudes from in-situ observations and within the framework of a chemical transport model (CTM). Spatially, we find that the location of the EDJ coincides with strong positive correlations between O3 and temperature and the largest O3-temperature sensitivity, while the highest O3 mixing ratios prevail ~5-10Â° south of the EDJ. Also examined are the net effects of cyclones and anticyclones light of differences in emissions in different regions and how these systems can ventilate O3 from the boundary layer. These large-scale transport features not only help explain the historical relationship between O3 and temperature, but also have implications for air quality under climate change. We discuss pathways by which this could occur as well as uncertainty associated with future air quality.
Evaluation of the online multiscale MONARCH model to forecast air quality over Europe
Summary:The Barcelona Supercomputing Center develops the Multiscale Online Nonhydrostatic AtmospheRe CHemistry model (MONARCH) to research on the atmospheric composition. MONARCH provides regional dust forecasts for the World Meteorological Organization (WHO), and for the WMO Sand and Dust Storm Warning Advisory and Assessment System (SDS-WAS). Additionally, MONARCH contributes to the International Cooperative for Aerosol Prediction (ICAP). Currently, it is a candidate in the European Air Quality Ensemble Forecast of Copernicus Atmosphere Monitoring Service (CAMS). This work presents the ongoing MONARCH evaluation activities to forecast European air quality. MONARCH uses the High-Elective Resolution Modeling Emission System (HERMESv3) to process the CAMS emission inventories and the Global Fire Assimilation System. It integrates MEGANv2.0.4 to estimate online the biogenic emissions. Initial and boundary conditions come from the Integrated Forecasting System of the European Centre for Medium-Range Weather Forecasts and the Copernicus global forecasting system for the chemistry. The MONARCH's evaluation uses concentration of key pollutants from the main European monitoring network. In addition, we compare MONARCH with current European operational prediction systems. Finally, to support evaluation in the frame of the European Air Quality Directive, we use the Delta tool developed by the Forum for Air Quality Modelling in Europe to benchmark the MONARCH against agreed quality standards.
Source apportionment modelling to unravel the origin of tropospheric ozone peaks over southwestern Europe
Summary: Spain is a Western Mediterranean country, which experiences a severe O3 issue due to the favorable climatological conditions for its formation. However, the contribution of the different sources of precursors to O3 formation within the country relative to the imported O3 is poorly quantified. This lack of quantitative knowledge prevents local authorities from effectively designing plans that reduce the exceedances of the O3 target value set by the European directive. This study aims at estimating the contribution of activity sectors to peak O3 events in Spain relative to the contribution of imported O3. We used the Integrated Source Apportionment Method within the CALIOPE air quality system at 4-km resolution using the bottom-up HERMESv2 emission model. Our study shows, for the first time, that imported O3 is the largest input to the surface O3 concentration in the Iberian Peninsula, accounting for 46%-68% of the daily mean O3 during exceedances of the European target value. However, during stagnant conditions, the local anthropogenic precursors control the O3 peaks in areas downwind of the main urban/industrial regions (up to 40% in hourly peaks). We also show that surface O3 concentrations are strongly affected by vertical mixing of O3-rich layers present in the free troposphere, which result from local/regional layering and accumulation, and continental/hemispheric transport.
Evaluation of AQ models: what we miss with limited information
Summary:Two major field campaigns - the NSF/NCAR and State of Colorado Front Range Air Pollution and Photochemistry Experiment (FRAPPÃ) and the NASA DISCOVER-AQ - took place jointly in summer 2014 to study the drivers of summertime ozone pollution in the Northern Colorado Front Range (NFR). A comprehensive suite of chemical and meteorological measurements was collected from in total four research aircraft, six heavily instrumented ground sites with in-situ and remote sensing instruments, additional surface ozone monitoring sites, six mobile labs as well as tethered balloons and ozone sondes to provide a 3D picture of the chemical and meteorological characteristics of the area. This contrast the about dozen of operational surface ozone monitoring sites in the NFR and the even fewer surface sites with CO or NOx measurements and the infrequent canister samples at two locations that are typically available for evaluating air quality models. Using the WRF-Chem model, we demonstrate how the additional information from the field campaign might change the conclusions drawn about model performance compared to findings based on evaluation with operationally available data alone. We will not only demonstrate the importance of available information above the surface but also the additional benefit from information on solar radiation and boundary layer heights.
Interactions between meteorology and chemistry during wildfire season over Western US
Summary:Air quality over Western US deteriorates severely during wildfire season (June to October) when mixture of air pollutants is released due to burning of biomass. Pollutants like particulate matter and ozone precursors are dispersed into the atmosphere and even carried away with the winds to locations otherwise unaffected directly by wildfires. Besides impacting air quality directly, these pollutants can alter meteorology through changing air temperature, relative humidity, cloud cover and photolysis rates which, in turn, can alter the atmospheric chemistry. In this study, we use the online Weather Research and Forecasting model (WRF-Chem; version 3.9.1) coupled with MOZART-MOSAIC chemistry option to investigate these impacts over western US. The wildfire emissions are incorporated using Fire Inventory from NCAR (version 1). Model results are first evaluated against a network of observations to provide confidence into model's ability to reproduce observed variations in meteorology and pollutant concentrations during wildfire season. Next, sensitivity simulations are conducted to analyse how wildfires affect meteorology and air quality.
