This paper is excellent as a large compilation of INP data that has been processed in a consistent manner. The effort is to be commended for that reason alone. It is also a very well written manuscript, and with most of the details one would wish for, and the abstract highlights several key points: well mixed populations that do not vary greatly overall between northern and southern continental and marine sites, shortterm variability dominating at all sites, certain site specific aerosol drivers of INPs, but no universal driving aerosol property driver, and no indication of anthropogenic influences. Nevertheless, as I read the paper as it is currently organized, I struggled in C1 ACPD Interactive comment Printer-friendly version Discussion paper knowing how to relate the method and results from the standard FRIDGE method to drop freezing assays (or the immersion mode method sometimes applied using the FRIDGE device), which are possibly the most widely used present method. It seems to me that two things are required to assist readers in understanding the nature of the results, and potentially how to consider them in relation to immersion freezing data. First, the title should explicitly describe the basis for INP measurements. In other words, "Long-term deposition/condensation freezing INP measurements.. ." or something to that effect. When one sees the INP versus ice supersaturation data in this manuscript, there is no discontinuity that occurs at water saturation (as the authors readily note), and so it seems apparent that immersion mode freezing is indeed not represented at all. The authors provide a discussion of the dominant mechanisms at play in the data and the likely underestimate in comparison to immersion freezing mode operation of the FRIDGE only very late in the paper. This is critically important in understanding if the findings can be ascribed only to deposition and condensation-freezing mode INPs, or if the same is expected for immersion freezing populations. I suggest in the specific comments that the methods used may indeed limit assessment of strong local/regional impacts, at least for biomass burning. Of course, it will not be possible to make a conclusion about what was not measured, but it should be highlighted as a question for future inspection. This should all be made crystal clear. Hence, the second recommended change is to bring a discussion forward of what types of INPs the data describe, and what types the generalized results may not describe. It will not detract from the great effort the authors have made to collect large quantities of ice nucleation data from multiple sites and discern answers to some of the key and enduring questions related to INP sources. However, I believe that it will better frame future needs. Specific Comments 1) Introduction