scholarly journals Exploring the potential of machine learning for simulations of urban ozone variability

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Narendra Ojha ◽  
Imran Girach ◽  
Kiran Sharma ◽  
Amit Sharma ◽  
Narendra Singh ◽  
...  
Keyword(s):  
Machine Learning ◽  
Atmospheric Chemistry ◽  
Model Performance ◽  
Meteorological Conditions ◽  
Earth System ◽  
The Earth ◽  
Model Complex ◽  
Quality Assessments ◽  
The Himalaya ◽  
Urban Ozone

AbstractMachine learning (ML) has emerged as a powerful technique in the Earth system science, nevertheless, its potential to model complex atmospheric chemistry remains largely unexplored. Here, we applied ML to simulate the variability in urban ozone (O3) over Doon valley of the Himalaya. The ML model, trained with past variations in O3 and meteorological conditions, successfully reproduced the independent O3 data (r2 ~ 0.7). Model performance is found to be similar when the variation in major precursors (CO and NOx) were included in the model, instead of the meteorology. Further the inclusion of both precursors and meteorology improved the performance significantly (r2 = 0.86) and the model could also capture the outliers, which are crucial for air quality assessments. We suggest that in absence of high-resolution measurements, ML modeling has profound implications for unraveling the feedback between pollution and meteorology in the fragile Himalayan ecosystem.

Learning ordinary differential equations from remote sensing data

2020 ◽  
Author(s):  
Jose E. Adsuara ◽  
Adrián Pérez-Suay ◽  
Alvaro Moreno-Martínez ◽  
Anna Mateo-Sanchis ◽  
Maria Piles ◽  
...  
Keyword(s):  
Machine Learning ◽  
Remote Sensing ◽  
Differential Equations ◽  
Ordinary Differential Equations ◽  
Remote Sensing Data ◽  
Earth System ◽  
Sensing Data ◽  
Learning From Data ◽  
The Earth ◽  
Parameters Of Interest

<p>Modeling and understanding the Earth system is of paramount relevance. Modeling the complex interactions among variables in both space and time is a constant and challenging endevour. When a clear mechanistic model of variable interaction and evolution is not available or uncertain, learning from data can be an alternative. </p><p>Currently, Earth observation (EO) remote sensing data provides almost continuous space and time sampling of the Earth system which has been used to monitor our planet with advanced, semiautomatic algorithms able to classify and detect changes, and to retrieve relevant biogeophysical parameters of interest. Despite great advances in classification and regression, learning from data seems an ilusive problem in machine learning for the Earth sciences. The hardest part turns out to be the extraction of their relevant information and figuring out reliable models for summarizing, modeling, and understanding variables and parameters of interest.</p><p> </p><p>We introduce the use of machine learning techniques to bring systems of ordinary differential equations (ODEs) to light purely from data. Learning ODEs from stochastic variables is a challenging problem, and hence studied scarcely in the literature. Sparse regression algorithms allow us to explore the space of solutions of ODEs from data. Owing to the Occam's razor, and exploiting extra physics-aware regularization, the presented method identifies the most expressive and simplest ODEs explaining the data. From the learned ODE, one not only learns the underlying dynamical equation governing the system, but standard analysis allows us to also infer collapse, turning points, and stability regions of the system. We illustrate the methodology using some particular remote sensing datasets quantifying biosphere and vegetation status. These analytical equations come to be self-explanatory models which may provide insight into these particular Earth Subsystems.</p>

Towards fast machine learning parameterizations of stratospheric ozone feedbacks in climate change simulations

2020 ◽  
Author(s):  
Peer Nowack ◽  
Nathan Luke Abraham ◽  
Peter Braesicke
Keyword(s):  
Climate Change ◽  
Machine Learning ◽  
Global Warming ◽  
Climate Sensitivity ◽  
Atmospheric Chemistry ◽  
Stratospheric Ozone ◽  
Environmental Research ◽  
Earth System ◽  
Geophysical Research ◽  
The Impact

