Gains and losses in surface solar radiation with dynamic aerosols in regional climate simulations for Europe

The solar resource can be highly influenced by clouds and atmospheric aerosol, which has been named by the IPCC as the most uncertainty climate forcing agent. Nonetheless, Regional Climate Models (RCMs) hardly ever model dynamically atmospheric aerosol concentration and their interaction with radiation and clouds, in contrast to Global Circulation Models (GCMs). The objective of this work is to evince the role of the interactively modeling of aerosol concentrations and their interactions with radiation and clouds in Weather Research and Forecast (WRF) model simulations with a focus on summer mean surface downward solar radiation (RSDS) and over Europe. The results show that the response of RSDS is mainly led by the aerosol effects on cloudiness, which explain well the differences between the experiments in which aerosol-radiation and aerosol-radiation-cloud interactions are taken into account or not. Under present climate, a reduction about 5% in RSDS was found when aerosols are dynamically solved by the RCM, which is larger when only aerosol-radiation interactions are considered. However, for future projections, the inclusion of aerosol-radiation-cloud interactions results in the most negative RSDS change pattern (while with slight values), showing noticeable differences with the projections from either the other RCM experiments or from their driving GCM (which do hold some significant positive signals). Differences in RSDS among experiments are much more softer under clear-sky conditions.


-Introduction
The WRF outputs were recorded every hour, in particular for the variables of interest here, namely Surface Downward Solar Radiation (RSDS) and Total Cloud Cover (CCT). We also compute AOD at 550 nm from the WRF-Chem outputs following Palacios-Peña et al (2019b). RSDScs and AODcs will denote the RSDS and AOD values under clear sky conditions, computed here at the daily time scale from those days with values of CCT lower than 1%. The RSDS and CCT data simulated by the driving GCM runs were used for comparison purposes. We also retrieved the AOD at 550 nm as seen by the GCM from the MACv2 data (Kinne et al 2019), whose anthropogenic changes are in accordance with the RCP8.5 while its coarse mode (of natural origin) was not allowed to change.
Summer (JJA: June-July-August) means of all the variables were used in the analysis. The analysis involving RSDScs and AODcs will be considered only over those grid points where at least 75% of the summer mean values in the time series (i.e. at least 15 records per period) are not missing values (which, according to our methodology, would occur only if all days within a summer season have CTT values ≥1%).

-Results
We focus on the summer season (JJA), when solar energy provides its most, AOD tipically reaches the highest values and the aerosol radiative effect has been proven to be strongest (Pavlidis et al 2020 in the RCM experiments) south and northward (up to 20 and 30% respectively), and negative values in between (10-15%, eventually up to 25%). Nonetheless, there still exist significant differences within the set of WRF experiments (Fig 1a-c), in which this research puts the focus.
The inclusion of interactive aerosols (ARI and ACI experiments) reduce the JJA mean values of RSDS in central and northern parts of our domain by a few percents as compared to the BASE experiment (Fig 1a,b). This reduction is generally stronger in ARI than in ACI. Consequently, the ACI minus ARI pattern (Fig 1c)  both ARI and ACI lead to more cloudiness in central and northern regions and less cloudiness southward, specially in ACI (Fig 1d-f), which is well correlated with the spatial distribution of the differences between experiments in RSDS. Also, the dynamic treatment of aerosols lead to noticeable differences (up to 10%) in the AOD values between ACI and ARI (Fig 1i), and the AOD climatologies from these two experiments provides a consistently non-nule picture (Fig 1g,h; nule values can be considered for BASE). However, the paterns for AOD do not correlate with those for RSDS. In fact, the temporal correlation at the grid point level between the series of differences in RSDS and in CTT is above 0.8 (negative) in most of the domain, while differences in AOD hardly Therefore, there is an overall a direct and predominant link between the aerosols effect on cloudiness and its impact on the amount of solar radiation reaching the surface. Contrary, the effect of interactive aerosols schemes on AOD seems to play a minor and more local role in certain locations, where it eventually can help to straightly explain the differences in RSDS between ACI and ARI as the matching between the RSDS and CCT differences devanishes. For instance, a closer look to local differences between ARI/ACI and BASE reveals regions (in central and eastern Mediterranean Europe) where, in spite of the less CCT simulated in experiments with interactive aerosols (Fig 1d,e), they also simulate less RSDS than BASE (Fig 1a,b). This could be explained by the differences in AOD and its locally relevant impact on RSDS over thse regions, as pointed out by Supp Fig 3d. Also, over areas of central Europe, while differences between ACI and ARI in CTT are small (Fig 1f), ACI provides higher values of RSDS than ARI (Fig 1c), which could be explained by the larger AOD values in the ARI simulation (Fig 1i).
Under clear-sky conditions (Fig 2), both the spatial correlations between patterns of AOD and RSDS differnces, and the temporal correlations between the respective series computed at the grid point level ( Supp Fig 3g-i, 5g-i and 6), support the relevant role of the AODcs variable for the simulation of RSDScs. Nonetheless, differences in RSDScs are lower than differences in RSDS, basically nule between ACI and ARI. It is important to note that this analysis considers coincident clear-sky dates between the pairs of experiment being faced (the percentage of days retained under this approach can be seen in Supp Fig 7). Without this restriction (see the percentage of days retained in Supp Fig 8)  Vulnerabilities and resilience of European power generation to 1.5 C, 2 C and 3 C warming.