Search NASA⌕ Search

SEARCH · Search NASA

Results for “(CHEM)”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 271 records · Page 15
All sourcesNASA NTRSDOE OSTI
NASA NTRS2003

Foam Sensor Structures Would be Self-Deployable and Survive Hard Landings

A document proposes systems of sensors encased in cold hibernated elastic memory (CHEM) structures for exploring remote planets. Lightweight structures that can be compressed for storage and later expanded, then rigidified for use are made from foams of shape-memory polymers (SMPs). According to the instant proposal, a CHEM sensor structure would be fabricated at full size from SMP foam at a temperature below its glass-transition temperature (Tg). It would then be heated above Tg and compacted to a small volume, then cooled below Tg and kept below Tg during launch, flight, and landing. At landing, the inelastic yielding of the rigid compacted foam would absorb impact energy, thereby enabling the structure to survive the landing. The structure would then be solar heated above Tg, causing it to revert to its original size and shape. Finally, the structure would be rigidified by cooling it below Tg by the cold planetary or space environment. Besides surviving hard landing, this sensor system will provide a soft, stick-at-the-impact-site landing to access scientifically and commercially interesting sites, including difficult and hard-to-reach areas.

Sokolowski, Witold↗
NASA NTRS2003

Soft Landing of Spacecraft on Energy-Absorbing Self-Deployable Cushions

A report proposes the use of cold hibernated elastic memory (CHEM) foam structures to cushion impacts of small (1 to 50 kg) exploratory spacecraft on remote planets. Airbags, which are used on larger (800 to 1,000 kg) spacecraft have been found to (1) be too complex for smaller spacecraft; (2) provide insufficient thermal insulation between spacecraft and ground; (3) bounce on impact, thereby making it difficult to land spacecraft in precisely designated positions; and (4) be too unstable to serve as platforms for scientific observations. A CHEM foam pad according to the proposal would have a glass-transition temperature (Tg) well above ambient temperature. It would be compacted, at a temperature above Tg, to about a tenth or less of its original volume, then cooled below Tg, then installed on a spacecraft without compacting restraints. Upon entry of the spacecraft into a planetary atmosphere, the temperature would rise above Tg, causing the pad to expand to its original volume and shape. As the spacecraft decelerated and cooled, the temperature would fall below Tg, rigidifying the foam structure. The structure would absorb kinetic energy during ground impact by inelastic crushing, thus protecting the payload from damaging shocks. Thereafter, this pad would serve as a mechanically stable, thermally insulating platform for the landed spacecraft.

Sokolowski, Witold↗
NASA NTRS2012

Tropospheric Bromine Chemistry: Implications for Present and Pre-industrial Ozone and Mercury

We present a new model for the global tropospheric chemistry of inorganic bromine (Bry) coupled to oxidant-aerosol chemistry in the GEOS-Chem chemical transport model (CTM). Sources of tropospheric Bry include debromination of sea-salt aerosol, photolysis and oxidation of short-lived bromocarbons, and transport from the stratosphere. Comparison to a GOME-2 satellite climatology of tropospheric BrO columns shows that the model can reproduce the observed increase of BrO with latitude, the northern mid-latitudes maximum in winter, and the Arctic maximum in spring. This successful simulation is contingent on the HOBr + HBr reaction taking place in aqueous aerosols and ice clouds. Bromine chemistry in the model decreases tropospheric ozone mixing ratios by <1-8 nmol/mol (6.5% globally), with the largest effects in the northern extratropics in spring. The global mean tropospheric OH concentration decreases by 4 %. Inclusion of bromine chemistry improves the ability of global models (GEOS-Chem and p-TOMCAT) to simulate observed 19th-century ozone and its seasonality. Bromine effects on tropospheric ozone are comparable in the present-day and pre-industrial atmospheres so that estimates of anthropogenic radiative forcing are minimally affected. Br atom concentrations are 40% higher in the pre-industrial atmosphere due to lower ozone, which would decrease by a factor of 2 the atmospheric lifetime of elemental mercury against oxidation by Br. This suggests that historical anthropogenic mercury emissions may have mostly deposited to northern mid-latitudes, enriching the corresponding surface reservoirs. The persistent rise in background surface ozone at northern mid-latitudes during the past decades could possibly contribute to the observations of elevated mercury in subsurface waters of the North Atlantic.

