An in-situ methodology to separate the contribution of soil water content and salinity to EMI-based soil electrical conductivity

Autovino, Dario; Coppola, Antonio; De Mascellis, Roberto; Farzamian, Mohammad; Basile, Angelo

doi:https://doi.org/10.5194/egusphere-2025-2696

Preprints

https://doi.org/10.5194/egusphere-2025-2696

Preprints

25 Jun 2025

| 25 Jun 2025

Status: this preprint is open for discussion and under review for SOIL (SOIL).

An in-situ methodology to separate the contribution of soil water content and salinity to EMI-based soil electrical conductivity

Dario Autovino, Antonio Coppola, Roberto De Mascellis, Mohammad Farzamian, and Angelo Basile

Abstract. Salt accumulation in the root zone limits agricultural productivity and can eventually lead to land abandonment. Therefore, monitoring the spatial distribution of soil water content and solution salinity is crucial for effective land and irrigation management. However, assessing soil water content and salinity at the field scale is often challenging due to the heterogeneity of soil properties.

Electromagnetic induction (EMI) offers a fast, non-invasive, in situ geophysical method to map spatial variability in soil. EMI instruments measure the apparent soil electrical conductivity (EC_a), which reflects the integrated contribution of the bulk electrical conductivity (σ_b) of different soil layers. By inverting the measured EC_a, it is possible to obtain the distribution of the σ_b along the soil profile, which provides indirect information on soil salinity. However, in saline soils, σ_b is influenced by both water content (θ) and soil solution electrical conductivity (σ_w) (the salinity), making it difficult to independently quantify these two variables through a single, straightforward procedure.

The objective of this study is to separate the respective contributions of θ and σ_w to σ_b, as obtained from the EMI inversion. To achieve this, EC_a was measured using a CMD-MiniExplorer instrument in two maize plots irrigated with saline and non-saline water, respectively, in an agricultural field in southern Italy. The dataset was then inverted in order to obtain the σ_b distribution. By employing a site-specific calibrated Rhoades linear model and assuming homogeneity between the two plots, the spatial distribution of θ and σ_w in the saline plot was successfully estimated. To validate the results, independent measurements of soil water content by Time Domain Reflectometry (TDR) and direct measurement of soil solution electrical conductivity, σ_w, were performed.

The proposed procedure enables the estimation of θ and σ_w with high accuracy along the soil profile, except in the soil surface, where EMI reliability is limited. These findings demonstrate that the integration of EMI with a site-specific θ - σ_b - σ_w model is a reliable and efficient in-situ approach for mapping soil salinity and water content at field scale, offering valuable insights for optimizing agricultural irrigation management in systems using saline water.

Received: 06 Jun 2025 – Discussion started: 25 Jun 2025

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Dario Autovino, Antonio Coppola, Roberto De Mascellis, Mohammad Farzamian, and Angelo Basile

Status: open (until 09 Oct 2025)

Post a comment Subscribe to comment alert

RC1:
'Comment on egusphere-2025-2696', Anonymous Referee #1, 08 Jul 2025 reply

Review of manuscript titled “An in-situ methodology to separate the contribution of soil water
content and salinity to EMI-based soil electrical conductivity” by Autovino et al.
The manuscript is interesting, aimed to partition ECa surveyed data and untangle the influence of soil moisture and soil salinity on bulk EC. I believe the manuscript does not require significant revision. Nevertheless, the following issues should be addressed:
L22-23: It seems contradicting having an assumption that field plots are homogeneous and then consider the spatial distribution of properties within.
L43: The most common field method…
L52: Please consider the use of the word However at the beginning of the sentence. It seems misplaced there.
L69: ECa (subscript a).
L91: Please change meters to m. Also in L105-115. There is no reason for writing units in full.
L91-93: Which year? I don’t think you give it anywhere.
L97: Please understand that EC of 1.6 dS/m can hardly be referred to as non-saline. What was the source of this water?
L99-102: When? How frequently?
L118: Please add ring sizes and reasons for collecting undisturbed samples.
L157: Above, there is no information about collection of disturbed samples, only undisturbed. As I believe the authors did not make the effort to collect soil cores only to destroy them later, some additional information on soil collection seems to be missing in section 2.1.
L193: Zc (subscript c).
L279: Please revise. Everything else seems to be given in dS/m.
L438: Present.

