Exploring the coupled ocean and atmosphere system with a data science approach applied to observations from the Antarctic Circumnavigation Expedition

The Southern Ocean is a critical component of Earth's climate system, but its remoteness makes it challenging to develop a holistic understanding of its processes from the small scale to the large scale. As a result, our knowledge of this vast region remains largely incomplete. The Antarctic Circumnavigation Expedition (ACE, austral summer 2016/2017) surveyed a large number of variables describing the state of the ocean and the atmosphere, the freshwater cycle, atmospheric chemistry, and ocean biogeochemistry and microbiology. This circumpolar cruise included visits to 12 remote islands, the marginal ice zone, and the Antarctic coast. Here, we use 111 of the observed variables to study the latitudinal gradients, seasonality, shorter-term variations, geographic setting of environmental processes, and interactions between them over the duration of 90 d. To reduce the dimensionality and complexity of the dataset and make the relations between variables interpretable we applied an unsupervised machine learning method, the sparse principal component analysis (sPCA), which describes environmental processes through 14 latent variables. To derive a robust statistical perspective on these processes and to estimate the uncertainty in the sPCA decomposition, we have developed a bootstrap approach. Our results provide a proof of concept that sPCA with uncertainty analysis is able to identify temporal patterns from diurnal to seasonal cycles, as well as geographical gradients and “hotspots” of interaction between environmental compartments. While confirming many well known processes, our analysis provides novel insights into the Southern Ocean water cycle (freshwater fluxes), trace gases (interplay between seasonality, sources, and sinks), and microbial communities (nutrient limitation and island mass effects at the largest scale ever reported). More specifically, we identify the important role of the oceanic circulations, frontal zones, and islands in shaping the nutrient availability that controls biological community composition and productivity; the fact that sea ice controls sea water salinity, dampens the wave field, and is associated with increased phytoplankton growth and net community productivity possibly due to iron fertilisation and reduced light limitation; and the clear regional patterns of aerosol characteristics that have emerged, stressing the role of the sea state, atmospheric chemical processing, and source processes near hotspots for the availability of cloud condensation nuclei and hence cloud formation. A set of key variables and their combinations, such as the difference between the air and sea surface temperature, atmospheric pressure, sea surface height, geostrophic currents, upper-ocean layer light intensity, surface wind speed and relative humidity played an important role in our analysis, highlighting the necessity for Earth system models to represent them adequately. In conclusion, our study highlights the use of sPCA to identify key ocean–atmosphere interactions across physical, chemical, and biological processes and their associated spatio-temporal scales. It thereby fills an important gap between simple correlation analyses and complex Earth system models. The sPCA processing code is available as open-access from the following link: https://renkulab.io/gitlab/ACE-ASAID/spca-decomposition (last access: 29 March 2021). As we show here, it can be used for an exploration of environmental data that is less prone to cognitive biases (and confirmation biases in particular) compared to traditional regression analysis that might be affected by the underlying research question.

Overview of the Antarctic Circumnavigation Expedition: Study of Preindustrial-like Aerosols and Their Climate Effects (ACE-SPACE)

Andrea Baccarini, Sebastian Johannes Heinz-Josef Landwehr, Robin Lewis Modini, Julia Schmale

