the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 29 Sep 2025
Moisture storage, transport and mixing processes after sprinkler irrigation of Pasture cropland: Understanding based on water isotopes
Abstract. The processes of water storage, migration, and mixing in agricultural fields are influenced by a combination of factors, including climatic conditions, soil properties, cropping structure, and field management practices. Sprinkler irrigation is a widely adopted method in agricultural fields globally. Studying the post-irrigation processes of sprinkler irrigation in specific regions can provide valuable insights for regional agricultural development and the conservation and utilization of water resources. In this study, we investigated the water storage, migration, and mixing processes in vegetation within arid irrigated areas. This was achieved by analyzing stable isotope data, using sprinkler-irrigated pastureland (alfalfa) as the research subject. The study results indicated that: (1) there was significant isotope depletion in soil moisture following irrigation, with soil moisture and isotope characteristics returning to their pre-irrigation state after an average of 9 days; (2) water transport in the soil was predominantly vertical, with a minimal proportion of horizontal movement; and (3) evaporation losses due to sprinkler irrigation accounted for 32 %, while losses from excess irrigation (infiltration into soil layers below 60 cm) comprised 5 %. In arid regions, sprinkler irrigation effectively controls infiltration losses; however, evaporation losses remain considerably high. We recommend promoting low-level multipoint sprinkler irrigation and nighttime irrigation practices to enhance water use efficiency and ensure agricultural sustainability.
- Preprint
(1181 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (extended)
- RC1: 'Comment on egusphere-2025-4004', Anonymous Referee #1, 05 Jan 2026 reply
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 316 | 51 | 23 | 390 | 16 | 16 |
- HTML: 316
- PDF: 51
- XML: 23
- Total: 390
- BibTeX: 16
- EndNote: 16
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This preprint uses stable water isotopes (δ2H, δ18O, lc‑excess) together with soil moisture observations to infer post‑sprinkler changes in soil water storage, transport, and mixing in an arid, sprinkler‑irrigated alfalfa pasture in the Jingdian Irrigation Area. The topic is relevant to Biogeosciences, and a well-documented isotope/soil‑moisture dataset from an arid irrigated system could be valuable; however, as written, the core quantitative results are not reproducible, and several headline conclusions (notably the 32% evaporative loss, 5% deep loss, “9‑day recovery,” and the vertical/horizontal transport interpretation) are not convincingly supported by the methods and data presented. Substantial revision is needed before the manuscript can be evaluated on its scientific merits.
A recurring problem is that key quantities are either (1) reported with unclear or implausible units, (2) derived using equations that are dimensionally inconsistent, or (3) asserted without showing the intermediate calculations and assumptions. In an open, permanently archived discussion format, I would strongly encourage the authors to make the analysis auditable: provide the raw data (or a repository link), define the sampling units and timeline, and show the calculations that lead to the main percentages and process interpretations.
Specific comments (scientific issues)
1. Sampling design and timeline are not clearly described, and they conflict with several conclusions. The Methods state that soil sampling occurred “once before sprinkler irrigation and for five consecutive days afterwards” at 10 depth intervals (0–100 cm), with four parallels per layer (Section 2.2.1). Yet the Results emphasize seasonal “monthly variation” plots (Figs. 2, 5, 7) and repeatedly claim that soil moisture and isotopic characteristics return to pre‑irrigation values after ~9 days. As written, these elements do not fit together. There needs to be a clear timeline table/figure listing (1) irrigation event dates, durations, and applied depths; (2) soil sampling dates relative to each event; (3) how the monthly plots were constructed (what dates go into each month; are these event‑based composites?). Without this, “9 days” and several other time‑based interpretations are not checkable.
2. Table 1 “Soil moisture (%)” values are implausible (or mislabeled), and this undermines the storage calculations. Table 1 reports soil moisture near 85–91% at essentially all depths (0–100 cm). For sandy loam/light loam mineral soils, that is not credible as gravimetric water content, and it is physically impossible as volumetric water content. This points to a unit/definition error (e.g., % of field capacity, or a decimal point shift).
It is important to note that soil water storage (Eq. 4–5; Fig. 5; Section 3.2) depends directly on these values. If the moisture values are not what the table label implies, the storage results cannot be interpreted.
The authors need to define moisture precisely (gravimetric vs volumetric vs relative), provide bulk density values by depth (measured but not reported), and ensure Table 1 aligns with the storage values in Fig. 5.
3. Equations (4)–(5) are dimensionally inconsistent as written; bulk density is missing; storage cannot be reproduced. Eq. (5) defines W with a “×100%” factor (i.e., W is a percent), but Eq. (4) uses S = R × W × H × 10 without dividing by 100. Either the “×100%” is wrong, or Eq. (4) is missing a conversion. In addition, R (bulk density) is required but never reported.
The authors would have to fix Eq. (4)–(5) so units are explicit and consistent, and report R (bulk density) by depth so storage values can be verified.