bromine↗
NASA NTRS2012

Climatic Effects of 1950-2050 Changes in US Anthropogenic Aerosols: Aerosol Trends and Radiative Forcing - Part 1

We calculate decadal aerosol direct and indirect (warm cloud) radiative forcings from US anthropogenic sources over the 1950-2050 period. Past and future aerosol distributions are constructed using GEOS-Chem and historical emission inventories and future projections from the IPCC A1B scenario. Aerosol simulations are evaluated with observed spatial distributions and 1980-2010 trends of aerosol concentrations and wet deposition in the contiguous US. Direct and indirect radiative forcing is calculated using the GISS general circulation model and monthly mean aerosol distributions from GEOS-Chem. The radiative forcing from US anthropogenic aerosols is strongly localized over the eastern US. We find that its magnitude peaked in 1970-1990, with values over the eastern US (east of 100 deg W) of −2.0Wm(exp−2 for direct forcing including contributions from sulfate (−2.0Wm−2), nitrate (−0.2Wm(exp−2), organic carbon (−0.2Wm(exp−2), and black carbon (+0.4Wm(exp−2). The uncertainties in radiative forcing due to aerosol radiative properties are estimated to be about 50 %. The aerosol indirect effect is estimated to be of comparable magnitude to the direct forcing. We find that the magnitude of the forcing declined sharply from 1990 to 2010 (by 0.8Wm(exp−2) direct and 1.0Wm(exp−2 indirect), mainly reflecting decreases in SO2 emissions, and project that it will continue declining post-2010 but at a much slower rate since US SO2 emissions have already declined by almost 60% from their peak. This suggests that much of the warming effect of reducing US anthropogenic aerosol sources has already been realized. The small positive radiative forcing from US BC emissions (+0.3Wm(exp−2 over the eastern US in 2010; 5% of the global forcing from anthropogenic BC emissions worldwide) suggests that a US emission control strategy focused on BC would have only limited climate benefit.

climate change↗
NASA NTRS2014

HEMCO v1.0: A Versatile, ESMF-Compliant Component for Calculating Emissions in Atmospheric Models

We describe the Harvard-NASA Emission Component version 1.0 (HEMCO), a stand-alone software component for computing emissions in global atmospheric models. HEMCO determines emissions from different sources, regions, and species on a user-defined grid and can combine, overlay, and update a set of data inventories and scale factors, as specified by the user through the HEMCO configuration file. New emission inventories at any spatial and temporal resolution are readily added to HEMCO and can be accessed by the user without any preprocessing of the data files or modification of the source code. Emissions that depend on dynamic source types and local environmental variables such as wind speed or surface temperature are calculated in separate HEMCO extensions. HEMCO is fully compliant with the Earth System Modeling Framework (ESMF) environment. It is highly portable and can be deployed in a new model environment with only few adjustments at the top-level interface. So far, we have implemented HEMCO in the NASA Goddard Earth Observing System (GEOS-5) Earth system model (ESM) and in the GEOS-Chem chemical transport model (CTM). By providing a widely applicable framework for specifying constituent emissions, HEMCO is designed to ease sensitivity studies and model comparisons, as well as inverse modeling in which emissions are adjusted iteratively. The HEMCO code, extensions, and the full set of emissions data files used in GEOS-Chem are available at http: //wiki.geos-chem.org/HEMCO.

GEOS-5↗
NASA NTRS2014

Fifteen-Year Global Time Series of Satellite-Derived Fine Particulate Matter

Ambient fine particulate matter (PM2.5) is a leading environmental risk factor for premature mortality. We use aerosol optical depth (AOD) retrieved from two satellite instruments, MISR and SeaWiFS, to produce a unified 15-year global time series (1998−2012) of ground-level PM2.5 concentration at a resolution of 1deg x 1deg. The GEOS-Chem chemical transport model (CTM) is used to relate each individual AOD retrieval to ground-level PM2.5. Four broad areas showing significant, spatially coherent, annual trends are examined in detail: the Eastern U.S. (−0.39 +/- 0.10 micro-g/cu m/yr), the Arabian Peninsula (0.81 +/- 0.21 micro-g/cu m/yr), South Asia (0.93 +/- 0.22 micro-g/cu m/yr) and East Asia (0.79 +/- 0.27 micro-g/cu m/yr). Over the period of dense in situ observation (1999− 2012), the linear tendency for the Eastern U.S. (−0.37 +/- 0.13 micro-g/cu m/yr) agrees well with that from in situ measurements (−0.38 +/- 0.06 micro-g/cu m/yr). A GEOS-Chem simulation reveals that secondary inorganic aerosols largely explain the observed PM2.5 trend over the Eastern U.S., South Asia, and East Asia, while mineral dust largely explains the observed trend over the Arabian Peninsula.

