General Assessment

The manuscript is well written, logically structured, and makes an important contribution by comparing AquaCrop (a crop model) and Noah-MP (a land surface model, LSM). This work effectively bridges the crop water and land surface modeling communities, providing valuable insights into irrigation, soil moisture, vegetation, and ET representation at both basin and field scales. The use of multiple observational and satellite products adds robustness. However, some methodological and interpretational aspects require clarification and further depth.

Major Comments

1. Section 4.4 – Uncertainties in satellite-based retrievals: The treatment of uncertainties is currently too brief. Methodological uncertainties should be discussed in more detail:

• Soil moisture representation: The use of surface soil moisture (SSM) is problematic for irrigation studies, since what is required is modeling of root-zone soil moisture (RZSM) and the management allowable depletion (MA) in irrigation modules to trigger irrigation. This mismatch and the limitation of SSM in representing RZSM should be explicitly acknowledged.
• Downscaling errors: Both SenET-derived ET and downscaled SMAP SSM rely on downscaling methods, which propagate errors into the reference datasets and thus into the model evaluation. It may be more consistent to use an ET reference dataset at a coarser resolution that avoids such downscaling errors.

• Implications: These uncertainties influence not only irrigation assessment but also the reliability of the ET and SSM benchmarks used in the study. A deeper discussion would strengthen the credibility of the conclusions.

2. Temporal aggregation of evaluation metrics: The aggregation of daily values into 15-day means is understandable to reduce the impact of uncertain irrigation timing, but it introduces serious limitations:

• Loss of dynamics: Short-term variability in ET and SSM, especially during crop water stress, is smoothed out. This may reduce the evaluation to potential ET comparisons.
• Bias masking: Aggregation may artificially improve correlations while masking systematic model deficiencies in simulating fast drying, irrigation events, or recovery after stress.

The performance should be quantified at multiple aggregation scales (e.g., daily, weekly, bi-weekly), which would also be more coherent with irrigation comparison at different temporal scales. Discuss implications for agricultural applications, where short-term stress is critical.

3. Guidance for the irrigation modeling community: Given the comparative scope, the study should conclude with more explicit guidance:

• How to reconcile crop models’ explicit treatment of management practices with generalized LSM schemes.
• Implications of shared assumptions (e.g., optimal irrigation rules at ~50% TAW) for large-scale irrigation simulations. Moreover, in the case of sprinkler irrigation, the adoption of such low MA thresholds, essentially at the onset of water stress, appears unjustified, since sprinkler systems enable the frequent application of small water amounts, as evidenced in Fig. 8d.
• Pathways to integrate crop model process detail into LSMs.

Minor comments.

Model inconsistency in vegetation vs ET (Fig. 6): The observation that AquaCrop overestimates vegetation (positive bias in Fig. 6k) while underestimating ET (Fig. 6l) is surprising, especially in quasi non-stressed conditions (as suggested by the MA threshold and Fig. 8c, where MA remains above 50% of TAW). Although possible causes are discussed in the article, this inconsistency deserves a more in-depth analysis.

L47-49: Right, but irrigation and other agricultural practices significantly alter water, energy, and carbon fluxes. LSMs need to better represent these processes.

L64: The statement ‘irrigation is not an addition to the model to better represent the water balance’ is unclear. In fact, in crop water models irrigation explicitly supplies water to the soil water storage, thereby improving the representation of water balance components (e.g., RZSM, ET) (Olivera-Guerra et al., 2018).”

L66: Beyond their use in estimating net irrigation requirements (i.e., crop water needs), crop models have also been applied to estimate actual irrigation water use under a range of regimes, from deficit to excess irrigation, making them highly relevant for large-scale irrigation assessments (Olivera-Guerra et al., 2023; Laluet et al., 2024).

L341-342: If runoff is not generated after irrigation in AquaCrop, percolation losses should be included explicitly as residuals in the water balance.

L356-359: Dari et al. (2023) account for evaporative losses in potential ET, similar to AquaCrop, which incorporates ET into the daily calculation of irrigation requirements.

