Please refer to the attached PDF for the detailed comment.

General comments

The manuscript “Downpour Dynamics: Outsized impacts of storm events on unprocessed atmospheric nitrate export in an urban watershed” by Bostic et al. expands on the existing literature using triple isotopes of nitrate to partition storm event loads into atmospheric and terrestrial fractions. The main finding is that stormflow exports more atmospheric nitrate and baseflow exports more terrestrial nitrate in an urban watershed and there is not much difference in a non-urban watershed. This is not necessarily a surprising finding, but the results are described well and presented with interesting and unique figures with the data to support them. I think the strength of the paper is in its comparative nature. While a few others have used D17O to partition loads into terrestrial and atmospheric fractions, it is exciting to see how these relative partitions vary between comparable urban and non-urban watersheds. I think this paper would benefit from a few more citations particularly in the methods section to further differentiate it from other work in the field.

Specific Comments

Lines 39-42: I don’t know if it is still correct to say that export is rarely partitioned into atmospheric and terrestrial sources. It is an important part of the literature, and this study adds to the cumulative knowledge in this area, but I wouldn’t frame it as something no one else has done.

Lines 59-60: I would caution against referring to D17O as triple oxygen isotopes. While three isotopes are relevant to the measurement of D17O so the method is sometimes known as triple oxygen isotope analysis of nitrate (Kaiser 2007), but the resulting value for D17O itself is not triple. D17O values could also be called the 17O anomaly (Michalski 2003). To add to the confusion, the way you are using these isotopes for quantifying loads is often referred to as triple nitrate isotopes, where D17O is one of the three isotopes along with d15N and d18O (Liu et al 2013, Hale et al 2014, Rose et al 2015). To minimize confusion here and throughout the paper, I would keep it as either “D17O values” or “triple oxygen isotope analysis.”

Line 79: What are "moderate frequency samples"? There is no reference point for the time interval.

Line 101: 45 minutes to 12 hours is a very wide range of sampling intervals. Is this the average among events with widely different sampling intervals, or does this change within a given event?

Lines 166 – 170: I am not sure what traditional methods you are referring to. Maybe a citation or two would help. The other papers I have seen that quantify NO3 loads use the discharge that corresponds with each individual grab sample (ie Hale 2014).

Lines 222-223: The fractional export plot seem like a very interesting method. Could you add a few citations for other that have used this, unless you are the first?

Figure 2: I would recommend using open and filled circles/triangle for your two sites. With the colors as they are, they will both print grey in black and white.  Also, is each point a single storm event? It is a bit confusing to have both NO3 atm and NO3 terrestrial plotted as they are directly inverse of each other.

Figure 3: Event mean does gloss over the changes in source load within a given event. Though I suppose it is necessary for the 1:1 comparison with baseflow for this plot. Might be worth discussing in the body of the text though.

Technical Corrections

Line 16:  Spell out 8