aerosol optical depth↗
NASA NTRS2015

Overview of Global/Regional Models Used to Evaluate Tropospheric Ozone in North America

Ozone (O3) is an important greenhouse gas, toxic pollutant, and plays a major role in atmospheric chemistry. Tropospheric O3 which resides in the planetary boundary layer (PBL) is highly reactive and has a lifetime on the order of days, however, O3 in the free troposphere and stratosphere has a lifetime on the order of weeks or months. Modeling O3 mixing ratios at and above the surface is difficult due to the multiple formation/destruction processes and transport pathways that cause large spatio-temporal variability in O3 mixing ratios. This talk will summarize in detail the global/regional models that are commonly used to simulate/predict O3 mixing ratios in the United States. The major models which will be focused on are the: 1) Community Multi-scale Air Quality Model (CMAQ), 2) Comprehensive Air Quality Model with Extensions (CAMx), 3) Goddard Earth Observing System with Chemistry (GEOS-Chem), 4) Real Time Air Quality Modeling System (RAQMS), 5) Weather Research and Forecasting/Chemistry (WRF-Chem) model, National Center for Atmospheric Research (NCAR)'s Model for OZone And Related chemical Tracers (MOZART), and 7) Geophysical Fluid Dynamics Laboratory (GFDL) AM3 model. I will discuss the major modeling components which impact O3 mixing ratio calculations in each model and the similarities/differences between these models. This presentation is vital to the 2nd Annual Tropospheric Ozone Lidar Network (TOLNet) Conference as it will provide an overview of tools, which can be used in conjunction with TOLNet data, to evaluate the complex chemistry and transport pathways controlling tropospheric O3 mixing ratios.

Mixing Ratios↗
NASA NTRS2014

Investigating the Sensitivity of Surface-Level Nitrate Seasonality in Antarctica to Primary Sources Using a Global Model

Determining the sources of total nitrate (TNIT NO-3 + HNO3) reaching Antarctica is a long-standing challenge. Here we analyze the monthly sensitivity of surface-level TNIT in Antarctica to primary sources using a global 3-D chemical transport model, GEOS-Chem, and its adjoint. Modeled seasonal variation of TNIT concentrations shows good agreement with several measurement studies, given that the lack of post-depositional processing in the model leads to an expected underestimate of maximum values in November through January. Remote NOx sources have the greatest impact May-July, during when the model background concentrations are sensitive to NOx emissions from fossil fuel combustion, soil, and lightning originating from 25S to 65S. In this season, NOx is transported to Antarctica as TNIT, which is formed above continental source regions at an altitude of 5-11 km. In other seasons, more NOx is transported as a reservoir species (e.g., peroxyacetyl nitrate, PAN) through the free troposphere, transforming into TNIT within a cone of influence that extends to 35S and above 4 km altitude. Photolysis of PAN over Antarctica is the main driver of modeled NOx seasonality. Stratospheric production and loss of tracers are relatively unimportant in monthly sensitivities in GEOS-Chem, driving only a few percent of surface level variability of TNIT. A small peak concentration in August is captured by the model, although some measured values in August fall outside the range of simulated concentrations. Modifications to the model to represent sedimentation of polar stratospheric clouds (PSCs) lead to increased surface level August TNIT concentrations. However, this simple representation does not explicitly account for PSC particle deposition or disappearance of the tropopause in the middle of winter, and thus the influence of stratospheric nitrate sources estimated in this study is likely a lower bound.

stratosphere↗
NASA NTRS2014

Assessing the Ability of Instantaneous Aircraft and Sonde Measurements to Characterize Climatological Means and Long-Term Trends in Tropospheric Composition

