The authors present a valuable technical note to the CRNS community on understanding local vs. global calibration uncertainty. The authors provide a comprehensive dataset from various CRNS experiments and calibrations across Europe. The article is well written, and data is appropriately documented in the appendix and online. I do have some suggestions and comments that need to be addressed prior to publication.

Line 20: Typically we don’t included dimensions for the a0, a1, and a2 as they are derived coefficients from Desilets 2010. The units make the equation balance, but I am not sure if they need to be included?

Line 123 eq 5. McJannet and Desilets 2023 recently addressed a similar topic on using different neutron monitors and correcting for spatial variability. I am not sure exactly how these approaches are the same or differ, but the citation should be added with a short discussion of that paper. Also in the original COSMOS project an fscaling factor was included. This was also tied to the appropriate reference pressure selected using the COSMOS online calculator (http://cosmos.hwr.arizona.edu/Util/calculator.php). Use of the long-term average pressure for reference pressure instead of the one from the COSMOS online calculator did cause some problems when using the rover at different locations in my experience.

McJannet DL, Desilets D (2023) Incoming Neutron Flux Corrections for Cosmic-Ray Soil and Snow Sensors Using the Global Neutron Monitor Network. Water Resources Research, 59(4):e2022WR033889. https://doi.org/10.1029/2022WR033889

Line 140. The vegetation correction for biomass is still an active area of research so would suggest a little more discussion here, particularly as its uncertainty is identified as being significant later in the manuscript. Hawdon et al. 2014 found a linear reduction in count rate at low biomass but it became more nonlinear for higher biomass and with sites with forest canopy. Franz 2015 found a linear reduction in count rate around 1% per kg/m^2 for croplands which is similar as reported by Baatz 2015. It seems there is a geometric effect for sites with a clumpy distribution of water, but I am not sure it is fully resolved yet (Franz et al 2013 provided some MCNPx simulations as well as Andreasen 2016, 2017). I am cautious about using the Baatz 2015 empirical equation for all vegetation types especially at high biomass sites within forests where this is likely a geometric factor reducing the impact of increasing biomass on the reduction in neutron counts (i.e. maybe a 0.3 or 0.5% in count rate per kg/m^2). I think so additional discussion of this effect some be added especially as it impacts the main conclusions of the paper.  

Hawdon A, McJannet D, Wallace J (2014) Calibration and correction procedures for cosmic-ray neutron soil moisture probes located across Australia. Water Resources Research, 50(6):5029–5043. https://doi.org/10.1002/2013wr015138

Franz TE, Wang T, Avery W, Finkenbiner C, Brocca L (2015) Combined analysis of soil moisture measurements from roving and fixed cosmic ray neutron probes for multiscale real-time monitoring. Geophysical Research Letters, 42https://doi.org/10.1002/2015GL063963

Franz TE, Zreda M, Rosolem R, Hornbuckle BK, Irvin SL, Adams H, Kolb TE, Zweck C, Shuttleworth WJ (2013) Ecosystem-scale measurements of biomass water using cosmic ray neutrons. Geophysical Research Letters, 40(15)https://doi.org/10.1002/grl.50791

Andreasen M, Jensen KH, Desilets D, Zreda M, Bogena HR, Looms MC (2017) Cosmic-ray neutron transport at a forest field site: the sensitivity to various environmental conditions with focus on biomass and canopy interception. Hydrology and Earth System Sciences, 21(4):1875–1894.

L238. Avery 2016 also provided an uncertainty analysis and provided a CONUS map of soil properties for use with CRNS rovers and compared local sampling vs. available continuous datasets.

Avery WA, Finkenbiner C, Franz TE, Wang T, Nguy-Robertson AL, Suyker A, Arkebauer T, Muñoz-Arriola F (2016) Incorporation of globally available datasets into the roving cosmic-ray neutron probe method for estimating field-scale soil water content. Hydrology and Earth System Sciences, 20(9):3859–3872. https://doi.org/10.5194/hess-20-3859-2016

In general: Crow 2012 describes the relationship between average soil moisture and its variance at different spatial aggregations. This relationship is asymmetric but parabolic shaped. Meaning that you would expect the highest standard deviation at intermediate soil moisture and low SD at low soil moisture and intermediate SD at high soil moistures. This information could be included in the expected range of SD across average soil moisture. I am not sure how this physical constraint would affect the local vs. general calibration suggestions. The same is also true for the bulk density and LW due to textural differences within a CRNS footprint. The point is there is both measurement error due to the instrument/method and natural variation due to spatial variability with a CRNS footprint.

Crow WT, Berg AA, Cosh MH, Loew A, Mohanty BP, Panciera R, Rosnay P de, Ryu D, Walker JP (2012) Upscaling Sparse Ground-Based Soil Moisture Observations For The Validation Of Coarse-Resolution Satellite Soil Moisture Products. Reviews of Geophysics, 50https://doi.org/10.1029/2011rg000372

Review

Technical Note: revisiting the general calibration of cosmic-ray neutron sensors to estimate soil water content

Maik Heistermann , Till Francke, Martin Schrön, and Sascha E. Oswald

General Comments

This is a nice exploration of the relative merits of local and general calibration approaches for cosmic ray neutron sensors. Other authors have made similar attempts to undertake generalised calibrations and reveal a single value of N0 but the approach presented here seems robust and draws on some large open source datasets. The ability to avoid a reference calibration for soil moisture will appeal in many circumstances. The revelation that a local calibration may not be the best option in some circumstances is a key finding. This is particularly relevant to low biomass sites. The drawback, and point that might restrict uptake more generally, is how best to specify errors, particularly around biomass and bulk density.

The paper is well written, and the figures are of a high quality. I have made only minor suggestions in the section below. The online calculator is a nice touch which really helps to grasp the concepts and explore the options. I have no problem with this paper being published with only minor modifications.

Specific Comments

Line 60: Sentence is a bit clunky - suggest reword to “This requires the user to account for the relative sensitivity of the neutron detector, the effects of other hydrogen pools in the sensor footprint, and the effects of geographic latitude, longitude, and altitude”

Line 130: An alternative method for correcting for spatial and altitudinal variability has been published by McJannet and Desilets and it may be useful here as it will allow user to apply even more widely to areas outside of Europe.

McJannet DL, Desilets D (2023) Incoming Neutron Flux Corrections for Cosmic-Ray Soil and Snow Sensors Using the Global Neutron Monitor Network. Water Resources Research, 59(4):e2022WR033889. https://doi.org/10.1029/2022WR033889

Line 145 suggest changing to “For our reference detector we chose a so called…”

Line 334: suggest “Altogether, we assume that we have considered the most relevant processes and variables…”

Line 335: suggest “On average, the general calibration function fits the weighted average of locally measured SWC per CRNS footprint fairly well, specifically for the apparent gravimetric SWC”

Line 340: suggest “Interestingly, though, we only become aware of these errors in the case where we apply a general calibration…”

Line 344: suggest “Based on the results of this study, we recommend considering both calibration options, local and general, and weighing the relative uncertainty of the one against the other”