Using machine learning to build temperature-based ozone parameterizations for climate sensitivity simulations

Nowack, Peer; Braesicke, Peter; Haigh, Joanna D.; Abraham, N. L.; Pyle, J. A.; Voulgarakis, Apostolos

doi:10.1088/1748-9326/aae2be

Cited by 66 publications

(69 citation statements)

References 67 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The increment is then applied as a forcing to the meteorology at every time step during a second run of the same simulation period. This simulation includes full interactive tropospheric and stratospheric chemistry from the Goddard Modeling Initiative (GMI) chemical mechanism (Nielsen et al, 2017) with output for the years 1980-2018 at 0.625 • × 0.5 • horizontal res-olution and 72 vertical levels (Orbe et al, 2017;Stauffer et al, 2019;Wargan et al, 2018). This simulation is referred to as "GEOS Replay".…”

Section: Introductionmentioning

confidence: 99%

A machine learning examination of hydroxyl radical differences among model simulations for CCMI-1

et al. 2020

View full text Add to dashboard Cite

Abstract. The hydroxyl radical (OH) plays critical roles within the troposphere, such as determining the lifetime of methane (CH4), yet is challenging to model due to its fast cycling and dependence on a multitude of sources and sinks. As a result, the reasons for variations in OH and the resulting methane lifetime (τCH4), both between models and in time, are difficult to diagnose. We apply a neural network (NN) approach to address this issue within a group of models that participated in the Chemistry-Climate Model Initiative (CCMI). Analysis of the historical specified dynamics simulations performed for CCMI indicates that the primary drivers of τCH4 differences among 10 models are the flux of UV light to the troposphere (indicated by the photolysis frequency JO1D), the mixing ratio of tropospheric ozone (O3), the abundance of nitrogen oxides (NOx≡NO+NO2), and details of the various chemical mechanisms that drive OH. Water vapour, carbon monoxide (CO), the ratio of NO:NOx, and formaldehyde (HCHO) explain moderate differences in τCH4, while isoprene, methane, the photolysis frequency of NO2 by visible light (JNO2), overhead ozone column, and temperature account for little to no model variation in τCH4. We also apply the NNs to analysis of temporal trends in OH from 1980 to 2015. All models that participated in the specified dynamics historical simulation for CCMI demonstrate a decline in τCH4 during the analysed timeframe. The significant contributors to this trend, in order of importance, are tropospheric O3, JO1D, NOx, and H2O, with CO also causing substantial interannual variability in OH burden. Finally, the identified trends in τCH4 are compared to calculated trends in the tropospheric mean OH concentration from previous work, based on analysis of observations. The comparison reveals a robust result for the effect of rising water vapour on OH and τCH4, imparting an increasing and decreasing trend of about 0.5 % decade−1, respectively. The responses due to NOx, ozone column, and temperature are also in reasonably good agreement between the two studies.

show abstract

Section: Introductionmentioning

confidence: 99%

A machine learning examination of hydroxyl radical differences among model simulations for CCMI-1

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Machine learning approaches range from interpretable options such as the random forest algorithm (Breiman, 2001) to less interpretable ones such as artificial neural networks (Gardner and Dorling, 1998). On the more interpretable end, machine learning algorithms are being used increasingly within environmental sciences, with recent examples including linear ridge regression and random forest models to replace computationally expensive processes (Keller and Evans, 2019;Nowack et al, 2018) and T. Sherwen et al: A machine-learning-based global sea-surface iodide distribution 1241 Large and Yeager (2009) Expansion of Acronyms. WOA: World Ocean Atlas, SeaWIFS: Sea-Viewing Wide Field-of-View Sensor, GEBCO: General Bathymetric Chart of the Oceans, NOAMADS: NOAA National Operational Model Archive and Distribution System.…”

