|Review of paper:|
Title: Dust aerosol radiative effects during summer 2012 simulated with a coupled regional aerosol-atmosphere-ocean model over the Mediterranean
Author(s): P. Nabat et al.
MS No.: acp-2014-527
This paper aims at characterizing the dust aerosol radiative effects over the Mediterranean. The focus is done over a specific period, the summer 2012, in which in-situ measurements were made as part of the campaign TRAQA/CHARMEX. This paper addresses several different questions: (i) the ability of a model dust emissions scheme to produce realistic fluxes and then aerosols concentrations, (ii) the impact of these dust concentrations on the daily radiation variability, (iii) the impact of the use of a deterministic scheme in place of a climatology, (iv) climatological differences between 'dusty' or not days, (v) a comparions between several AOD satellite products.
This leads to a very long paper, mixing several concepts: climate and a specific studied case, model sensitivity tests, model comparisons (with MACC), data comparisons, impact of aerosols on radiation, comparisons between model outputs and measurements. The result is sometimes not very clear and several issues has to be corrected before publications. Removing unnecessary parts could make the article more clear and precise.
The authors made a complete revision and proposed answers to all questions adressed by the reviewers.
However, a large number of problems are not really solved. If some sections were moved or reorganized, the paper has the same length (24 pages in pre-print format) and the same number of figures. Note that for Figure 11, the geopotential height disappears (but is cited in the caption).
1. The "climatological" point of view remains in the whole manuscript (except in the title). Even if the authors remind us the definition of climate, a study over a summer is not really a climate study.
2. The domain size. The justification here is that the domain was enlarged compared to the previous ALADIN configuration. To enlarge a non-adaptated domain is not convincing if the domain remains not large enough. But we can consider here that the domain may cover the largest part of aerosols observed over the Mediterranean sea,
3. Use of the MACC outputs for the comparison. The section was shortened but not removed, the authors considering this is important.
4. The definition of the dusty days was changed. The use of the Angstrom coefficient could be useful. This is strange to read that a very important parameter is not used to define a model criterio because the model does not compute this parameter. In this case, instead of an unsuitable criterion, change directly your model. In my frist review I wrote "the results are difficult to understand". The authors answered "We admit that the method could be difficult to understand". There is here a confusion: the method is not "difficult to understand", it is not robust. Thus, the results (not the method) could be difficult to understand.
Finally: I suggest that the authors make minor revisions in the manuscript. It is important to remind more clearly the limitations of their study, by adding more precise arguments in the section "4.5 Discussion". Indeed, as it is written, the paper tends to oversell what is really inside. It is needed to remind that the model is a climate model and, thus, has some poor or simplified parametrizations: for example, the number of bins is low and may conduct to erroneous conclusions when the 'aerosol fine mode' is needed.