Reviewer Blind Comments to Author In the manuscript entitled “Groundwater discharge and its impact on primary production in the shallow zone of a semi-enclosed coastal sea (Obama Bay, Japan)”, the authors focused on the submarine groundwater discharge (SGD) in Obama Bay, Japan. They adopted 222Rn as a latest powerful tracer of SGD, and reported the seasonal and spatial variations of concentrations of 222Rn as well as chlorophyll-a (Chl-a), nutrients, and salinity. They proposed different sources of 222Rn in the bay such as SGD and rivers and potential impact of SGD on aquatic biota.

General Comments The main conclusion of this manuscript is that not all 222Rn in the bay was derived from SGD. This seems to be reasonable, but was already discussed in more detail in previous studies. For example, Sugimoto et al. (2016) used the box model and calculated the 222Rn fluxes of both the river water and groundwater in the same bay, which the authors did not applied. At this time, this manuscript seems to be relatively descriptive and needs major revision. The authors also proposed four possible sources of 222Rn in the bay. I think these sources are poorly organized. SGD should be divided into fresh one (SFGD) and recirculated one (RSGD), which the author mentioned ONLY in the result section. I fail to understand the difference between (2) and (3). Are they subsurface flow or baseflow? The authors also reported that high-222Rn seawater around the zone C-D is moved toward the G3 station by the northern wind, which seems to be reasonable for me. But there are no firm evidences that the original high-222Rn seawater was derived by groundwater, as the authors proposed. How is the impact of river water? More importantly, this wind-derived advection seems to be just a local event, in limited area between the station C and G3. Is this advection common and important for aquatic biota “over the whole bay”, which should be the overall theme of this manuscript? Significant correlation between 222Rn and Chl-a concentrations is another main conclusion of this manuscript. However, I guess this is just a spurious relationship, of which a hidden factor is major non-metal nutrients (PO4 in particular, as the authors discussed in page 20). Is Rn necessary for biotas, like Cu or Zn (these elements work as micro-nutrients, involved in various enzyme processes)? It took me a while to understand the datasets of this manuscript. I found many basic mistakes and inappropriate descriptions in the figures and captions (e.g., too small letters, and discrepancy between the figure and main text). For details, please see the specific comments. I think that the authors’ datasets of SGD along the coastline in different seasons are important for further understandings of not only SGD but biogeochemical cycle in this bay, but just not enough. In order to provide a significant improvement on previous knowledge, I strongly recommend the authors to discuss in more detail: for example, the mass balance approach for each source of 222Rn, spatial and seasonal variations of groundwater flow streamlines as discussed in Robinson et al. (2007), Adv. water. resour., or relationships between SGD fluxes and actual phytoplankton growth. I am sorry that I cannot be more positive about this manuscript at this time.

Specific comments: L. 101: “Figure 1a” should be mentioned before “Figure 1b”. L. 136-144: In this section, the authors showed intervals of chemical analysis (salinity, Chl-a, and 222Rn) along the coastline. But this explanation is difficult to follow for me, when I compare both the main text and figures (Fig.2 to 4). For example, the measurement intervals of both salinity and Chl-a were 1 min. But Chl-a datasets shown in Fig.4 seem to be too sparse compared to salinity datasets in Fig.3. Please clarify. L. 146: Different from this text, the authors mentioned the possible impact of tidal period on 222Rn variations in Fig.8 and L. 248-254! In addition, they should clarify the timescales of “tidal period” (diurnal or monthly). L. 170-189: What are analytical errors? L. 200-308 (result section): The result section should include only straightforward reporting of your findings. Some paragraphs (e.g., mixing line between seawater and groundwater, and SFGD rates) should belong to the discussion section. L. 267-270: This text in very confusing. What is “0.05”? What is “the whole monitoring period”? L. 272: What is the difference between “submarine groundwater discharge” and “seepage flux”, both of which are abbreviated to SGD? L. 289: Please show me possible factors for this abrupt decrease of the seepage flux in May. L. 318-322: No 222Rn datasets are shown in Fig. 10! L. 346-352: What is the impact of weather (rain in particular)? L. 365-367: I suggest that the authors add N:P:Si ratios of aquatic biota in the bay, for further understandings of SGD impact on the ecosystem. For example, is this different from the Redfield ratio? Fig. 1b: Plots are busy and confusing. Fig. 2-4: I strongly recommend the authors to use the same scales among different seasons. For example, it took me a while to understand that 222Rn concentrations were highest in March. Please also see my comments for L. 136-144. Fig. 5: Does this figure show the same datasets as Fig. 2 and 4? Why are the Chl-a datasets around the zone C-D in March lacking? Fig. 7: This figure have many problems. In the caption, the lines are described as “solid and dashed lines”, while they are described as just “bold lines” in the main text (L. 237). Two open squares (RW) are plotted. The endpoint for oceanic water (L.240-241) is lacking. I think that the discussion “half the 222Rn plots were located below the conservative mixing line” (L. 334) is not convincing. The authors adopted the annual average values of 222Rn in the river water, groundwater, and seawater proposed by Sugimoto et al. (2016) as the endpoints of this figure. When I consider the seasonal variations of these values, the relationships of position between the mixing lines and the author’s datasets are not significant. What kind of “groundwater” are the authors mentioning (fresh, recharged, or wind-derived one)? In addition, I think that the residence time of water in the bay, that is, mixing time of seawater with river water/seawater is also important because 222Rn is a short-lived radioisotope. Fig. 8-9: The arguments as to why the authors represented the datasets of different seasons (summer for Fig. 8 and spring for Fig. 9) are not clear. Please also see my comments for L. 289. Fig. 10: The letters are too small to read. What do the y-axis in the right side represent? Table 2: How did the authors calculate the SFGD values, even though the SGD values are not determined in some samples?