<p>There is a plethora of ways in which the representation of upper tropospheric and stratospheric ozone (‘ozone feedbacks’) can affect the outcome of climate change simulations. Prominent examples include modulations of the tropospheric and stratospheric circulation, climate sensitivity, cloud formation, and stratospheric water vapour (e.g. [1-8]). Here I first revisit some recent work providing evidence for such effects. I then provide an update on a recently developed machine learning parameterization for ozone using the UK Earth System Model (UKESM1, [9]). Such a parameterization could adequately represent ozone feedbacks without adding the high computational expense of a fully interactive atmospheric chemistry scheme. The parameterization could also provide several notable scientific advantages, for example concerning the treatment of important chemistry-climate model biases. Finally, I put my results into the context of several other methods suggested as potential means for addressing ozone-related effects in idealized climate sensitivity simulations, also considering the still substantial uncertainties related to modelling ozone [10,11] and associated climate feedbacks [5,12].</p><p>References:</p><p>[1] Son et al. (2008), The impact of stratospheric ozone recovery on the Southern Hemisphere westerly jet. Science 320, 1486, doi:10.1126/science.1155939.</p><p>[2] Dietmüller et al. (2014), Interactive ozone induces a negative feedback in CO2-driven climate change simulations, Journal of Geophysical Research: Atmospheres 119, 1796-1805, doi:10.1002/2013JD020575.</p><p>[3] Chiodo & Polvani (2016), Reduction of climate sensitivity to solar forcing due to stratospheric ozone feedback, Journal of Climate 29, 4651-4663, doi:10.1175/JCLI-D-15-0721.1.</p><p>[4] Chiodo & Polvani (2017), Reduced Southern Hemispheric circulation response to quadrupled CO2 due to stratospheric ozone feedback, Geophysical Research Letters 43, 465-474, doi:10.1002/2016GL071011.</p><p>[5] Nowack et al. (2015), A large ozone-circulation feedback and its implications for global warming assessments. Nature Climate Change 5, 41-45, doi:10.1038/nclimate2451.</p><p>[6] Nowack et al. (2017), On the role of ozone feedback in the ENSO amplitude response under global warming, Geophysical Research Letters 44, doi:10.1002/2016GL072418.</p><p>[7] Nowack et al. (2018), The impact of stratospheric ozone feedbacks on climate sensitivity estimates. Journal of Geophysical Research: Atmospheres 123, 4630-4641, doi:10.1002/2017JD027943.</p><p>[8] Rieder et al. (2019), Is interactive ozone chemistry important to represent polar cap stratospheric temperature variability in Earth-System Models?, Environmental Research Letters 14, 044026, doi: 10.1088/1748-9326/ab07ff.</p><p>[9] Nowack et al. (2018), Using machine learning to build temperature-based ozone parameterizations for climate sensitivity simulations, Environmental Research Letters 13, 104016, doi:10.1088/1748-9326/aae2be.</p><p>[10] Chiodo & Polvani (2019), The response of the ozone layer to quadrupled CO2 concentrations: implications for climate, Journal of Climate 31, 3893-3907, doi:10.1175/JCLI-D-17-0492.1.</p><p>[11] Keeble et al. (2020), Evaluating stratospheric ozone and water vapour changes in CMIP6 models from 1850-2100, Atmospheric Chemistry and Physics Discussions.</p><p>[12] Dacie et al. (2019), A 1D RCE study of factors affecting the tropical tropopause layer and surface climate. Journal of Climate 32, 6769-6782, doi:10.1175/JCLI-D-18-0778.1.</p>

Automating the pre-processing of time-domain induced polarization data using machine learning

2020 ◽  
Author(s):  
Adrian S. Barfod* ◽  
Jakob Juul Larsen
Keyword(s):  
Machine Learning ◽  
Time Domain ◽  
Induced Polarization ◽  
Geophysical Data ◽  
Neural Net ◽  
Earth System ◽  
Polarization Data ◽  
The Earth ◽  
Time Modeling ◽  
Wide Range

<p>Exploring and studying the earth system is becoming increasingly important as the slow depletion of natural resources ensues. An important data source is geophysical data, collected worldwide. After gathering data, it goes through vigorous quality control, pre-processing, and inverse modelling procedures. Such procedures often have manual components, and require a trained geophysicist who understands the data, in order to translate it into useful information regarding the earth system. The sheer amounts of geophysical data collected today makes manual approaches impractical. Therefore, automating as much of the workflow related to geophysical data as possible, would allow novel opportunities such as fully automated geophysical monitoring systems, real-time modeling during data collection, larger geophysical data sets, etc.</p><p>Machine learning has been proposed as a tool for automating workflows related to geophysical data. The field of machine learning encompasses multiple tools, which can be applied in a wide range of geophysical workflows, such as pre-processing, inverse modeling, data exploration etc.</p><p>We present a study where machine learning is applied to automate the time domain induced polarization geophysical workflow. Such induced polarization data requires pre-processing, which is manual in nature. One of the pre-processing steps is that a trained geophysicist inspects the data, and removes so-called non-geologic signals, i.e. noise, which does not represent geological variance. Specifically, a real-world case from Grindsted Denmark is presented. Here, a time domain induced polarization survey was conducted containing seven profiles. Two lines were manually processed and used for supervised training of an artificial neural network. The neural net then automatically processed the remaining profiles of the survey, with satisfactory results. Afterwards, the processed data was inverted, yielding the induced polarization parameters respective to the Cole-Cole model. We discuss the limitations and optimization steps related to training such a classification network.</p>

scholarly journals Lifecycle of light-absorbing carbonaceous aerosols in the atmosphere

