General comments

This paper reports on the development of a model that simulates hydropower generation in France at 30-minute intervals at individual power plants, using a hydrological model used in meteorological and climatological studies. The primary boundary conditions of the model are electricity demand and meteorological information. Electricity demand is allocated to three different types of power plants: run-of-river, reservoir, and pumped-hydro-storage. River discharge and electricity generation are intensively validated and discussed.

　Hydrologic models used in weather and climate studies are theoretically accurate, but often fail to achieve a simplistic objective of reproducing observed flow rates. The model constructed in this paper successfully achieved a reasonably accurate estimation of dam inflows in mountainous areas where flow reproduction is difficult. In addition, the experiment itself, in which dam releases nationwide are adjusted to match fluctuating electricity demand at 30-minute intervals, is significantly novel. At least, I have never seen or heard of such an experiment before.

　While the research is excellent, the manuscript is extremely difficult to read. First, the manuscript is long. It seems the authors are trying to describe everything that was implemented and devised. The manuscript needs to be shortened, for example, by moving the treatment of exceptions (e.g., lines 180-185) to the supporting materials. Second, the structure is not standardized. In particular, Methods and Results are written in a mixed style. Sections 2 and 3 should be put together as Methods, and Sections 4 and 5 should be put together as Results. Then, the part corresponding to introducing results in Sections 2 and 3 (e.g., Fig. 7-10) should be moved to Results, and the methodological descriptions in Sections 4 and 5 (e.g., lines 505-516) should be moved to Methods. Finally, some technical terms seem to be not adequately defined or consistently used throughout the text (e.g., dispatch, dispatchable production, poundage, and reservoir). Further polishing is needed to enhance readability. Although I am convinced that the study is novel and reliable in general, I could not fully understand the details due to the above issues. Perhaps I will provide further technical comments after the text is fully revised.

Specific comments

Line 166 “we thus select as feeding reservoir the one that maximizes the potential function…”: Why did the authors seek the feeding reservoirs by using an algorithm? Collecting the published facts or inferring from satellite images sounds more reasonable and accurate. A bit more elaboration is needed here.

Line 180-185: I appreciate the authors providing every detail, but too much information hampers readability. Perhaps this part can be further shortened or moved to supplementary material.

Line 280 “the result of the dispatch of the total power demand”: What does “dispatch” mean? A clear definition should be added.

Figure 12 legend: is “bias<-50%” correct? From the text, I read “Abs(bias)>50%”.

Figure 14: First, is “poundage production” identical to reservoir production? Second, if this is the case, why must run-of-river and reservoir production be shown simultaneously (they are discussed in two different subsections)? Third, the legend “differences” should be reconsidered because it is hard to know what variable was compared here. Finally, I feel that too much information is crammed into one figure. Maybe it might be more understandable if you split it into multiple panels.

Figure 15: Is “pilotable production” identical to “dispatchable production” in Figure 14? The line for “target production” is hard to see.

The authors presented a model to simulate hydropower within the routing module of land surface models with a more detailed representation of hydropower plants and their operations. Specifically, the model is validated to produce hydropower at a 30-minute time-step for individual hydropower plants of three types: run-of-river, reservoir, and pumped-hydro-storage. Such a level of detailed representation in hydropower simulation is quite impressive. However, the authors did not demonstrate the value of using such a model. For example, how could such a model be useful to dispatch hydropower in the presence of intermittent renewable resources like wind and solar, and what could be its implications for other types of storage (e.g., batteries)? Answering such questions could be a valuable scientific contribution of the paper. Instead, much of the paper demonstrated the validation of the model to replicate observed river discharge and hydropower at the power plants. Also, the authors’ claim of “operations at the scale of a power grid” seems a bit misleading. The hydropower is simulated based on exogenous (observed) demand for hydropower but not based on the operation of a power grid, i.e., the model did not dispatch hydropower to meet grid-level demand considering other generation, storage, and transmission facilities. Moreover, the paper is not well-written. It is long and written in the form of a technical report (e.g., there are 18 figures and quite a few sections/paragraphs of a single sentence). In summary, I suppose the study requires extended experiments and analysis to demonstrate the value of the proposed model, while the manuscript itself needs to be substantially improved.