L359-360: While assuming a 100% irrigated fraction for irrigated pixels could lead to overestimation of irrigation, this does not align with the reported values. Furthermore, it may be inconsistent with the approach described in L345–L347, where a conveyance efficiency factor is applied to adjust Noah-MP estimates to match management data.

L505-506: The conclusion that both models approximate average irrigation rates after accounting for losses seems overstated, given substantial biases. However, the results do not fully support this assertion. Significant errors arise not only from the accounted losses but also from assumptions inherent to each model, including the overestimation of irrigated areas noted above. Importantly, both models apply the same irrigation rule—filling the soil water storage once a critical soil moisture threshold (~50% of TAW) is reached, effectively simulating optimal irrigation, which constitutes a major source of bias.

L520: Add reference for SM_WP and SM_FC being tuned to obtain accurate ET fluxes.

L522: It is unclear how the use of ensembles can better represent irrigation. Moreover, it feels out of context.

L570-571: Note that irrigation has been evaluated at finer scales (weekly/bi-weekly) by Dari et al. (2022, 2023), Laluet et al. (2024), and Olivera-Guerra et al. (2023).

L575-576: Highlight Olivera-Guerra et al. (2023)’s approach for calibrating temporal dynamics of irrigation at the district scale, which is compatible with large-scale applications using either crop models or land surface models.

This manuscript presents a timely and well-structured comparison of AquaCrop and Noah-MP, making a valuable contribution by bridging crop and land surface modeling perspectives. The study has strong potential, but the conclusions are currently weakened by the use of heavily aggregated (15-day) ET and SSM comparisons and by limited discussion of uncertainties in satellite-based reference datasets. Expanding the uncertainty analysis, exploring multiple aggregation windows, and offering explicit recommendations to the irrigation community would substantially strengthen the paper.

References:

Dari et al. (2022) : https://doi.org/10.1016/j.agwat.2022.107537

Dari et al. (2023) : https://doi.org/10.5194/essd-15-1555-2023

Laluet et al. (2024) : https://doi.org/10.1016/j.agwat.2024.108704

Olivera-Guerra et al. (2018) : https://doi.org/10.1016/j.agwat.2018.06.014

Olivera-Guerra et al. (2023) : https://doi.org/10.1016/j.agwat.2022.108119

General Assessment

This manuscript is well-written, logically structured, and makes an important contribution by comparing AquaCrop and Noah-MP. This work effectively bridges the crop water and land surface modelling communities, providing valuable insights into the representation of irrigation, soil moisture, vegetation, and ET at both basin and field scales. The use of multiple observational and satellite products adds significant robustness. Overall, the paper is of high quality and is generally suitable for publication, requiring minor to moderate revisions focused on strengthening methodological justifications and expanding the interpretation of specific results.

Major comments

Given that Section 3.2 focuses on field-based evaluation, yet Section 4.2 later refers to in situ irrigation data as “sometime unreliable”, the authors must enhance Section 3.2. A detailed overview of the field data used for comparison is needed, specifically detailing measurement or estimation protocols and the associated uncertainties. This justification supports why these field measurements are considered more reliable than the satellite retrievals, especially considering the better correlation observed with the model simulations. Alternatively, if the data reliability, protocols, and uncertainties are described in a published document or paper, a precise citation to that source should be included with a summary in Section 2.3.4

Minor comments

Since the acronym TAW is first introduced in the caption of Figure 1, please provide its full definition and briefly explain its significance upon this first appearance.

For Figures 2 and 3, I recommend employing shared ranges and colorbars (at least per variable) to allow the reader to accurately compare differences between the datasets.

At Line 355, the manuscript states that the Dari's dataset does not include evaporative losses. However, the subsequent methodology utilizes evapotranspiration (ET) rates in the water balance to estimate irrigation. Its implications must be explicitly considered when comparing the results against AquaCrop and Noah-MP.

At Line 381, I suggest including a brief explanation for the decision to evaluate the additional DMP data specifically during the months from January to June.

I think the paper could close with an overview of implications on this paper on the future investigations, regarding the utility of estimating the irrigation and the gap between crop and land surface models.

