Skip to content

At The Bleeding Edge Of Climate Modeling: Aerosols In NorESM

  • INES 
Ingvild Aukland photo
Ingvild Aukland

The largest uncertainties in the radiative forcing of climate are caused by aerosol and aerosol-cloud interactions1. As part of the infrastructure efforts in INES2-interim, Ingvild Sofie Sundby Aukan is collaborating with several researchers at MET Norway, on identifying and constraining uncertainties related to the various methods of representing aerosols in climate models. We asked her some questions about her work.

What are aerosols?
Aerosol is the term we use for tiny particles, liquid or solid, that are suspended in air. Aerosols can be of both natural and anthropogenic origin, and their properties may vary depending on their constituents, and where in the atmosphere we find them.

Why do we care?
In climate research, aerosols are important because of how they interfere with solar radiation directly through scattering and absorption, and because they are essential for the formation of clouds, which in turn impact the Earth’s energy balance. Through a process called activation, some aerosols can become cloud condensation nuclei, upon which water vapor can condense to form cloud droplets. Aerosols can also serve as ice nucleating particles, by initiating the formation of ice crystals from supercooled water droplets or water vapor in the atmosphere.
Although there are a range of ways aerosols impact climate, the overall net effect is cooling, and anthropogenic emissions have been shown to mask a portion of the warming from greenhouse gasses in the industrial era. However, there are relatively large uncertainties related to climate model simulations of aerosol impact on climate, partly due to differences in how aerosols are represented in climate models.

Figure 1. Artist’s rendition of aerosol sources and aerosol–cloud-radiation interactions. Figure by Ingvild Aukan.

How are aerosols modeled in climate models?
Aerosols exist in sizes that range from a few nanometers to several micrometers, and aerosol processes occur on different timescales. Additionally, there is a range of chemical and radiative properties to consider. The complexity of sizes, time scales, chemical and radiative properties make it challenging to include aerosols in climate models. The focus on allocating computational resources and deciding the level of complexity in representing aerosols depends on the model’s purpose and its intended use. Aerosols can be included in climate models using three common methods. First, a bulk scheme is a simplified way of incorporating aerosols, by simulating only the mass of constituents and prescribing all other properties. However, a bulk scheme lacks aerosol microphysics, and consequently, it does not alter the size distribution of constituents as they evolve after emission. Second, in a sectional scheme, the aerosol size distribution is represented by discrete size bins, and no a priori assumptions about the distribution is needed. However, this approach often entails high computational costs. A third approach is to use a modal scheme to represent the size distribution by a sum of independent modes, usually defined by lognormal functions. Such modal schemes are generally favored in climate models because they require fewer tracers and are computationally much cheaper.

How do we model aerosols in NorESM?
OsloAero, the aerosol module in NorESM, is largely a modal scheme, but with elements of the sectional scheme. This method involves lognormal “background” modes representing the aerosol size distribution that can change in size due to processes like condensation, coagulation and aerosol-cloud interactions. In addition to the background modes, we have something we call “process” tracers, that are compounds that change the shape and chemical composition of the background modes. A combination of a background mode and a process tracer is referred to as a mixture.

During transport, background modes and process tracers are handled independently. The mass of the process tracers is after transport distributed onto the background modes.
To combine the information about the mixtures, we use AeroTab, an offline component of the aerosol module, which provides information about aerosol properties, such as optical parameters needed for the radiative balance calculation in the atmospheric model. AeroTab calculates these properties by combining the mass from background modes and the mass from the process tracers in a sectional scheme. With this method, the size distribution of a mixture and the fraction of aerosol compounds within it can change for specific sizes, rather than being uniformly distributed across the background mode. In addition, we have a version of OsloAero, called OsloAeroSec, which handles the transition from very small particles created by nucleation from the gas phase, by using a sectional scheme, until the particles grow large enough to enter the smallest background mode.