Reply

Citation: https://doi.org/10.5194/egusphere-2025-2696-RC1
- AC1:
  'Reply on RC1', Dario Autovino, 17 Jul 2025 reply
  Review of manuscript titled “An in-situ methodology to separate the contribution of soil water content and salinity to EMI-based soil electrical conductivity” by Autovino et al.
  The manuscript is interesting, aimed to partition ECa surveyed data and untangle the influence of soil moisture and soil salinity on bulk EC. I believe the manuscript does not require significant revision. Nevertheless, the following issues should be addressed:
  Reply: We thank the referee for their positive and constructive review of our manuscript. Below we address each of the specific comments in detail.
  
  L22-23: It seems contradicting having an assumption that field plots are homogeneous and then consider the spatial distribution of properties within.
  Reply: Although the concepts of heterogeneity and spatial variability may appear contradictory, they are not. Heterogeneity refers to distinct or abrupt differences in soil classification or morphology within a given area. In contrast, spatial variability describes the continuous, often subtle variations in soil properties (e.g., pH, salinity, moisture content) that occur within or between otherwise homogeneous units. While spatial variability is intrinsic to all soils - even at fine spatial scales - our case study defines homogeneity as the overall similarity in soil type, classification, and horizon depth between saline and non-saline plots. To avoid potential confusion, we have adopted the term pedological homogeneity to emphasize this classification-based similarity.
  
  L43: The most common field method…
  Reply: We agree with the reviewer. In the revised version of the manuscript, we will modify the sentence to begin with “The most common field method to evaluate…"
  
  L52: Please consider the use of the word However at the beginning of the sentence. It seems misplaced there.
  Reply: We agree with the reviewer and will make this change in the revised version.
  
  L69: ECa (subscript a).
  Reply: We agree with the reviewer and will make this change in the revised version.
  
  L91: Please change meters to m. Also in L105-115. There is no reason for writing units in full.
  Reply: We agree with the reviewer and will make this change in the revised version.
  
  L91-93: Which year? I don’t think you give it anywhere.
  Reply: The experiment was conducted in 2018. This information will be added to the revised manuscript.
  
  L97: Please understand that EC of 1.6 dS/m can hardly be referred to as non-saline. What was the source of this water?
  Reply: The irrigation water with EC = 1.6 dS/m corresponds to the local well water typically used in the study area. We will reword the revised text to more clearly indicate that the "non-saline" plot was irrigated with low-salinity well water without the addition of additional salt.
  
  L99-102: When? How frequently?
  Reply: Leaf water potential was measured nine times during the growing season from June 11 to July 29. This information will be added to the methodology in the revised manuscript.
  
  L118: Please add ring sizes and reasons for collecting undisturbed samples.
  Reply: We thank the reviewer for noticing the typo. In the revised manuscript, we will clarify that disturbed soil was sampled at the TDR measurement points. These samples were used for laboratory determinations of soil solution salinity (σw) using the 1:2 extraction method.
  
  L157: Above, there is no information about collection of disturbed samples, only undisturbed. As I believe the authors did not make the effort to collect soil cores only to destroy them later, some additional information on soil collection seems to be missing in section 2.1.
  Reply: As reported in the previous point, we will indicate that disturbed soil samples were collected for soil solution extraction and salinity analysis.
  
  L193: Zc (subscript c).
  Reply: We agree with the reviewer and will make this change in the revised version.
  
  L279: Please revise. Everything else seems to be given in dS/m.
  Reply: We will revise the unit presentation in this section and ensure that all electrical conductivity values are uniformly reported in dS/m.
  
  L438: Present.
  Reply: We will revise the verb tenses in this section for consistency with the present tense style used elsewhere in the manuscript.
  
  Reply
  
  Citation: https://doi.org/10.5194/egusphere-2025-2696-AC1
  - RC2: 'Reply on AC1', Anonymous Referee #1, 21 Jul 2025 reply
    
    Thank you for considering my comments. I have nothing else to add.
    