Over four decades of measurements exist that sample the 3-D composition of reactive trace gases in the troposphere from approximately weekly ozone sondes, instrumentation on civil aircraft, and individual comprehensive aircraft field campaigns. An obstacle to using these data to evaluate coupled chemistry-climate models (CCMs)the models used to project future changes in atmospheric composition and climateis that exact space-time matching between model fields and observations cannot be done, as CCMs generate their own meteorology. Evaluation typically involves averaging over large spatiotemporal regions, which may not reflect a true average due to limited or biased sampling. This averaging approach generally loses information regarding specific processes. Here we aim to identify where discrete sampling may be indicative of long-term mean conditions, using the GEOS-Chem global chemical-transport model (CTM) driven by the MERRA reanalysis to reflect historical meteorology from 2003 to 2012 at 2o by 2.5o resolution. The model has been sampled at the time and location of every ozone sonde profile available from the Would Ozone and Ultraviolet Radiation Data Centre (WOUDC), along the flight tracks of the IAGOSMOZAICCARABIC civil aircraft campaigns, as well as those from over 20 individual field campaigns performed by NASA, NOAA, DOE, NSF, NERC (UK), and DLR (Germany) during the simulation period. Focusing on ozone, carbon monoxide and reactive nitrogen species, we assess where aggregates of the in situ data are representative of the decadal mean vertical, spatial and temporal distributions that would be appropriate for evaluating CCMs. Next, we identically sample a series of parallel sensitivity simulations in which individual emission sources (e.g., lightning, biogenic VOCs, wildfires, US anthropogenic) have been removed one by one, to assess where and when the aggregated observations may offer constraints on these processes within CCMs. Lastly, we show results of an additional 31-year simulation from 1980-2010 of GEOS-Chem driven by the MACCity emissions inventory and MERRA reanalysis at 4o by 5o. We sample the model at every WOUDC sonde and flight track from MOZAIC and NASA field campaigns to evaluate which aggregate observations are statistically reflective of long-term trends over the period.

Atmospheric composition↗
NASA NTRS2016

The Asian Tropopause Aerosol Layer: Balloon-Borne Measurements, Satellite Observations and Modeling Approaches

Satellite observations and numerical modeling studies have demonstrated that the Asian Summer Monsoon (ASM) can provide a conduit for gas-phase pollutants in south Asia to reach the lower stratosphere. Now, observations from the CALIPSO satellite have revealed the Asian Tropopause Aerosol Layer (ATAL), a summertime accumulation of aerosols associated with ASM anticyclone, in the upper troposphere and lower stratosphere (UTLS). The ATAL has potential implications for regional cloud properties, climate, and chemical processes in the UTLS. Here, we show in situ measurements from balloon-borne instrumentation, aircraft and satellite observations, combined with trajectory and chemical transport model (CTM) simulations to explore the origin, composition, physical and optical properties of aerosols in the ATAL. In particular, we show balloon-based observations from our BATAL-2015 field campaign to India and Saudi Arabia in summer 2015, including in situ backscatter measurements from COBALD instruments, and some of the first observations of size and volatility of aerosols in the ATAL layer using optical particle counters (OPCs). Back trajectory calculations initialized from CALIPSO observations point to deep convection over North India as a principal source of ATAL aerosols. Available aircraft observations suggest significant sulfur and carbonaceous contributions to the ATAL, which is supported by simulations using the GEOS-Chem CTM. Source elimination studies conducted with the GEOS-Chem indicate that 80-90% of ATAL aerosols originate from south Asian sources, in contrast with some earlier studies.

Fairlie, T. D.↗
NASA NTRS2016

Characterizing the Asian Tropopause Aerosol Layer (ATAL) Using Satellite Observations, Balloon Measurements and a Chemical Transport Model

Satellite observations and numerical modeling studies have demonstrated that the Asian Summer Monsoon (ASM) provide a conduit for gas-phase pollutants in south Asia to reach the lower stratosphere. Now, observations from the CALIPSO satellite have revealed the Asian Tropopause Aerosol Layer (ATAL), a summertime accumulation of aerosols in the upper troposphere and lower stratosphere (UTLS), associated with the ASM anticyclone. The ATAL has potential implications for regional cloud properties, climate, and chemical processes in the UTLS. Here, we show in situ measurements from balloon-borne instruments, aircraft, and satellite observations, together with trajectory and chemical transport model (CTM) simulations to explore the origin, composition, physical, and optical properties of aerosols in the ATAL. In particular, we show balloon-data from our BATAL-2015 field campaign to India and Saudi Arabia in summer 2015, which includes in situ backscatter measurements from COBALD instruments, and the first observations of size and volatility of aerosols in the ATAL layer using optical particle counters (OPCs). Back trajectory calculations initialized from CALIPSO observations point to deep convection over North India as a principal source of ATAL aerosols. Available aircraft observations suggest significant sulfur and carbonaceous components to the ATAL, which is supported by simulations using the GEOS-Chem CTM. Source elimination studies conducted with the GEOS-Chem indicate that ATAL aerosols originate primary from south Asian sources, in contrast with some earlier studies.