This manuscript presents a comparison between AquaCrop and Noah-MP within NASA’s LIS framework for irrigation estimation over the Po Valley. The integration of AquaCrop into LIS is very interesting and  innovative contribution, opening new opportunities for the community to apply AquaCrop in regional, high-resolution hydrological contexts.

In my opinion (sorry),  at the same time, several conceptual and methodological aspects limit the scientific strength of the study in its current form.  In my opinion the most critical point is the crop representation strategy, which strongly influences irrigation and land-surface processes.  The use of a generic transplanted crop as the representative vegetation for the Po Valley is not adequately justified. Since AquaCrop is inherently crop-specific and phenology-driven, this choice strongly influences canopy development, rooting depth evolution, timing and magnitude of irrigation. In my opinion, considering that only summer crops (and sometime Alfalfa) are irrigated in Po Valley, my proposal is to use a realistic summer crop (e.g., maize), or justify in detail why this was not feasible.

Re-running simulations with a realistic crop in my opinion can  greatly enhance the quality and the interest of the comparative analysis.

Intermediate-minor concern

Parametrization: The two models rely on different soil hydraulic parameterizations. Many differences in simulated irrigation or soil moisture may therefore reflect parameter space differences rather than model process differences. A clearer discussion would help guide interpretation.

The 45% irrigation threshold, coming from earlier studies, may not be optimal for AquaCrop or for the selected generic crop. Some justification, sensitivity analysis, or recalibration would strengthen the results (note that in several models the threshold is below 33% pf the TAW

The use of GVF and CC thresholds produces growing seasons that do not always correspond to real agricultural calendars in the Po Valley, mainly because there are (I imagine frequently) cases with cells that have inside winter and summer crop. Some problem still can arise when a winter crop like Barley or Triticale (not irrigated) are harvested early for forage and after is followed by maize

The manuscript is very long and at times overly detailed. Several paragraphs could be substantially condensed without loss of information. Some sections (especially the Results) are verbose and occasionally difficult to follow.

In my opinion several graphs (the maps)  are difficult to interpret due to different scales for the same variables. Some caption is too synthetic and full of acronyms, some other is too long (I think that we don’t need the explication of what is a boxplot in the caption, and what is the reason to omit the outliers ?

I also attach the annotated PDF, where I have included several rough (apologies) but hopefully useful comments and margin detailed notes that may help the authors to understand my criticism and to improve the manuscript.

Review on paper « On the gap between crop and land surface models: comparing irrigation and other land surface estimates from AquaCrop and Noah-MP over the Po Valley » by Busschaert et al. submitted for publication in HESS

This study aims to compare the ability of a land surface model and a crop model to predict irrigation at the regional scale and other variables related to the water cycle including soil moisture, evapotranspiration and dry matter production. The paper is well written and well organized and is of interest for the HESS community. Nevertheless, I have some major concerns listed below.

Positioning of the research question

The objective of the study is described as « estimating irrigation at the regional scale » but it is not clear what’s the purpose behind estimating irrigation at the regional scale. If it is for water management, the model should relate the water withdrawal for irrigation to the surface and ground reservoir. If it is for improving the land surface condition for atmospheric models, spatial resolution of 1 km² is probably too fine. It cannot be for irrigation scheduling as the resolution is too coarse … Please elaborate on this point. Still considering the positioning of the research question, what’s the objective of comparing a Land Surface Model to an agronomic model as many LSM, thanks to recent developments, include irrigation modules quite similar in their basic principle to aquacrop (refill the reservoir when available water fall below a given threshold) ? By contrast, I see significant difference in the way the crop growing processes are represented and I was wondering if the contrasting results were not related more to the difference in terms of dynamic vegetation than in terms of irrigation prediction.

Methodological aspects of the comparison

The authors choosed to use the models « out of the box » with their own parameterization and input variables without trying to adjust the inputs and parameterization in order that both models runs in close configuration seems questionnable to me. Is it not an obvious result that the model provide with quite different irrigation estimates considering the large differences in crop phenology, soil moisture dynamic and thus probably root depth dynamic,? It is even not necessary to run the models to show that irrigation estimates with NOAH will be much higher than with Aquacrop. Having a look to the TAW values for both model is sufficient (as written by the authors at L470).