Section: Introductionmentioning

confidence: 99%

A machine-learning-based global sea-surface iodide distribution

Sherwen

Chance

Tinel

et al. 2019

Earth Syst. Sci. Data

View full text Add to dashboard Cite

Abstract. Iodide in the sea-surface plays an important role in the Earth system. It modulates the oxidising capacity of the troposphere and provides iodine to terrestrial ecosystems. However, our understanding of its distribution is limited due to a paucity of observations. Previous efforts to generate global distributions have generally fitted sea-surface iodide observations to relatively simple functions using proxies for iodide such as nitrate and sea-surface temperature. This approach fails to account for coastal influences and variation in the bio-geochemical environment. Here we use a machine learning regression approach (random forest regression) to generate a high-resolution (0.125∘×0.125∘, ∼12.5km×12.5km), monthly dataset of present-day global sea-surface iodide. We use a compilation of iodide observations (1967–2018) that has a 45 % larger sample size than has been used previously as the dependent variable and co-located ancillary parameters (temperature, nitrate, phosphate, salinity, shortwave radiation, topographic depth, mixed layer depth, and chlorophyll a) from global climatologies as the independent variables. We investigate the regression models generated using different combinations of ancillary parameters and select the 10 best-performing models to be included in an ensemble prediction. We then use this ensemble of models, combined with global fields of the ancillary parameters, to predict new high-resolution monthly global sea-surface iodide fields representing the present day. Sea-surface temperature is the most important variable in all 10 models. We estimate a global average sea-surface iodide concentration of 106 nM (with an uncertainty of ∼20 %), which is within the range of previous estimates (60–130 nM). Similar to previous work, higher concentrations are predicted for the tropics than for the extra-tropics. Unlike the previous parameterisations, higher concentrations are also predicted for shallow areas such as coastal regions and the South China Sea. Compared to previous work, the new parameterisation better captures observed variability. The iodide concentrations calculated here are significantly higher (40 % on a global basis) than the commonly used MacDonald et al. (2014) parameterisation, with implications for our understanding of iodine in the atmosphere. We envisage these fields could be used to represent present-day sea-surface iodide concentrations, in applications such as climate and air-quality modelling. The global iodide dataset is made freely available to the community (https://doi.org/10/gfv5v3, Sherwen et al., 2019), and as new observations are made, we will update the global dataset through a “living data” model.

show abstract

“…On the more interpretable end, machine learning algorithms are being used increasingly within environmental sciences, with recent examples including linear Ridge Regression and Random Forest models to replace computationally-expensive processes (Keller and Evans, 2018;Nowack et al, 2018) and Gaussian Process emulation to explore model biases on a global scale (Lee et al, 2011;Revell et al, 2018).…”

Section: Introductionmentioning

confidence: 99%

A machine learning based global sea-surface iodide distribution

Sherwen¹,

Chance²,

Tinel³

et al. 2019

Preprint

View full text Add to dashboard Cite

Abstract. Iodide in the sea-surface plays an important role in the Earth system. It modulates the oxidising capacity of the troposphere and provides iodine to terrestrial ecosystems. However, our understanding of its distribution is limited due to a paucity of observations. Previous efforts to generate global distributions have generally fitted sea-surface iodide observations to relatively simple functions of sea-surface temperature (Chance et al., 2014; MacDonald et al., 2014). This approach fails to account for coastal influences and variation in the bio-geochemical environment. Here we use a machine learning regression approach (Random Forest Regression) to generate a high resolution (0.125° x 0.125°, ∼ 12.5 km), monthly dataset of present-day global sea-surface iodide. We use a compilation of iodide observations (Chance et al., 2019b) that is 45 % larger than has been used previously (Chance et al., 2014) as the dependent variable and co-located ancillary parameters (temperature, nitrate, phosphate, salinity, shortwave radiation, topographic depth, mixed layer depth, and chlorophyll-a) from global climatologies as the independent variables. We investigate the regression models generated using different combinations of ancillary parameters and select the ten best-performing models to be included in an ensemble prediction. We then use this ensemble of models, combined with global fields of the ancillary parameters, to predict a new high resolution global sea-surface iodide field. Sea-surface temperature is the most important variable in all of the top ten models. We estimate a global average sea-surface iodide concentration of 106 nM (with an uncertainty of ∼ 20 %), which is within the range of previous estimates (60–130 nM). Similar to previous work, higher concentrations are predicted for the tropics than for the extra-tropics. Unlike the previous parameterisations, higher concentrations are also predicted for shallow areas such as coastal regions and the South China Sea. Compared to previous work, the new parameterisation better captures observed variability. The iodide concentrations calculated here are significantly higher (40 % on a global basis) than the commonly used MacDonald et al. (2014) parameterisation, with implications for our understanding of iodine in the atmosphere. The global iodide dataset is made freely available to the community (DOI: https://doi.org/10/gfv5v3) and as new observations are made, we will update the global dataset through a "living data" model.

show abstract

Using machine learning to build temperature-based ozone parameterizations for climate sensitivity simulations

Cited by 66 publications

References 67 publications

A machine learning examination of hydroxyl radical differences among model simulations for CCMI-1

A machine learning examination of hydroxyl radical differences among model simulations for CCMI-1

A machine-learning-based global sea-surface iodide distribution

A machine learning based global sea-surface iodide distribution

Contact Info

Product

Resources

About