4. Table 2 contains internal statistical inconsistencies (at least one SD is mathematically impossible). Table 2 contains internal statistical inconsistencies (at least one SD is mathematically impossible). For irrigation water δ2H, Table 2 reports Max = −53.47, Min = −72.80, Mean = −66.87, SD = 16.80. Given the stated min–max range (19.33‰), an SD of 16.80‰ is not just “large”; it is inconsistent with the range (i.e., it cannot occur for any dataset bounded by those min/max values under standard SD definitions).
The authors should re-check Table 2 calculations and transcription; report N for each water type and for each soil depth bin; and clarify whether statistics pool both years, multiple events, etc. Endmember statistics must be correct because they propagate into the mixing/infiltration calculations.
5. Headline loss fractions (32% evaporation; 5% “excess irrigation” below 60 cm) are asserted but not derived transparently. The Abstract and Conclusions state evaporation losses of 32% and deep losses of 5%. The manuscript defines PET (Eq. 2) and an isotope mass-balance approach for infiltration (Eq. 6–8), but it does not show how those lead to the reported percentages, nor does it report the necessary bookkeeping terms (irrigation applied in mm, precipitation, Δ storage, ET or evaporation estimates, drainage).
The authors need to present an explicit water balance (with units and uncertainty) that produces these percentages. If the percentages are isotope-derived, show the full calculation chain and assumptions (including how “loss” is defined).
6. The infiltration/mixing calculation (Eq. 6–8) rests on strong assumptions that are not defended, and the reported output (“infiltration into each layer”) is not clearly defined. Eq. (6–8) assumes post‑irrigation water storage in a layer is a two‑component mixture of (i) pre‑irrigation water in that layer and (ii) infiltrated irrigation water, diagnosed from δ18O. In reality, each layer is part of a through‑flow system (water enters from above and exits below), and isotopes are also affected by evaporation and root water uptake over days. Without explicitly restricting the time window or modeling those processes, it is not clear how Eq. (6–8) could uniquely identify “infiltration amount” per layer.
Also, Eq. (8) appears to compute a fraction (Wi/Wa × 100%) rather than an amount (mm), but the text interprets it as a quantification of infiltrated water amount.
The authors need to define exactly what Wi represents (mm, fraction, or both), justify the two‑endmember assumption over the sampling interval, and provide a sensitivity/uncertainty analysis (endmember variability, fractionation effects, etc.).
7. Claims about horizontal transport are not identified by the data presented. The manuscript concludes that transport is “predominantly vertical” with “minimal” horizontal movement, but the sampling design described is vertical profiling at a single location/event set (no lateral transect, no spatial mapping). The discussion in Section 3.3 reads as a conceptual description (supported by a conceptual figure), not an inference constrained by measurements.
The authors need to either remove/soften the horizontal-transport conclusion, or provide spatial sampling/analysis capable of detecting lateral redistribution.
8. Correlations involving lc‑excess are partly tautological, and p-values likely overstate evidence due to non-independence. The manuscript reports significant correlations between SW lc‑excess and δ2H/δ18O (Section 3.1). But lc‑excess is defined as a linear combination of δ2H and δ18O (Eq. 3), so correlation with its components is expected and does not constitute independent process evidence. In addition, pooling depths and times creates strong non‑independence (repeated measures along profiles and through time), so standard p-values are not meaningful unless the sampling unit and degrees of freedom are defined correctly.
The authors need to remove (or reframe) lc‑excess vs δ correlations; relate lc‑excess to independent drivers (RH, VPD, temperature, time since irrigation) using an analysis that respects the repeated‑measures structure (e.g., mixed effects or event-level aggregation).
9. Endmembers for “groundwater mixing” are not adequately reported, making the mixing interpretation unsupported. The text argues that 0–60 cm soil water is mainly irrigation water, while 60–100 cm reflects irrigation water mixing with groundwater (Section 3.4). However, groundwater isotope statistics are not presented in Table 2, and no quantitative mixing analysis with uncertainty is shown.
The authors need to report groundwater isotope values (N, mean, SD; and temporal variability if relevant), show them clearly in the isotope plots, and present a quantitative mixing framework if “mixing” is a key conclusion.
10. Isotope laboratory methods and QA/QC are under-described for soil water extracted by vacuum distillation. The manuscript states soil water was extracted using a “low-temperature vacuum condensation extraction system” but then specifies 180°C (Section 2.3). This discrepancy needs clarification. More importantly, laser spectroscopy of extracted soil water can be affected by organics and matrix effects; the manuscript does not describe checks/corrections (standards bracketing, drift, VSMOW–SLAP normalization details beyond Eq. 1, memory correction, organic contamination screening).
The authors need to provide the QA/QC procedures sufficient for readers to trust that the reported shifts (often a few per mil in δ18O) are not methodological artifacts.
Technical corrections (typos, formatting, clarity)
If the authors address the unit inconsistencies, fix the descriptive statistics, clearly document the sampling timeline, and provide a transparent and reproducible derivation of the key loss percentages (with uncertainty), the study could become much more informative. As it stands, the main conclusions outpace what the presented data and analyses can support.