    Reply
    
    Citation: https://doi.org/10.5194/egusphere-2025-2696-RC2
RC3:
'Comment on egusphere-2025-2696', Anonymous Referee #2, 10 Sep 2025 reply
This paper (“An in-situ methodology to separate the contribution of soil water content and salinity to EMI-based soil electrical conductivity”) addresses a critical challenge in agricultural management: monitoring soil water content (θ) and solution salinity (σ_w) at the field scale. While Electromagnetic Induction (EMI) is a non-invasive geophysical method used to map soil spatial variability by measuring apparent soil electrical conductivity (ECa), the bulk electrical conductivity (σ_b) derived from EMI is influenced by both θ and σ_w. This dual dependency makes it difficult to quantify these two variables independently.
The study's primary objective is to develop and validate an EMI-based methodology capable of separating the respective contributions of θ and σ_w to σ_b. To achieve this, the authors conducted an experiment using two adjacent maize plots, one irrigated with saline water and the other with non-saline water. The proposed procedure involves measuring ECa with a CMD-MiniExplorer, inverting the data to obtain the spatial σ_b distribution, and then employing a site-specific calibrated Rhoades linear model, alongside an assumption of homogeneity in water content between the two plots, to estimate the spatial distribution of θ and σ_w in the saline plot. The results indicate that this integrated approach estimates θ and σ_w along the soil profile with reasonable accuracy, except at the immediate soil surface where EMI reliability is limited.
That said, I have several major concerns and questions regarding the methodology, assumptions, and interpretation of results:
Which linear form of Topp’s equation do you use to compare your results to in Fig. 3? It does not seem correct. This is the correct Topp equation: swc = -5.3*10**(-2) + 2.92*10**(-2)*eps - 5.5*10**(-4)*eps**2 + 4.3*10**(-6)*eps**3. The linear form of Ferre is: swc = 0.1181*√(eps)-0.1841. Based on my calculations, the curve (or line in case of linearization) should be around the Ap line on your plot. E.g., for √(eps) = 4, SWC = 0.29 cm3 cm-3 (see figure attached). Please verify, correct and clarify this.

Methodology of lab analyses: both 2.2.1 and 2.2.2 use PVC cylinders with 15 cm height.
2.2.1: Is it common to saturate the soil column from the bottom, hoping for a uniform wetting, instead of packing the soil with pre-homogenized soil? Similar question for the evaporation process, during which the measurements were taken: I would expect the risk of a vertical SWC gradient to be high?

2.2.2: Here you state that each sample was wetted with 15 ml of a solution. How was this done? From the top or bottom? Or was the soil homogenized first and then put in the PVC cylinder? The latter one would be the optimal approach I think, since top or bottom wetting may result in non-uniform distribution of water and salinity.

θ(σ_b) relation

Fig. 4: the data range is very limited (0.3-0.45 cm3 cm-3); do you think these ranges are sufficient? Do you assume that you can extrapolate the linear relation?
Why can you assume that this relation would be linear? I would expect a sigmoidal/concave upward curve, starting flat at low θ, rising sharply, then flattening again near saturation. You can see this also in your Eq. 1, if you write theta in function of σ_b, in principle you assume that the ratio of σ_s and σ_w is constant and independent of theta, which is not the case I think.

This range is in contrast with your experiment description in 2.2.2, where you mention SWCs of 0.06 to 0.46 cm3 cm-3.

In Fig. 7, you estimated SWC based on this relation, and here you get a SWC range of 0.19-0.25 cm3 cm-3; which was not in the range of the soil-specific θ(σ_b) calibration relation. How can you be sure this is correct?

--> I would propose to redo the experiment to know the relation over a wider range of SWCs.

Some smaller comments or questions that I have after reading the manuscript:
L56: abbreviate TDR

L75: over time and space

L80-82: add comma or start new sentence

L111: TDR with abbreviation

You mention taking undisturbed soil samples in the saline plot – what happened with them? Later you also mention soil samples from the non-saline plot? Clarify this.

Figure 1: Improve this figure.

Add subplot letters

Plant row is not visible

North arrow unclear what it belongs to. If it belongs to the map, improve the map (lat, lon) or don’t show Italy at all. The Wikipedia reference is not optimal.

σ_w measurement?

TDR measurement are not on here?

Why only measure 1-2 rows?

Why no σ_w or TDR measurements in non-saline plot?

