The paper has improved a lot compared to the Discussion status, making a clearer case regarding the design and purpose of the suggested model type. But below I list a number of further amendments and clarifications that should be considered, which are mainly about more cautious formulations.
Lay out upfront that the suggested novel model is currently a mere concept, not yet tested and implemented in practice. In some parts it reads as if it was already part of an IAM, but other parts suggest this is not the case. Related to this: Especially section 3.2, which uses some equations and model structural explanations, comes across as a more formal model description than other parts (which are more informal and wordy). That said, the text meanders a bit between a model description, a more general description of a model concept, and an actual implementation of a model within existing schemes. This description could be more ‘standardised’ to further clarify where the (proposed) scheme currently stands.
The Abstract focuses on agent-based modelling. I’m not sure this is the focus of the main text (especially there seems to be no discussion regarding existing agent-based models of which there are many); on the other hand, the Abstract could mention more explicitly the main advantages of the proposed model framework. So many of these are listed by the end of the paper in section 4 – in my opinion too many, and some of them are rather general, so this section can also be focused; especially since much of it is currently kind of speculative, as no such model applications have been tested, i.e. it is not sure whether it all will work so straightforwardly once implemented.
You could make better use of the schematic figures explaining your model concept, i.e. guide the reader in the main text what they actually show.
Section 2.1. This is mainly about increasing the resolution of models; not so exciting or new, but okay. More importantly, you may want to link this more directly to section 2.3, which is also about trade.
Page 5, lines 7ff: Is it really so, IAMs mainly focused on GHG emissions with only a few now incorporating also agriculture (in what way actually – is that about more than just GHG emissions from agriculture) – but what about water, ecosystems, carbon cycling etc., aren’t these components of IAMs too for a while already? Similarly section 3.1.1: A bit confusing what the grid-based ecohydrological model is; isn’t that part of the IAM anyway? See also comment above. This section seems to require better integration with 3.1.2 (and 3.2) as it is not yet clear to me how model components interact and which are already existing / standard and which is the new component. Also page 8, lines 1ff: Isn’t LPJmL also a hydrological model?
Same page, line 30ff: There is a new interesting paper just published (http://www.nature.com/nclimate/journal/v7/n7/full/nclimate3310.html) which seems to strongly support your claim here, also using an example application.
Same page, bottom lines: said before; there are other such redundancies (repeating general remarks already made in the Introduction) in some places, please check.
Page 9 lines 18ff: It is not clear whether the text here (and in some other places) is meant to be a general remark or whether it is actually part of your model concept.
Many, many references: any potential to leave out some? |