Simulation of the land-atmosphere exchange during persistent cold air pool events in Salt Lake Valley, Utah
Summary:The multiday stagnant meteorological conditions associated with persistent cold air pools (PCAPs), especially the vertical structure of the atmospheric boundary layer (ABL) below 1 km in the atmosphere, have been notoriously difficult to capture using numerical weather prediction models. The land-atmosphere exchange controls the energy exchange at the interface of surface and atmosphere, which impacts the ABL development. Thus, well-represented land-atmosphere exchange in numerical models is a prerequisite for reliable meteorological predictability. We characterized surface turbulent fluxes during PCAPs and investigated the surface atmospheric stability impacts on surface turbulent fluxes. The simulated surface exchange coefficient (Ch), one crucial parameter used to simulate surface sensible heat fluxes in the Weather Research and Forecast (WRF) model, was compared with values estimated from observations. Suppressed surface turbulence with lower magnitudes of sensible (H) and latent (LE) heat fluxes are observed during strong PCAP events compared with non-PCAPs. The surface turbulent fluxes are impacted by net radiation, the presence of a PCAP, as well as the PCAP type (cloudy or dry). The significantly overestimated H in WRF is related to an overestimated value for Ch. The non-dimensional vertical temperature gradients used in the stability functions based on the Monin-Obukhov similarity theory are underestimated in WRF and are responsible for the overestimated Ch.
Preliminary Assessment of Micro-Pulse LiDAR Measurements of the Mixed Layer Height in the San Joaquin Valley
Summary:The mixed layer height is a critical meteorological variable, serving as a diagnostic for uncertainties in models for aerosol dispersion and determining the volume pollutants are mixed due to turbulence. Estimation of the mixed layer height (MLH) using the vertical distribution of aerosols, however, is complicated by nocturnal residual layers and local transient aerosol layers. In this work we utilize an adaptive covariance wavelet transform of lidar signals to determine continuous paths of sunrise to sunset MLHs in the San Joaquin Valley (Fresno Supersite) over three seasons. We also apply a novel method -cyclostationary analysis - to the data set to separate diurnal components of vertical distributions. This provides insight into the types and frequencies of several carry-over events and how these contribute to ground level aerosol concentrations. The method also gives an alternative estimate of MLH based on the dominant diurnal component. A comparison of these results is presented along with previous estimates of convective boundary layer heights in the San Joaquin Valley.
How would a regional nuclear war affect the global climate?
Summary:Urban firestorms in the aftermath of the detonation of nuclear weapons produce plumes of black carbon (BC) in the atmosphere. Atmospheric BC can cause global climate impacts including surface cooling, increased ultraviolet radiation flux, and decreased precipitation. Previous climate modeling demonstrates that the magnitude and duration of the climate impacts from BC are dependent on how much BC lofts into the stratosphere in the days to weeks following a detonation. The amount of BC that lofts into the stratosphere is dependent on the evolution of the BC plume-particularly its height-in the seconds to hours after a detonation, which cannot be resolved by a global climate model. We take a new approach to estimating how much BC lofts into the stratosphere for a previously studied scenario in which India and Pakistan each detonate 50 fifteen kiloton nuclear weapons over urban areas. Our method combines BC plume rise modeling using the Weather Research and Forecasting Model (WRF) large-eddy simulations (LES) with climate simulations using the Energy Exascale Earth System Model (E3SM). The disparate modeling techniques are unified by a common atmospheric state, allowing for a more complete test of the sensitivity of the plume rise into the stratosphere to atmospheric conditions at the moment of detonation. Prepared by LLNL under Contract DE-AC52-07NA27344.
Empirical estimation of posterior emission flux errors
Summary:Uncertainty of the posterior estimation is essential for interpreting the solution from an inverse analysis. Some inversion methods (e.g., analytical inversion) allows direct computation of posterior errors, but observation-based evaluation of these estimations often lacks. Evaluation of an inverse analysis is routinely performed by comparison in the concentration space, which does not readily provide information about the flux uncertainties. Here, we present a method to perform evaluation in the flux space by applying the gain matrix to independent evaluation data, providing an empirical estimate of the posterior flux errors. We perform a global inversion of methane emission flux on a 4x5 grid using GOSAT satellite observations from 2010-2017 and evaluate the posterior flux errors with AIRS satellite observations. We find a general agreement in the flux errors between the empirical and analytic estimation. Particularly, both find smaller relative errors in regions with large averaging kernel sensitivities such as Amazon and Eastern China. The implication and limitation of the method are also discussed.
Assessment of Climate change impact over California for wintertime using dynamic downscaling with a bias correction technique
Summary:The regional WRF model is used to dynamically downscale the global climate model (CESM) results from CMIP5 to a resolution of 4km for a present decade (2003-2012) and a future decade (2046-2055) to explore climate change impacts on meteorological variables and atmospheric phenomena (such as precipitation and stagnation events) during wintertime over California. The CESM results are bias-corrected using an advanced statistical method prior to the downscaling process. California is divided into different geographical sub-regions and the climate projections over these sub-regions are discussed separately. Comparison between WRF downscaled results for the present and future decades suggests an up to 2 K increase in winter mean surface temperature over California by the 2050s. Our results show a winter precipitation increase (decrease) for the northern (southern) California. The connections between the future changes in precipitation and landfalling atmospheric rivers on the U.S. West Coast will be explored. Winter mean surface relative humidity is projected to increase in the future, especially over the northern California, which can be attributed to the precipitation increase over this region. Under climate change, future boundary layer height and ventilation index are projected to decrease over California, which would exacerbate the winter particulate matter problem within the Central Valley. Future potential changes in winter stagnation events will also be investigated.