Clouds, climate and the Roaring 40s

By Richard Allan

In our new research we have traced large and long-standing biases in computer simulations of climate, affecting the tempestuous Southern Ocean, to errors in cloud that emerge rapidly within the atmospheric models. Biases evolve over time through knock on effects that shift the location of the battering winds known as the Roaring 40s. Our new method combines detailed computer simulations with observations of energy exchanged between the oceans and atmosphere that allowed us to better understand how deficiencies in climate models emerge in this key region for climate. This offers a route to improve the complex simulations necessary to make reliable climate change projections of the future.

The Southern Ocean is a pivotal component of the global climate system yet it is poorly represented in climate simulations, with significant biases in upper-ocean temperatures, clouds and winds. It plays an important role in the uptake of excess heat and carbon dioxide generated through human activities. However most coupled atmosphere-ocean climate models have substantial warm biases in Southern Ocean Sea Surface Temperature (SST) (see Figure 1) that have been linked to a lack of reflective super-cooled liquid water clouds in simulations. Our work has helped to elucidate the link back from the SST biases to cloud-related errors in absorbed sunlight and we identified a slower response of the region of intense winds affecting the Southern Ocean that further modify the biases.

 

Figure 1: Warm biases in simulated sea surface temperatures cover the Southern Ocean (large orange region near the bottom of the map) (IPCC AR5 Chapter 9, Figure 9.2(b)).

In our study we find that coupled climate simulations with warm biases in the Southern Ocean also receive too much heat flux at the surface in simulations using just the atmospheric part of the model (Figure 1). This suggests deficiencies that develop rapidly in the atmosphere are strongly linked with the long-term climatological bias in the simulations. Further analysis identified that too much sunlight due to unrealistic cloud is primarily to blame, consistent with previous research.

Figure 2: Link between sea surface temperature (SST) biases in coupled CMIP5 simulations and surface energy flux bias in the atmospheric component of the simulations (AMIP5) from Hyder et al. (2018)

To interpret the results a detailed framework was developed that resulted in a candidate for longest methods section of the year award! We attempted to summarise the main points in a schematic where we assume SST biases are linked to the energy budget of the upper mixed layer of the ocean as similarly applied in studies understanding ocean temperature variability. A further finding is that although initial deficiencies in cloud develop rapidly in simulations, the overall biases also relate to a response in the location of the “Roaring 40s” or more specifically the latitude of maximum westerly wind. As can be seen in Figure 3 below, positional errors in this “zonal wind maxiumum latitude” (ZWML) are also correlated with errors in the surface energy fluxes in the atmospheric simulations.

Figure 3: Errors in coupled model “zonal wind maxiumum latitude” (ZWML) correlate with errors in the surface energy fluxes in atmospheric simulations from Hyder et al. (2018)

Importantly, further detailed analysis demonstrates that our interpretive framework can be applied in targeting improvements to climate simulations that avoid “pasting over cracks” where one bias compensates for another. This offers a route to further improve the climate model simulations that are vital in providing realistic projections of how climate will change over the coming decades. The work was led by colleagues at the Met Office, involved a collaboration of many scientists and was conduced as part of the NERC DEEP-C and SMURPHS projects. The detailed research is available as an open access research paper in Nature Communications.

 

This entry was posted in Climate, Climate change, Climate modelling, Clouds, earth observation, Energy budget, Numerical modelling, Oceans, Solar radiation. Bookmark the permalink.