2020 ◽  
Vol 3 (1) ◽  
Author(s):  
Dantong Liu ◽  
Cenlin He ◽  
Joshua P. Schwarz ◽  
Xuan Wang
Keyword(s):  
Optical Properties ◽  
Radiative Forcing ◽  
Water Solubility ◽  
Model Performance ◽  
Global Scale ◽  
Climate Impacts ◽  
Carbonaceous Aerosols ◽  
Earth System ◽  
The Earth ◽  
Improved Model

Abstract Light-absorbing carbonaceous aerosols (LACs), including black carbon and light-absorbing organic carbon (brown carbon, BrC), have an important role in the Earth system via heating the atmosphere, dimming the surface, modifying the dynamics, reducing snow/ice albedo, and exerting positive radiative forcing. The lifecycle of LACs, from emission to atmospheric evolution further to deposition, is key to their overall climate impacts and uncertainties in determining their hygroscopic and optical properties, atmospheric burden, interactions with clouds, and deposition on the snowpack. At present, direct observations constraining some key processes during the lifecycle of LACs (e.g., interactions between LACs and hydrometeors) are rather limited. Large inconsistencies between directly measured LAC properties and those used for model evaluations also exist. Modern models are starting to incorporate detailed aerosol microphysics to evaluate transformation rates of water solubility, chemical composition, optical properties, and phases of LACs, which have shown improved model performance. However, process-level understanding and modeling are still poor particularly for BrC, and yet to be sufficiently assessed due to lack of global-scale direct measurements. Appropriate treatments of size- and composition-resolved processes that influence both LAC microphysics and aerosol–cloud interactions are expected to advance the quantification of aerosol light absorption and climate impacts in the Earth system. This review summarizes recent advances and up-to-date knowledge on key processes during the lifecycle of LACs, highlighting the essential issues where measurements and modeling need improvement.

scholarly journals Biological regulation of atmospheric chemistry en route to planetary oxygenation

2017 ◽  
Vol 114 (13) ◽  
pp. E2571-E2579 ◽  
Author(s):  
Gareth Izon ◽  
Aubrey L. Zerkle ◽  
Kenneth H. Williford ◽  
James Farquhar ◽  
Simon W. Poulton ◽  
...  
Keyword(s):  
Organic Carbon ◽  
Atmospheric Chemistry ◽  
Atmospheric Oxygen ◽  
Earth System ◽  
Atmospheric Methane ◽  
Biological Regulation ◽  
Biological Source ◽  
The Earth ◽  
Methane Fluxes ◽  
Haze Formation

Emerging evidence suggests that atmospheric oxygen may have varied before rising irreversibly ∼2.4 billion years ago, during the Great Oxidation Event (GOE). Significantly, however, pre-GOE atmospheric aberrations toward more reducing conditions—featuring a methane-derived organic-haze—have recently been suggested, yet their occurrence, causes, and significance remain underexplored. To examine the role of haze formation in Earth’s history, we targeted an episode of inferred haze development. Our redox-controlled (Fe-speciation) carbon- and sulfur-isotope record reveals sustained systematic stratigraphic covariance, precluding nonatmospheric explanations. Photochemical models corroborate this inference, showing Δ36S/Δ33S ratios are sensitive to the presence of haze. Exploiting existing age constraints, we estimate that organic haze developed rapidly, stabilizing within ∼0.3 ± 0.1 million years (Myr), and persisted for upward of ∼1.4 ± 0.4 Myr. Given these temporal constraints, and the elevated atmospheric CO2 concentrations in the Archean, the sustained methane fluxes necessary for haze formation can only be reconciled with a biological source. Correlative δ13COrg and total organic carbon measurements support the interpretation that atmospheric haze was a transient response of the biosphere to increased nutrient availability, with methane fluxes controlled by the relative availability of organic carbon and sulfate. Elevated atmospheric methane concentrations during haze episodes would have expedited planetary hydrogen loss, with a single episode of haze development providing up to 2.6–18 × 1018 moles of O2 equivalents to the Earth system. Our findings suggest the Neoarchean likely represented a unique state of the Earth system where haze development played a pivotal role in planetary oxidation, hastening the contingent biological innovations that followed.