Considering the comparison with the Dari data set, I see a major concern as well. One of the main conclusion of the study is that the fact that both model irrigation estimates tend to correlate better with each other than with the satellite data sets by aggregating at larger time scale (the best correlation is obtained for correlation computed over the whole summer) demonstrate that the model estimate are more consistent than the satellite data set of Dari. To my opinion, it just show that the atmospheric evaporative demand, governed by the forcing meteorological data set that is the same for both model only … I agree that satellite irrigation estimates are really uncertain but uncertainties are high for model estimates as well (forcing, irrigation map, soil characteristics …) and this study does not demonstrate that the model estimate are of better quality for the reasons above. The next section showing the comparison to in situ irrigation is not sufficiently conving by lack of important information. Could you tell the reader a little bit more about the representativity of the in situ irrigation data, we don’t know how the irrigation data were gathered and we don’t even know the crops that were grown on the sample fields. Could you provide these information ?

In addition, the manuscript was at times hard to follow, as (1) important details on model parameterization and validation data are not provided, and simply referring to previous papers is not sufficient. Readers cannot reasonably be expected to spend several days reviewing the literature in order to understand a single study and (2) some choice of model setup appear questionable and are not justified. Finally, there is no clear conclusion apart that more in situ data should be gathered and that the remote sensing community should improve the satellite retrieval of irrigation. Could the authors suggest a few research directions to improve the models?

To my opinion, the paper has potential for publication but I recommend major revision including significant rewritting.

Detailed comments

Both models were run as sprinkler irrigation (meaning that irrigation is added to rainfall, right ?) whereas surface irrigation is the dominant practice in the study area (L96). This inconsistency raises questions about how representative the simulations are. In addition, part of the discussion focuses on canopy interception differences, which are simulated in Noah but not in AquaCrop; however, given that the prevailing irrigation method is surface rather than sprinkler, it is not clear what the added value of this discussion is. This is one the point among others that seems to complicate the message unnecessarily.

L120 : Why a spin up is needed for Aquacrop as the reservoir is refilled every winter ?

It would be interesting to further the comparison with the data set of Dari et al. including the spatial patterns (fig. 2) and on the validation sites (Fig 9, G1, G2)

Legend of figure 1 : explicit SHPs

Table 1 : explicit CC and GVF

L128 : please justify the bilinear interpolation by providing the elevation values of the study region, at least the range of elevation.

L140 : « More specifically, the Noah-MP SHPs are based on Cosby et al. (1984) and further adapted to intentionally increase total available water (TAW; also referred to as plant-available water) in the context of largescale simulations by Chen and Dudhia (2001). » Why is it necessary to intentionally increase TAW in the context of large scale simulations ?

L159 : the rooting  depth not only depends on the geographic location but also on the type of crops. The sentence should be reformulated.

L170 : the choice of C3 is questionable in this region that is the main producer of maize. I understand that it won’t change the conclusion of your study but It doesn’t enhance its credibility from an agronomic point of view either. If you prefer considering C3, wheat is sown in autumn and harvested by the end of June. I understand that for this exercise we’re trying to simplify as much as possible by applying the same crop across the whole study area, but why not, at the very least, choose realistic cropping periods?

L201 : January 1st to the end of March (80 calendar days) is more than 80 days.

L203 : « except for the radiation transfer scheme » please explain

L204 : « The latter enables a dynamic definition of the growing season in line with AquaCrop » At least explain how LAI (which is the prognostic variable, right ?) is computed and the basic principle for the start of the growing season. Is there a seeding date defined ? (January 1st like Aquacrop ?). We finally learn at L402 that harvest is not modeled but what about seeding ? Is there a photosynthesis module and growth start when carbon assimilation is enough to maintain existing tissues and fill some carbon reservoirs ?

L218 : the discretization of the soil  vertical profile is much finer for Aquacrop than for NOAH. How Irrigation applied is computed in Noah ? For instance, if root depth is larger than the third layer, let’s say 65 cm, do you refill the four layer including the last one going down to one meter entirely ?

L230 : « This threshold is defined as 40% » This is the threshold for the start of the growing season or the threshold when irrigation can be triggered ?