L184: Clarify that Tektronix = TDR

The steps are not very clear and it’s confusing that they are not in the same order in the results section. Also, step 6 is spread over 3 parts. Please improve structure.

It is unclear how you calibrated or fitted the parameters in Table 1: is this based on a least squares method? Also, the last paragraph of 2.2.3 is quite vague and is based on reading another paper (“In order to calibrate the model for deriving the soil-specific a, b and σs parameters, the procedure reported in Malicki and Walczak (1999) was applied, by using the same experiment reported therein at the point 2.2.2. Finally, the obtained θ - σb - σw data were fitted to the Rhoades model to finalize the calibration procedure.”)

L239: Check sentence.

L242-243: R² is sensitive to the range of the data!

Figure 6

Add subplot letters. C is not in the caption.

What are the lines in Fig. 6a-b? They don’t seem to fit the data points – is this the result of a moving average filter? Clarify in the text and caption.

Fig. 6c: You have ECa measurements at 30 cm depth; how does the model extrapolate to the top 20 cm? You have ECa measurements at 1.2 m depth, why does the inverse model results only go up to 1.05 m depth?

Consider using a different colormap; this one is not greyscale (printer) friendly.

You say the 0-30 cm layer is influenced by surface drip irrigation, while also the 30-80 cm is ‘directly wetted by drip irrigation’. At what depth was the drip line?

Be consistent in the plots: horizontal position of 0-16 m (Fig. 6) or 1-17 m (Fig. 7)

Fig. 8: Add subplot letters.

L347: Fig. 9, not 10

L349 & Fig. 9: You applied a moving average filter to the sigma_w-SS measurements; which window size did you use and why? Is this necessary? It feels like you might be artificially changing measurement data.

L360: 0.23 is not in the figure? I read 0.10

L376: Variance of wSS does not increase? It even slightly decreases with depth. (Table 2)

Check consistency in symbols! (sigma_w,SS or sigma_w-SS, sigma_w^SS-f)

Fig. 10: incomplete legend (thick solid line, thin solid line), unclear what are the two theta-EMI, the two X’s, not all plots have both?

L402: Check sentence.

L467: Check sentence.

Overall, the manuscript addresses an important problem and proposes a creative approach, but at present the methodology and interpretation are not sufficiently clear or convincing. In particular, the assumptions about the θ–σb relation, the limited calibration range, and the lack of detail in sample preparation raise doubts about the robustness of the conclusions.
Furthermore, the practical applicability of the method seems constrained: it requires a non-saline plot immediately adjacent to a saline one, with identical soil properties and moisture conditions; a situation that is rarely feasible in practice.
I encourage the authors to clarify the methodology, provide additional experimental data across a broader SWC range, and carefully reconsider the assumptions that are made. With these improvements, the study could make a valuable contribution to EMI-based soil monitoring.

Reply
Citation: https://doi.org/10.5194/egusphere-2025-2696-RC3

Dario Autovino, Antonio Coppola, Roberto De Mascellis, Mohammad Farzamian, and Angelo Basile

Viewed

Total article views: 546 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
471	56	19	546	14	32

HTML: 471
PDF: 56
XML: 19
Total: 546
BibTeX: 14
EndNote: 32

Views and downloads (calculated since 25 Jun 2025)

Month	HTML	PDF	XML	Total
Jun 2025	91	14	3	108
Jul 2025	111	27	12	150
Aug 2025	109	7	3	119
Sep 2025	160	8	1	169

Cumulative views and downloads (calculated since 25 Jun 2025)

Month	HTML	PDF	XML	Total
Jun 2025	91	14	3	108
Jul 2025	111	27	12	150
Aug 2025	109	7	3	119
Sep 2025	160	8	1	169

Viewed (geographical distribution)

Total article views: 541 (including HTML, PDF, and XML) Thereof 541 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 15 Sep 2025

Short summary

In this article, we developed a method to better understand how soil water moisture and salt content affect electrical signals measured from the surface by electromagnetic induction technique. This helps farmers manage irrigation, especially in areas using salty water. By combining field and lab data, we could tell how much each factor—water or salt—affected the signal. This technique offers a faster, easier way to track soil health and could improve how we use water in farming.


Total:	0
HTML:	0
PDF:	0
XML:	0