Regional vs Global Models From The Perspective of a Polar Climate Scientist

By: Charlotte Lang

There is a debate in the world of polar climate and ice sheet surface modelling about global (GCM) versus regional (RCM) models and each side is trying to convince the other that they do better: global modellers insist that regional models don’t add much to the topic while regional modellers try to convince them that their models can complement or even improve the outputs of global models.

Although originally I was a regional modeller as a user of MAR (Modèle Atmosphérique Régional, Fettweis et al., 2017) in Liège, Belgium, I have since joined the dark side as a global modeller in December 2020 when I started a postdoc at NCAS. 

Do these 2 types of models really have to be opposed? Can’t they be used in a complementary way?

Let’s review a few of the advantages of each of them.

Resolution: advantage RCMs

For a long time, the biggest advantage of regional climate models was their ability to run at a higher spatial resolution (~10 km) than highly time consuming GCMs (~100 km), allowing them to better represent smaller scale processes and to be calibrated for specific climates (Fig. 1). For Greenland for example, it means a better representation of the narrow ablation zone, the marginal area of the ice sheet where the amount of melt occurring at the surface in summer exceeds the mass gained in winter through snowfall. In Svalbard, it means a better representation of the hilly topography that gave its name to the main island Spitsbergen (“pointed mountains” in Dutch) and a better representation of orographic precipitation.

Figure 1: Representation of GCMs and RCMs. Source: Ambrizzi et al. (2018); Figure 1.

On the other hand, running on a limited area means RCMs have to be told what happens at their boundaries, which have to be “forced” at regular time intervals by the outputs of a global model. That is the biggest criticism one can have against RCMs: they can only be as good as the model used to force their boundaries. Feed them with a “good” GCM and they might even improve their results (and therefore a better representation of orographic precipitation); feed them with a biased GCM and their outputs will display larger biases as well.

Ice sheet surface processes: advantage RCM

When the progress in computing allowed GCMs to increase their spatial resolution, us polar climate and ice sheet surface modellers could still argue that we could better simulate the interactions between the climate and the ice sheets surface with RCMs. Indeed, some RCMs such as Liège’s MAR or IMAU’s RACMO (Noël et al., 2018) include snow modules allowing explicit simulation of the energy and mass transfer between the atmosphere and the surface of the ice sheets. Global models for their part didn’t include such complex models and users had to resort to forcing simpler and often empirical snow models with their climate, missing some of the feedback between the climate and the snow surface.

The rise of Earth System Models: advantage GCMs

More recently, regional climate models suffered a blow with the development of a new class of global models, Earth System Models (ESM). ESMs, like the British UKESM (Sellar et al., 2019), are complex models coupling many components of the Earth System: atmosphere, ocean, vegetation, biogeochemistry… UKESM also includes a complex model for the surface of ice sheets coupled to a thermo-mechanical ice sheet model (Smith et al., 2021) simulating the dynamics of the ice sheets, allowing it to take into account the effect of changing ice sheets on the climate and vice-versa. This feature is of particular interest in long term future projections (>2100) as the surface of the Greenland ice sheet is expected to lower as it melts, further increasing the near-surface temperature and therefore surface melt through the melt-elevation feedback that RCMs and their fixed geometry can’t represent (Fig. 2). A changing ice sheet geometry could also modify the atmospheric circulation on and around the ice sheet, including a weakening (resp. strengthening) of the katabatic winds, which could further enhance (resp. dampen) the positive feedback loop.

Figure 2: Melt-elevation feedback in a fixed vs dynamical ice sheet geometry.

Efforts have been made to couple RCMs to ice sheet models, like the PARAMOUR-EOS project (https://www.elic.ucl.ac.be/users/klein/PARAMOUR/index.html) aiming at coupling several RCMs to ice sheet and ocean models over the Greenland and Antarctic ice sheets but the use of coupled models is still quite marginal in polar climatology.

Furthermore, the system of elevation classes in UKESM allows it to downscale the climate and surface variables needed to force the ice sheet model from the lower resolution atmospheric grid onto a much higher resolution ice sheet grid. And here one of the last advantages of working with an RCM disappears.

Yet regional climate models are still useful in many aspects. Forced by reanalyses, they can rapidly produce high resolution simulations of regional weather events like the extreme melt of the surface of the Greenland Ice sheet in 2012 or the deadly floods that affected Belgium and Germany this summer. Forced by an ESM climate, they can rapidly run with a focus on specific surface ice sheet processes whose numerical representation need improving or are yet to be included in the models. RCMs are not dead (yet)!

References:

[1] Ambrizzi, T., M. Reboita, R. Rocha, and M. Llopart, 2018:  The state-of-the-art and fundamental aspects of regional climate modeling in South America. Annals of the New York Academy of Sciences, 1436, https://doi.org/10.1111/nyas.13932.

[2] Fettweis, X., and Coauthors, 2017: Reconstructions of the 1900–2015 Greenland ice sheet surface mass balance using the regional climate MAR model. The Cryosphere, 11, 1015–1033, https://doi.org/10.5194/tc-11-1015-2017.

[3] Noël, B., and Coauthors, 2018: Modelling the climate and surface mass balance of polar ice sheets using RACMO2 – Part 1: Greenland (1958 – 2016). The Cryosphere, 12, 811–831, https://doi.org/10.5194/tc-12-811-2018.

[4] Sellar, A. A., and Coauthors, 2018: UKESM1: Description and evaluation of the U.K. Earth System Model. Journal of Advances in Modeling Earth Systems, 11, 4513–4558, https://doi.org/10.1029/2019MS001739.

[5] Smith, R. S., and Coauthors, 2021: Coupling the U.K. Earth System Model to dynamic models of the Greenland and Antarctic ice sheets. Journal of Advances in Modeling Earth Systems, 13, e2021MS002520, https://doi.org/10.1029/2021MS002520.

 

This entry was posted in Climate, Climate modelling, Cryosphere, Polar. Bookmark the permalink.