This paper is an excellent approach to calibrating a proxy system model, and its method is a valuable contribution to the proxy-system model community. The extensive observational dataset used to validate the calibration is impressive. I have mostly very minor comments. My only large-scale ish comment would be that the paper felt slightly unfocused at times. That isn’t the end of the world, as it is a methods paper and likely part of a broader project to understand these lake systems throughout different past climates, but I think adding a bit more (a paragraph at most should be more than enough) discussion about a couple things could help focus this manuscript and make your results more relevant to a wider audience:

1. What do your results reveal about lake dynamics/PRYSM? Do the parameter ranges that perform best reveal anything about these lakes, lakes in general, or the proxy system model? 
2. It seems like your results suggest that deuterium is fairly sensitive to both forcing inputs and parameter choice. This would imply to me that using lake proxy system models for isotope-based proxies doesn’t work very well. Is that a fair takeaway? Can anything be done to improve the PSM performance?

Otherwise I think this paper is pretty much ready to go! Great work. 

Minor comments:

Line 36: ‘i.e.’ should be ‘e.g.’ (unless I am misunderstanding what is meant by this parenthetical) 

Line 46 when you say ‘individual changes’ do you mean the changes are applied individually, and not all at once?

Line 46-47 this should either be ‘increased/decreased’ or ‘increase/decrease in’ for each of these examples 

Lines 55-58 the sentence ‘resulting in a proxy time series that is translated into climate variables’ is confusing to me, because my understanding of forward models was that they translate climate variables into proxy units rather than the other way around. But more broadly  I am not sure you need to go into too much detail about Bayesian modeling schema given that you are not using one. You can just say something like ‘some proxies, like terrestrial pollen, don’t require PSMs that incorporate lake dynamics but, x and y that use water isotopes, do’ 

Line 230 could you explain a bit more how this creek is relevant for your runoff model? Does it runoff into either of your lakes? Does the proxy system model include catchment-wide area information? 

Table 2 could you include the distance of each source of observational data from to your lake sites? Or does this not apply if they are within the same grid cell for ERA5?

Line 388 are these RMSE values the average of all 1000 simulations or the 30 you selected? Could you add text about why Bear lake may have such a high error compared to red pond to the discussion?

Line 600 does the turbidity (and thus shortwave extinction coefficient) of the lakes change over the course of the year? Does the parameter related to this value also change month to month, or is it a single number?

Lake sediments preserve valuable information about past climate change, such as air temperature and precipitation, but how to interpret them properly is challenging because it is affected by multiple climate signal as well as lake dynamics. The authors integrated a simple runoff-isotope module by using multi-year observational datasets from the two lakes with contrasting morphometry and hydrology. This study about the modern process of runoff-lake water isotope is very useful to interpret proxies in lake sediment. It is worthy of being published after minor revision. Here are some comments:

1, Language is difficult to understand for me. It would be good to improve it and made it more readable.

2, This study simulates the runoff-lake water isotopes on a very high resolution (hourly or daily) by using comprehensive in-situ observations of climate variables and lake water isotope and temperature. However, lake sediment is usually inter-annual, decadal or centennial time scales. Factors influencing lake water isotope may differ on different time scale. How to balance the time scale should also be considered.