Uncertainty in radiative forcing caused by aerosol–cloud interactions is about twice as large as for CO2 and remains the least well understood anthropogenic contribution to climate change. A major cause of uncertainty is the poorly quantified state of aerosols in the pristine preindustrial atmosphere, which defines the baseline against which anthropogenic effects are calculated. The Southern Ocean is one of the few remaining near-pristine aerosol environments on Earth, but there are very few measurements to help evaluate models. The Antarctic Circumnavigation Expedition: Study of Preindustrial-like Aerosols and their Climate Effects (ACE-SPACE) took place between December 2016 and March 2017 and covered the entire Southern Ocean region (Indian, Pacific, and Atlantic Oceans; length of ship track >33,000 km) including previously unexplored areas. In situ measurements covered aerosol characteristics [e.g., chemical composition, size distributions, and cloud condensation nuclei (CCN) number concentrations], trace gases, and meteorological variables. Remote sensing observations of cloud properties, the physical and microbial ocean state, and back trajectory analyses are used to interpret the in situ data. The contribution of sea spray to CCN in the westerly wind belt can be larger than 50%. The abundance of methanesulfonic acid indicates local and regional microbial influence on CCN abundance in Antarctic coastal waters and in the open ocean. We use the in situ data to evaluate simulated CCN concentrations from a global aerosol model. The extensive, available ACE-SPACE dataset (https://zenodo.org/communities/spi-ace?page=1&size=20) provides an unprecedented opportunity to evaluate models and to reduce the uncertainty in radiative forcing associated with the natural processes of aerosol emission, formation, transport, and processing occurring over the pristine Southern Ocean.

2019

Dry and ambient aerosol properties at the Jungfraujoch

Remo Nessler

Atmospheric aerosol is defined as a suspension of solid or liquid particles in air. The major concerns about aerosol particles are their adverse health effects and their role in the Earth's climate system. Atmospheric particles influence the Earth's radiation budget in two ways: either directly by scattering and absorbing incoming solar radiation (and to a far less extent longwave radiation emitted from the Earth) or indirectly by their ability to act as cloud condensation nuclei (CCN). With increasing particle concentrations, caused by, e.g., anthropogenic emissions, more CCN are available to form cloud droplets. For a fixed water content in the atmosphere this leads to more but smaller droplets and therefore to enhanced cloud albedos as well as to rain suppression. Both effects have been recognized in recent years to be of great importance for understanding the observed climate change. Nevertheless, they remain poorly understood and quantified. The Jungfraujoch (JFJ, 3580 m asl; 46.548°N, 7.984°E) High Alpine Research Station is equipped with instruments that have for decades performed important Global Climate Observations. A major part of the aerosol instruments perform in-situ measurements. However, due to the harsh weather conditions, they usually cannot be run outdoors. The ambient air has to be inducted into a housing, such that the aerosol is sampled at a temperature (T) and relative humidity (RH) different from the ambient values. Therefore, the measured aerosol properties may considerably differ from the ambient — the climate-relevant — ones. During the first Cloud and Aerosol Characterization Experiment (CLACE) at the JFJ in-situ aerosol size distributions were measured simultaneously indoor at dry conditions (T ≈ 25 °C and RH < 10%) and, for the first time in such an exposed place, outdoor at ambient conditions (T < –5 °C) by means of two scanning mobility particle sizers (SMPS). The data set was completed by measurements of hygroscopic growth factors using a hygroscopicity tandem differential mobility analyzer (H-TDMA). A comparison of dry and ambient size distributions shows two main features: First, the dry total number concentration is often considerably smaller (on average 28%) than the ambient total number concentration. This is most likely due to the evaporation of volatile material at the higher indoor temperature. These particle losses mainly concern small particles (dry diameter Ddry ≤ 100 nm), and therefore have only a minimal affect on the surface and volume concentrations. Second, the dry number size distribution is shifted towards smaller particles, reflecting the hygroscopic behavior of the aerosols. This shift was modeled using a modified Köhler equation adapted to the measured hygroscopic growth factors. The corrected dry surface and volume concentrations are in good agreement with the ambient measurements for the whole RH range, but the correction works best for RH < 80%. The results indicate that size distribution data measured at indoor conditions (i.e. dry and warm) may be successfully corrected to reflect ambient conditions. By combining these results with earlier JFJ aerosol measurements (e.g. chemical composition), it was possible to model dry and ambient optical aerosol properties for a summer and a winter case. The model calculations were based on a total of 496 size distributions representative for the JFJ aerosol, hygroscopic growth parameterizations and the following simplified aerosol structures: The fine mode (Ddry ≤ 1 µm) aerosol was assumed to consist of a water-insoluble spherical core concentrically encased by a coating comprising all water-soluble compounds, and the coarse mode (Ddry > 1 µm) particles were assumed to be water-insoluble homogeneous spheres. Ambient scattering coefficients σsca(RH) were found to be considerably different from the dry scattering coefficients σsca(RH = 0). The scattering enhancement factors ξ(RH) = σsca(RH)/σsca(RH = 0) strongly depend on the particle size. At RH = 85% they vary for example between ≈ 1.2 and ≈ 3.8. It was possible to establish a parameterization of ξ(RH) with the dry Ångström exponent å (based on scattering only). Since å follows directly from the dry scattering measurements, the parameterization allows to derive ambient scattering coefficients from the dry scattering coefficients without the need of any additional measurement other than ambient RH. RH also affects the absorption coefficient σabs. Depending on particle size and wavelength (investigated between 370 nm and 950 nm) absorption enhancement factors χ(RH) = σabs(RH)/σabs(RH = 0) were found to range from 0.84 to 1.78. However, it was possible to demonstrate that, even though the humidity effect on absorption is substantial, its contribution to the humidity effect on extinction and the single scattering albedo is negligible at the JFJ. The EPFL Raman lidar installed at the JFJ provides profiles of extinction and backscattering coefficients at the wavelengths 355 nm, 532 nm and 1064 nm for altitudes between typically 4 km and 15 km asl. These measurements are ambient but afflicted with more uncertainties than the dry in-situ measurements. A method was developed to derive microphysical aerosol properties from optical lidar data. This is a so-called ill-posed problem, i.e., small errors in the lidar data lead to huge errors in the inverted microphysical properties, unless special mathematical tools are applied. Therefore the method makes use of regularization techniques. It is shown that microphysical parameters can be inverted with acceptable accuracy from an optical dataset as it is provided by a lidar of e.g. the EPFL type.