To become cloud condensation nuclei (CCN) or ice nucleating particles (INPs) the mixtures are sent through an activation scheme. Depending on size and composition, some of the particles within the mixtures will be found suitable as CCN or INPs, and go on to create clouds.


Figure 2: The two current versions of the NorESM aerosol scheme OsloAero and OsloAeroSec: OsloAero represents the aerosol as a superposition of aerosol modes. OsloAeroSec puts nucleating particles in a sectional scheme, where they grow by coagulation and condensation. When particles reach the largest bin, they are transferred into the OsloAero modal scheme. This method was implemented to capture the growth of the smallest particles in a more realistic way. Figure from Blichner et al. [2020]3.

How do we constrain aerosol uncertainty?

With a climate model like NorESM, we have a unique tool for investigating different aspects of aerosol-climate interactions. We aim to constrain the uncertainties around aerosol effect on climate, meanwhile improving our understanding, mainly by focusing on three aspects:

Identifying the processes that are important for aerosol effects on climate, find the best and most efficient ways of implementing these processes in the model, and considering how our aerosol module interacts with the host model (i.e. CAM-Nor). Aerosol models are often developed to work with a certain host model, which can obscure the source of uncertainties; is it a cause of differences between atmospheric host models, or is it a result of how the aerosol module is implemented? We are currently working on making the aerosol module in NorESM (OsloAero) more independent from the host model (CAM-Nor), so that the aerosol module can more easily be exchanged with other aerosol modules, or modified and evaluated in itself. This effort follows an initiative from NCAR, where they are working on making their atmospheric model, CAM, compatible with multiple aerosol packages, and the GIANTS project4, which sets out to build a generalized interface between the host model and the aerosol module. This requires a substantial effort in refactoring and reorganizing our aerosol code, but will allow us to evaluate aerosol processes and assess our model in an unprecedented way, while also being at the forefront of climate model development.

References:

1: Forster, P., T. Storelvmo, K. Armour, W. Collins, J.-L. Dufresne, D. Frame, D.J. Lunt, T. Mauritsen, M.D. Palmer, M. Watanabe, M. Wild, and H. Zhang, 2021: The Earth’s Energy Budget, Climate Feedbacks, and Climate Sensitivity. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Masson-Delmotte, V., P. Zhai, A. Pirani, S.L. Connors, C. Péan, S. Berger, N. Caud, Y. Chen, L. Goldfarb, M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O. Yelekçi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 923–1054, doi: 10.1017/9781009157896.009.

2: Kirkevåg, A., Grini, A., Olivié, D., Seland, Ø., Alterskjær, K., Hummel, M., Karset, I. H. H., Lewinschal, A., Liu, X., Makkonen, R., Bethke, I., Griesfeller, J., Schulz, M., and Iversen, T.: A production-tagged aerosol module for Earth system models, OsloAero5.3 – extensions and updates for CAM5.3-Oslo, Geosci. Model Dev., 11, 3945–3982, https://doi.org/10.5194/gmd-11-3945-2018, 2018.

3: Blichner, S. M., Sporre, M. K., Makkonen, R., Berntsen, T. (2020). Implementing a sectional scheme for early aerosol growth from new particle formation in the Norwegian Earth System Model v2: comparison to observations and climate impacts. 10.5194/gmd-2020-357. https://doi.org/10.5194/gmd-14-3335-2021

4: Hodzic, A., Mahowald, N., Dawson, M., Johnson, J., Bernardet, L., Bosler, P., Fast, J., Fierce, L., Liu, X., Ma, P., Murphy, B., Riemer, N. and Schulz, M. (2023). GeneralIzed Aerosol/chemistry iNTerface (GIANT): a community effort to advance collaborative science across weather and climate models. To be published in Bulletin of the American Meteorological Society . Available at: https://journals.ametsoc.org/view/journals/bams/aop/BAMS-D-23-0013.1/BAMS-D-23-0013.1.xml