scholarly journals A machine learning-guided adaptive algorithm to reduce the computational cost of atmospheric chemistry in Earth System models: application to GEOS-Chem versions 12.0.0 and 12.9.1

2021 ◽  
Author(s):  
Lu Shen ◽  
Daniel J. Jacob ◽  
Mauricio Santillana ◽  
Kelvin Bates ◽  
Jiawei Zhuang ◽  
...  
Keyword(s):  
Machine Learning ◽  
Atmospheric Chemistry ◽  
Computational Cost ◽  
Coupled System ◽  
Atmospheric Composition ◽  
Earth System ◽  
Major Barrier ◽  
System Models ◽  
Computational Bottleneck ◽  
Earth System Models

Abstract. Atmospheric composition plays a crucial role in determining the evolution of the atmosphere, but the high computational cost has been the major barrier to include atmospheric chemistry into Earth system models. Here we present an adaptive and efficient algorithm that can remove this barrier. Our approach is inspired by unsupervised machine learning clustering techniques and traditional asymptotic analysis ideas. We first partition species into 13 blocks, using a novel machine learning approach that analyzes the species network structures and their production and loss rates. Building on these blocks, we pre-select 20 submechanisms, as defined by unique assemblages of the species blocks, and then pick locally on the fly which submechanism to use based on local chemical conditions. In each submechanism, we isolate slow species and unimportant reactions from the coupled system. Application to a global 3-D model shows that we can cut the computational costs of the chemical integration by 50 % with accuracy losses smaller than 1 % that do not propagate in time. Tests show that this algorithm is highly chemically coherent making it easily portable to new models without compromising its performance. Our algorithm will significantly ease the computational bottleneck and will facilitate the development of next generation of earth system models.

scholarly journals Representativeness of single lidar stations for zonally averaged ozone profiles, their trends and attribution to proxies

2018 ◽  
Vol 18 (9) ◽  
pp. 6427-6440 ◽  
Author(s):  
Christos Zerefos ◽  
John Kapsomenakis ◽  
Kostas Eleftheratos ◽  
Kleareti Tourpali ◽  
Irina Petropavlovskikh ◽  
...  
Keyword(s):  
Atmospheric Chemistry ◽  
Southern Oscillation ◽  
Correlation Coefficients ◽  
The Arctic ◽  
Earth System ◽  
Quasi Biennial Oscillation ◽  
The Earth ◽  
Ozone Trends ◽  
Anthropogenic Contributions ◽  
The Arctic Oscillation

Abstract. This paper is focusing on the representativeness of single lidar stations for zonally averaged ozone profile variations over the middle and upper stratosphere. From the lower to the upper stratosphere, ozone profiles from single or grouped lidar stations correlate well with zonal means calculated from the Solar Backscatter Ultraviolet Radiometer (SBUV) satellite overpasses. The best representativeness with significant correlation coefficients is found within ±15∘ of latitude circles north or south of any lidar station. This paper also includes a multivariate linear regression (MLR) analysis on the relative importance of proxy time series for explaining variations in the vertical ozone profiles. Studied proxies represent variability due to influences outside of the earth system (solar cycle) and within the earth system, i.e. dynamic processes (the Quasi Biennial Oscillation, QBO; the Arctic Oscillation, AO; the Antarctic Oscillation, AAO; the El Niño Southern Oscillation, ENSO), those due to volcanic aerosol (aerosol optical depth, AOD), tropopause height changes (including global warming) and those influences due to anthropogenic contributions to atmospheric chemistry (equivalent effective stratospheric chlorine, EESC). Ozone trends are estimated, with and without removal of proxies, from the total available 1980 to 2015 SBUV record. Except for the chemistry related proxy (EESC) and its orthogonal function, the removal of the other proxies does not alter the significance of the estimated long-term trends. At heights above 15 hPa an “inflection point” between 1997 and 1999 marks the end of significant negative ozone trends, followed by a recent period between 1998 and 2015 with positive ozone trends. At heights between 15 and 40 hPa the pre-1998 negative ozone trends tend to become less significant as we move towards 2015, below which the lower stratosphere ozone decline continues in agreement with findings of recent literature.

Sign in / Sign up

Export Citation Format

Share Document