EPFL2004

Incorporating an advanced aerosol activation parameterization into WRF-CAM5: Model evaluation and parameterization intercomparison

Athanasios Nenes, Xiuwei Zhang, Yi Zhang

Aerosol activation into cloud droplets is an important process that governs aerosol indirect effects. The advanced treatment of aerosol activation by Fountoukis and Nenes (2005) and its recent updates, collectively called the FN series, have been incorporated into a newly developed regional coupled climate-air quality model based on the Weather Research and Forecasting model with the physics package of the Community Atmosphere Model version 5 (WRF-CAM5) to simulate aerosol-cloud interactions in both resolved and convective clouds. The model is applied to East Asia for two full years of 2005 and 2010. A comprehensive model evaluation is performed for model predictions of meteorological, radiative, and cloud variables, chemical concentrations, and column mass abundances against satellite data and surface observations from air quality monitoring sites across East Asia. The model performs overall well for major meteorological variables including near-surface temperature, specific humidity, wind speed, precipitation, cloud fraction, precipitable water, downward shortwave and longwave radiation, and column mass abundances of CO, SO2, NO2, HCHO, and O3 in terms of both magnitudes and spatial distributions. Larger biases exist in the predictions of surface concentrations of CO and NOx at all sites and SO2, O3, PM2.5, and PM10 concentrations at some sites, aerosol optical depth, cloud condensation nuclei over ocean, cloud droplet number concentration (CDNC), cloud liquid and ice water path, and cloud optical thickness. Compared with the default Abdul-Razzack Ghan (2000) parameterization, simulations with the FN series produce ~107-113% higher CDNC, with half of the difference attributable to the higher aerosol activation fraction by the FN series and the remaining half due to feedbacks in subsequent cloud microphysical processes. With the higher CDNC, the FN series are more skillful in simulating cloud water path, cloud optical thickness, downward shortwave radiation, shortwave cloud forcing, and precipitation. The model evaluation identifies several areas of improvements including emissions and their vertical allocation as well as model formulations such as aerosol formation, cloud droplet nucleation, and ice nucleation. © 2015. American Geophysical Union. All Rights Reserved.

Wiley-Blackwell2015