2.3.3 Evapotranspiration product / I understand that you were looking for a publicly available data set and you didn’t wanted to recompute an evapotranspiration product on your own. Nevertheless, SenET is based on 1km² surface temperature product derived from Sentinel-3 and downscaled with Sentinel-2 at 100 m and you average again to compare to your 1km grid point …. Please, at least, point out this aberration.

L240 : I don’t understand the first half of this sentence « . The GV F in Noah-MP is based on the dynamically simulated LAI via an exponential relation (Equation 8) whereas the CC in AquaCrop is the driving variable and is based on the climatic and environmental conditions » in Noah-MP, GVF is directly derived from the dynamical variable LAI that depends as well on the climatic and environmental conditions, right ? The message of this sentence is not clear to me.

L245 : « Note that all these reference products have their uncertainties, in particular over irrigated areas » Is this sentence really useful considering that all data are uncertain all the more when considering remote sensing products. Could you explain why the data uncertainties should be higher on irrigated areas ?

L251 : Specify that these are the model estimates.

Fig3. Please use the same color scale

L287 : please explain how these irrigation data were gathered and provide with the crops and irrigation technic of each fields.

L296 : Please provide temporal and spatial resolution of the irrigation data set as well as the time period of availability.

L297 : The first part of the sentence « As the first regional irrigation product for the Po Valley, it is subject to notable uncertainties, especially since » is useless without providing an assessment of this data set.

L335 : please show the spatial patterns of precipitations and radiation. It would be interesting for comparison purpose and they can be easily inserted within an appendix.

L341 : « Noah-MP includes the simulation of interception » is this really problematic as most of irrigated field are through surface irrigation ?

L341 : « Noah-MP includes the simulation of interception and runoff losses that AquaCrop does not simulate, and which are important for sprinkler irrigation ». Why runoff is important for sprinkler irrigation ?

L351 : quite obvious as both models have the same meteorological forcing, right ?

L363-364 : knowing if there have been some administrative order restricting agricultural water use in 2021 could strengthen the credibility of the Dari dataset, which shows a decrease in irrigation water demand in that particular year.

L364 : I would add « and could face water use restriction during drought year »

L389 : « The irrigation practices applied in rice fields differ strongly from the sprinkler irrigation modeled by AquaCrop and Noah-MP » But I though that the main irrigation technics was surface irrigation ?

L403 «The choice of vegetation module options in this study may have resulted » again, this is not a report for the NOAH team but a scientific article for the scientific community. What are the different options ?

L431 « Note also that in Noah-MP, the irrigation water lost through runoff is not compensated, meaning that field capacity is never reached in the model after an irrigation event» It is not clear to me what runoff are you refering. Is it runoff due to infiltration excess ? Meaning that the irrigation rate is too strong with regards to the soil infiltration capacity ? Is it realistic ?

Fig 9 : could you please show the satellite irrigation retrieval as well.

L500 : « Overestimations were also found in other studies » of what by which model ?

L501 : « due to the dominance of fruit trees » meaning that the in situ data are gathered on fruit trees ?

L500-502 : « Overestimations were also found in other studies (Dari et al., 2023; Modanesi et al., 2022) and are likely due to […] localized methods (drip irrigation), which are more efficient than sprinkler irrigation / Probably but it can be simulated by aquacrop through the fw (wet fraction) factor.

L505 : « and application losses that are included in water records » I don’t understand the importance of mentioning the water losses through transportation as you compare your model predictions to in situ irrigation data and satellite retrieval that both represent the water really applied at the field scale.

L570 : « but this only allows to validate coarse-scale (temporal and spatial) irrigation estimates » In relation to my general comment about the positioning of the research question, this could be an acceptable objective but it needs to be explained in the introduction.

L576 : « First, the native spatial resolution of the remote sensing products (relatively coarse compared to fields) complicates the detection of irrigation signals (in the SSM or vegetation; Ozdogan et al., 2010) » I have two comment on this sentence. The size of your grid point is 1km² similar to the resolution of the Dari dataset so probably much higher than the typical field scale as well. Once again, what is the objective of this study ?