Fairlie, T. D.↗
NASA NTRS2016

Analysis of Summer-Time Ozone and Precursor Species in the Southeast United States

Ozone (O3) is a greenhouse gas and toxic pollutant which plays a major role in air quality and atmospheric chemistry. The understanding and ability to model the horizontal and vertical structure of O3 mixing ratios is difficult due to the complex formation/destruction processes and transport pathways that cause large variability of O3. The Environmental Protection Agency has National Ambient Air Quality Standards for O3 set at 75 ppb with future standards proposed to be as low as 65 ppb. These lower values emphasize the need to better understand/simulate the transport processes, emission sources, and chemical processes controlling precursor species (e.g., NOx, VOCs, and CO) which influence O3 mixing ratios. The uncertainty of these controlling variables is particularly large in the southeast United States (US) which is a region impacted by multiple different emission sources of precursor species (anthropogenic and biogenic) and transport processes resulting in complex spatio-temporal O3 patterns. During this work we will evaluate O3 and precursor species in the southeast US applying models, ground-based and airborne in situ data, and lidar observations. In the summer of 2013, the UAH O3 Differential Absorption Lidar (DIAL) (part of the Tropospheric Ozone Lidar Network (TOLNet)) measured vertical O3 profiles from the surface up to approximately 12 km. During this period, the lidar observed numerous periods of dynamic temporal and vertical O3 structures. In order to determine the sources/processes impacting these O3 mixing ratios we will apply the CTM GEOS-Chem (v9-02) at a 0.25 deg x 0.3125 deg resolution. Using in situ ground-based (e.g., SEARCH Network, CASTNET), airborne (e.g., NOAA WP-3D - SENEX 2013, DC-8 - SEAC4RS), and TOLNet lidar data we will first evaluate the model to determine the capability of GEOS-Chem to simulate the spatio-temporal variability of O3 in the southeast US. Secondly, we will perform model sensitivity studies in order to quantify which emission sources (e.g., anthropogenic, biogenic, lighting, wildfire) and transport processes (e.g., stratospheric, long-range, local scale) are contributing to these TOLNet-observed dynamic O3 patterns. Results from the evaluation of the model and the study of sources/processes impacting observed O3 mixing ratios will be presented.

Optical Radar↗
NASA NTRS2017

Evaluating a Priori Ozone Profile Information Used in TEMPO (Tropospheric Emissions: Monitoring of Pollution) Tropospheric Ozone Retrievals

A primary objective for TOLNet is the evaluation and validation of space-based tropospheric O3 retrievals from future systems such as the Tropospheric Emissions: Monitoring of Pollution (TEMPO) satellite. This study is designed to evaluate the tropopause-based O3 climatology (TB-Clim) dataset which will be used as the a priori profile information in TEMPO O3 retrievals. This study also evaluates model simulated O3 profiles, which could potentially serve as a priori O3 profile information in TEMPO retrievals, from near-real-time (NRT) data assimilation model products (NASA Global Modeling and Assimilation Office (GMAO) Goddard Earth Observing System (GEOS-5) Forward Processing (FP) and Modern-Era Retrospective analysis for Research and Applications version 2 (MERRA2)) and full chemical transport model (CTM), GEOS-Chem, simulations. The TB-Clim dataset and model products are evaluated with surface (0-2 km) and tropospheric (0-10 km) TOLNet observations to demonstrate the accuracy of the suggested a priori dataset and information which could potentially be used in TEMPO O3 algorithms. This study also presents the impact of individual a priori profile sources on the accuracy of theoretical TEMPO O3 retrievals in the troposphere and at the surface. Preliminary results indicate that while the TB-Clim climatological dataset can replicate seasonally-averaged tropospheric O3 profiles observed by TOLNet, model-simulated profiles from a full CTM (GEOS-Chem is used as a proxy for CTM O3 predictions) resulted in more accurate tropospheric and surface-level O3 retrievals from TEMPO when compared to hourly (diurnal cycle evaluation) and daily-averaged (daily variability evaluation) TOLNet observations. Furthermore, it was determined that when large daily-averaged surface O3 mixing ratios are observed (65 ppb), which are important for air quality purposes, TEMPO retrieval values at the surface display higher correlations and less bias when applying CTM a priori profile information compared to all other data products. The primary reason for this is that CTM predictions better capture the spatio-temporal variability of the vertical profiles of observed tropospheric O3 compared to the TB-Clim dataset and other NRT data assimilation models evaluated during this study.

