The manuscript presents the results of investigation of dissolved organic carbon and several major ions (sulphate, magnesium, calcium) concentrations, electrical conductivity measurements relationships with catchment discharge and seasonality using general additive models (GAMs) in the permafrost regions in Alaska.  The authors tried to link the selected stream chemistry parameters with discharge and seasonality. I am not sure if the GAMs are applicable in this context. For me personally, It was difficult to understand the presented results (e.g. figure 5 and 6) that used the derived statistical modelling such as sDOY and CV ratios, thus it was accordingly difficult to follow the discussion. The authors discuss the links of water chemistry with permafrost extent, topographic gradients, active layer. Perhaps, they can add catchment characteristics to their statistical model. As of now, the discussion reads more as a bare speculation. Significance of the work in terms of environmental implications must be more strongly highlighted.  I provided some detailed suggestions on how to improve the work.

The introduction is very long, LL54-104 I would recommend to include in the discussion section or made more focused. The portion of text may better fit into methods (LL76-84).

Section 2.1. Perhaps, it is better to combine the figures 1 and 2. Sections 2.1 and 2.2 are mostly land cover description, I would merge them.

Section 2.2. Table 1 seam more of land cover classification map, not catchment characteristics. I would move it to the supplement.

Section 2.3. for consistency LL203 replace sulphate with SO4 and elsewhere in the figures and text.

The main: You talk about discharge while presenting runoff data mm/day in the strict sense (see section 2.4). That’s confusing. Discharge is flow data (m3/sec) though a catchment outlet, while runoff is calculated per catchment area? I think if you talk about specific discharge in mm/day (e.g. LL260 you state this explicitly) you need to describe this in methods (section 2.4). Given the high variability of contribution area to runoff in this region, use of specific discharge is maybe not appropriate.  

LL259-260. Which figure or table do you refer to for median specific discharge numbers?

Fig. 7 and 8. I would recommend combining the figures, or better revising the figure to incorporate the topographical gradients and permafrost extent together, if possible.

The authors present a unique dataset of seasonal stream chemistry of 10 catchments spanning a latitudinal and permafrost gradient in the Yukon. The objectives were to unravel the drivers behind the seasonality of major ion and DOC concentrations in catchments, while accounting for stream discharge variability. The manuscript is generally well written and organized, however, the presentation of the results makes it difficult to link the findings to the conclusions and conceptual models. Some additional analysis and figure adjustments may help to clarify this point. It is worth noting that multi-year datasets such as this are challenging to produce and are a rare occurrence in the North. This is a valuable contribution to the literature, but some improvements could be made as suggested below.

General:

1. I was surprised about the lack of time series data presented. I think it would be worth presenting it, at least in supplemental data. It was difficult to grasp exactly when samples were taken over which years for what catchments.
2. I found it difficult to interpret the figures in results, even with the description in the results section. I think there is a missed opportunity to better link your results to your conceptual model. Can some of the key drivers you mention be included in these figures through colour, annotation, or organization? Some specific suggestions are below.
3. While the conceptual model intuitively makes sense, and the interpretation of the results seem valid, there is still a bit of a tenuous connection between the drivers of seasonality. 10 catchments is a relatively high number, so could there be some additional univariate or multivariate statistics be done to more firmly link the drivers to the results? As a very simple example, correlation between average catchment slope and DOC variability?

Section 2.2: This section could be removed and the information be combined with the previous section since there is already detailed descriptions of each of the catchments in 2.1.

Line 188: Can you be more specific about timing of the sampling? Was this just once in May/June and a second July/August? How are you defining spring and summer?

Line 204: Was this measured in all catchments continuously throughout the entire study period? Was this data used in the analysis or did you just use the YSI data? If continuous data for this exists for each catchment, I think it would be worth presenting.

Figure 3: Can the data be colour coded to permafrost extent or other key catchment characteristic? It might help tease out if there are other relationships going on and support the conceptual model.

Figure 4: Similarly, could you put these in order from lowest to highest permafrost coverage or catchment slope?

Figure 5: Can you include annotations or an additional example box that explains how to interpret these curves? Perhaps it could just be added to the caption as well.

464-467: Right now, your figures do not demonstrate this readily.

Figures 7 and 8: Remove the description from the figure itself. This should just go into the figure caption.