In this measurement report, the number size distribution of sub-40 nm particles derived from different instruments in the Amazon rainforest has been reported, including the diurnal pattern and the sources of nanoparticles, especially for the particles below 10 nm. This study provides valuable information in understanding the particle sources in Amazon region and highlight the necessity about the study on nanoparticles. However, further discussion needs to be added to make the results more robust. I recommend this paper being accepted after the below issues are addressed.

1. Line 45, It is necessary to clearly state that with or without a core sampling device, how much the sampling efficiency was changed for PSM.
2. Section 2.2.4, is there any calibration was conducted before or after for the intercomparison between nCNC and TSI SMPS? Or the two instruments' inherent system error was lower than 20%? It would be better if the authors can the information about if there are any other studies did such intercomparison, otherwise, how could we know the 20% discrepancy is a “good agreement”. In figure 1, please make sure the slope is positive or negative.
3. Line 210, it should be better to give the definition of “Amazonian bananas” here, otherwise, it may lead the reader to mistakenly think it refers to banana shaped NPF.
4. Line 220-227，is the “high RH range (90-100%)” corresponding to the precipitation? How much is the data excluded from the dataset due to precipitation or high RH? As the mean RH during this measurement was ~85%, that means the probability of the RH exceeding 85% is high.
5. Line 240, have you looked in to the information about radiation and cloud cover in this study? In addition, as the authors have mentioned the high RH during the measurement, it can also be a reason why no typical NPF was observed. As water vapor can also contribute to the high condensation sink in the ambient air, whereas the CS is normally calculated based on the dry PNSD.
6. Line 263-264, is that reasonable that the December data can represent the dry season as the mean RH was approximate 79%? The authors may refer to the previous literature to check the typical RH levels during the dry season in Amazon.
7. Line 368, it is addressed that on clean days the photochemistry is favorable for the production of nucleated particles. however, it is also stated that there is no typical NPF cases occurred. So that means the burst of nucleation existed, but without clear growth processes. Please give more explanation about the reason.
8. Line 405, the first sentence is not rigorous, as the authors also mentioned other studies (such as Wimmer et al., 2018 and Zhou et al., 2002) also conducted the particle size measurement below 10 nm. Please highlight the difference between this work and the others.
9. Table 1, the unit of particle number concentration should be given

This measurement report presents size distribution measurements of aerosol particles and ions with a particular focus on sub 10nm sizes captured at a measurement tower in the central Amazon rainforest. The measurements are taken with different state-of-the-art instruments during a period representing the wet and dry season of the region. The data is analyzed with respect to seasonal differences, diurnal variations and the influence of pollution. The paper provides valuable insights in aerosol particle concentrations and size distributions, particularly for diameters below 10 nm, for the Amazon rainforest. However, major improvements are required. I recommend to accept the paper after the following improvements have been addressed:

1. Section 2.2.4: Was the intercomparison also repeated during and/or after the measurement campaign to ensure that the instruments didn’t change during the measurement campaign? Also, the presented intercomparison is only considering two of the three instruments. It would be good to have an intercomparison between all the three instruments.
2. Line 209-210: The combined size distribution in Figure 2c disagree significantly for the sizes around the stitching point of NAIS and SMPS at 40 nm for most of the measurement period. For some periods the disagreement is more than one order of magnitude. The disagreement can also be seen in Table 1, where SMPS reports a median of 27 #/cm³ in the size-range 10-40 nm for “all period”. However, for NAIS (particle) the size range 12-40 (i.e. the subtraction of the concentrations in the size range 2-4 and 4-12 from the concentration for the size range 2-40) reveals a value of 310, which is one order of magnitude larger than the SMPS value. Why is there such a large disagreement and which instrument is correct?
3. Line 210-211: It would be good to explain the term “Amazonian bananas” and to mention, where the "Amazonian bananas" can be seen in Figure 2c.
4. Line 226-227: What are the criteria to reject the data. In Figure S3, outlier with higher concentrations are also visible for relative humidity values as low as 60% - were these datapoints also rejected. Could a to restrictive rejection of such periods also by mistake reject a NPF event? It would be good to also mention how much data was rejected?
5. Line 259: December concentration for size-range “large (4-12 nm)” of 491 cm^-3 does not agree with Table 1, where 497 cm^-3 is written.
6. Line 269-271: From Dec to Jan, the median concentration of particles with diameters between 1.5 and 3.5 nm measured with PSM is increasing from 371 to 573. This contradicts the findings with the NAIS instrument, where the concentration in the intermediate (2-4nm) size range is decreasing from 931 to 558. Why is there a contradicting trend from the two instruments covering a comparable size range?
7. Line 284-285: Where is the number concentration of 397 cm^-3 coming from? It is not one of the values from the Table 1. It should be better explained, how this value was derived. Also: Table S1 lists two values 450 and 610 labeled as “this study”. How are these values derived, since they also differ from the value mentioned in Table 1?
8. Line 329-330: The described trend of the particles larger than 15 nm (D_15-40nm) is very hard to see in the plot - especially for the Dec. period. The scale for the colormap starts at 1e1. However, the colors between 1e1 and 1e2 are not present in the plot. Starting the color scale at 1e2 would help to make it easier to see the mentioned trends.
9. Line 353-354: Where does the BC mass concentration come from and how was it measured? More explanation is required.
10. Line 374–379: “plume” should be changed to “polluted” and “clear” should be changed to “clean”
11. Line 374-375: It would be good to indicate the time of the large number of nucleation mode particles, since there is also a period with a very low nucleation mode number concentration from approx. 05:00 to 11:00. For Figure 10, also the color scale starts at 1e1. However, a min. value of 1e2 would most likely help to see features and trends better.
12. Line 375-376: “Figure 10a” should be changed to “left panel of Figure 10” and “Fig. 10b” should be changed to “right panel of Figure 10”.
13. Line 405-407: In the “Results and discussion” chapter, many of the sub 10nm findings of this study are compared to measurements reported in the literature. It would be good to specify more precisely which aspect of the paper is the “first-time” presentation.
14. Table 1: Since the sizes represented by the different instruments are not all comparable (as mentioned in 217-218) it would be good to indicate, that the sizes reported for PSM are different compared to the sizes reported by the other listed instruments.