Pollution↗
NASA NTRS2018

How Can TOLNet Help to Better Understand Tropospheric Ozone? A Satellite Perspective

Potential sources of a priori ozone (O3) profiles for use in Tropospheric Emissions: Monitoring of Pollution (TEMPO) satellite tropospheric O3 retrievals are evaluated with observations from multiple Tropospheric Ozone Lidar Network (TOLNet) systems in North America. An O3 profile climatology (tropopause-based O3 climatology (TB-Clim), currently proposed for use in the TEMPO O3 retrieval algorithm) derived from ozonesonde observations and O3 profiles from three separate models (operational Goddard Earth Observing System (GEOS-5) Forward Processing (FP) product, reanalysis product from Modern-Era Retrospective analysis for Research and Applications version 2 (MERRA2), and the GEOS-Chem chemical transport model (CTM)) were: 1) evaluated with TOLNet measurements on various temporal scales (seasonally, daily, hourly) and 2) implemented as a priori information in theoretical TEMPO tropospheric O3 retrievals in order to determine how each a priori impacts the accuracy of retrieved tropospheric (0-10 km) and lowermost tropospheric (LMT, 0-2 km) O3 columns. We found that all sources of a priori O3 profiles evaluated in this study generally reproduced the vertical structure of summer-averaged observations. However, larger differences between the a priori profiles and lidar observations were observed when evaluating inter-daily and diurnal variability of tropospheric O3. The TB-Clim O3 profile climatology was unable to replicate observed inter-daily and diurnal variability of O3 while model products, in particular GEOS-Chem simulations, displayed more skill in reproducing these features. Due to the ability of models, primarily the CTM used in this study, on average to capture the inter-daily and diurnal variability of tropospheric and LMT O3 columns, using a priori profiles from CTM simulations resulted in TEMPO retrievals with the best statistical comparison with lidar observations. Furthermore, important from an air quality perspective, when high LMT O3 values were observed, using CTM a priori profiles resulted in TEMPO LMT O3 retrievals with the least bias. The application of time-specific (non-climatological) hourly/daily model predictions as the a priori profile in TEMPO O3 retrievals will be best suited when applying this data to study air quality or event-based processes as the standard retrieval algorithm will still need to use a climatology product. Follow-on studies to this work are currently being conducted to investigate the application of different CTM-predicted O3 climatology products in the standard TEMPO retrieval algorithm. Finally, similar methods to those used in this study can be easily applied by TEMPO data users to recalculate tropospheric O3 profiles provided from the standard retrieval using a different source of a priori.

Satellite↗
NASA NTRS2018

Atmospheric Chemistry Modeling Using a Regression Forest Model

Atmospheric chemistry is central to many environmental issues such as air pollution, climate change, and stratospheric ozone loss. Chemistry Transport Models (CTM) are a central tool for understanding these issues, whether for research or for forecasting. These models split the atmosphere in a large number of grid-boxes and consider the emission of compounds into these boxes and their subsequent transport, deposition, and chemical processing. The chemistry is represented through a series of simultaneous ordinary differential equations, one for each compound. Given the difference in life-times between the chemical compounds (milli-seconds for O1D to years for CH4) these equations are numerically stiff and solving them consists of a significant fraction of the computational burden of a CTM. We have investigated a machine learning approach to solving the differential equations instead of solving them numerically. From an annual simulation of the GEOS-Chem model we have produced a training dataset consisting of the concentration of compounds before and after the differential equations are solved, together with some key physical parameters for every grid-box and time-step. From this dataset we have trained a machine learning algorithm (regression forest) to be able to predict the concentration of the compounds after the integration step based on the concentrations and physical state at the beginning of the time step. We have then included this algorithm back into the GEOS-Chem model, bypassing the need to integrate the chemistry. This machine learning approach shows many of the characteristics of the full simulation and has the potential to be substantially faster. There are a wide range of application for such an approach - generating boundary conditions, for use in air quality forecasts, chemical data assimilation systems, centennial scale climate simulations etc. We discuss our approches' speed and accuracy, and highlight some potential future directions for improving this approach.

Keller, Christoph A.↗
NASA NTRS2019

Atmospheric Chemistry Modeling Using Machine Learning

Atmospheric chemistry models are a central tool to study the impact of chemical constituents on the environment, vegetation and human health. These models split the atmosphere in a large number of grid-boxes and consider the emission of compounds into these boxes and their subsequent transport, deposition, and chemical processing. The chemistry is represented through a series of simultaneous ordinary differential equations, one for each compound. Given the difference in life-times between the chemical compounds (milli-seconds for O1D to years for CH4) these equations are numerically stiff and solving them consists of a significant fraction of the computational burden of a chemistry model. We have investigated a machine learning approach to emulate the chemistry instead of solving the differential equations numerically. From a one-month simulation of the GEOS-Chem model we have produced a training dataset consisting of the concentration of compounds before and after the differential equations are solved, together with some key physical parameters for every grid-box and time-step. From this dataset we have trained a machine learning algorithm (regression forest) to be able to predict the concentration of the compounds after the integration step based on the concentrations and physical state at the beginning of the time step. We have then included this algorithm back into the GEOS-Chem model, bypassing the need to integrate the chemistry. This machine learning approach shows many of the characteristics of the full simulation and has the potential to be substantially faster. There are a wide range of application for such an approach - generating boundary conditions, for use in air quality forecasts, chemical data assimilation systems, etc. We discuss speed and accuracy of our approach, and highlight some potential future directions for improving it.

Keller, Christoph A.↗
NASA NTRS2019

Machine Learning Application to Atmospheric Chemistry Modeling

Atmospheric chemistry models are a central tool to study the impact of chemical constituents on the environment, vegetation and human health. These models split the atmosphere in a large number of grid-boxes and consider the emission of compounds into these boxes and their subsequent transport, deposition, and chemical processing. The chemistry is represented through a series of simultaneous ordinary differential equations, one for each compound. Given the difference in life-times between the chemical compounds (milli-seconds for O (sup 1) D (Deuterium) to years for CH4) these equations are numerically stiff and solving them consists of a significant fraction of the computational burden of a chemistry model. We have investigated a machine learning approach to emulate the chemistry instead of solving the differential equations numerically. From a one-month simulation of the GEOS-Chem model we have produced a training dataset consisting of the concentration of compounds before and after the differential equations are solved, together with some key physical parameters for every grid-box and time-step. From this dataset we have trained a machine learning algorithm (regression forest) to be able to predict the concentration of the compounds after the integration step based on the concentrations and physical state at the beginning of the time step. We have then included this algorithm back into the GEOS-Chem model, bypassing the need to integrate the chemistry. This machine learning approach shows many of the characteristics of the full simulation and has the potential to be substantially faster. There are a wide range of application for such an approach - generating boundary conditions, for use in air quality forecasts, chemical data assimilation systems, etc. We discuss speed and accuracy of our approach, and highlight some potential future directions for improving it.

Keller, Christoph A.↗
NASA NTRS2019

Atmospheric Chemistry Modeling and Air Quality Forecasting Using Machine Learning

Atmospheric chemistry models are a central tool to study the impact of chemical constituents on the environment, vegetation and human health. These models split the atmosphere in a large number of grid-boxes and consider the emission of compounds into these boxes and their subsequent transport, deposition, and chemical processing. The chemistry is represented through a series of simultaneous ordinary differential equations, one for each compound. Given the difference in life-times between the chemical compounds (milli-seconds for O1D to years for CH4) these equations are numerically stiff and solving them consists of a significant fraction of the computational burden of a chemistry model.We have investigated a machine learning approach to emulate the chemistry instead of solving the differential equations numerically. From a one-month simulation of the GEOS-Chem model we have produced a training dataset consisting of the concentration of compounds before and after the differential equations are solved, together with some key physical parameters for every grid-box and time-step. From this dataset we have trained a machine learning algorithm (regression forest) to be able to predict the concentration of the compounds after the integration step based on the concentrations and physical state at the beginning of the time step. We have then included this algorithm back into the GEOS-Chem model, bypassing the need to integrate the chemistry.This machine learning approach shows many of the characteristics of the full simulation and has the potential to be substantially faster. There are a wide range of application for such an approach - generating boundary conditions, for use in air quality forecasts, chemical data assimilation systems, etc. We discuss speed and accuracy of our approach, and highlight some potential future directions for improving it.

Keller, Christoph A.↗