General comments:

This study provides valuable insights on how to optimally merge remotely sensed observations with a widely used land surface model (Noah-MP) using an irrigation scheme to improve simulated irrigation and effected hydrologic states and fluxes. The study concluded that the evaluated data assimilation system is largely affected by errors in simulated irrigation. The authors pose that inclusion of dynamic crop information and the assimilation of backscatter data per orbit can improve the results presented in this study. A very interesting aspect of this study is the assimilation of backscatter (rather than soil moisture) which allows assimilation to be performed in the observation space. It would be helpful for the authors to elaborate on this decision and if they think there would be degraded performance if SM was assimilated instead. Specific comments and questions are included below.

Specific comments:

C1: Were any bias correction methods employed (e.g., LSM calibration)? If not were there any steps taken to check if there are systematic biases between the observed and modeled backscatter that could violate the Kalman filter assumption (i.e., line 236). It seems it would be useful to run an open loop Noah-MP simulation forced with precipitation + known irrigation to see if simulated backscatter errors are truly random relative to the observations.

C2: The relationship between Noah-MP simulated soil moisture and vegetation with the assimilated variable, backscatter, is vitally important to this analysis. It would be very beneficial to include equations that show how backscatter is related to these variables, and then how the assimilation is used to ‘correct’ each state. What assumptions are made within these steps that can affect irrigation estimates?

C3: The EnKF is a commonly used data assimilation algorithm and certainly has proven useful. However, from a mass-balance perspective, particle assimilation algorithms (e.g., Abolafia-Rosenzweig et al., 2019) may be more appropriate. For instance, in particle DA algorithms, all model states are corrected in a physically consistent manner (e.g., rather than choosing to only update surface soil moisture or empirically decide how to update states and fluxes related to the observation). Can you discuss why the EnKF was used and potential limitations of this data assimilation strategy in the context of irrigation quantification and simulating irrigation signals? In future steps that seek to employ the lessons of this study, considering other DA algorithms can also be beneficial.

Reference:

Abolafia‐Rosenzweig, R., Livneh, B., Small, E.E., Kumar, S.V., 2019. Soil Moisture Data Assimilation to Estimate Irrigation Water Use. J. Adv. Model. Earth Syst. 11, 3670–3690. https://doi.org/10.1029/2019MS001797

C4: The timing of irrigation (e.g., continuous vs. applied only during morning hours) can greatly affect the amount of irrigation required to achieve a specified (or observed) soil and vegetation moistness. Is the irrigation timing assumed from Noah-MP reasonable, or is this likely to introduce errors? If so, are ‘corrected’ errors from DA a sign of skill or are they compensating for other errors?

C5: What is the footprint of irrigation at the study sites relative to the observed footprint? How could this affect the amount of information provided to the LSM via observations?

C6: Why use ASCAT to evaluate Noah-MP surface soil moisture instead of finer resolution data such as SMAP-S1 (which has been shown to have irrigation signals in Jalilvand et al., 2021) or SMAP which was shown to have irrigation signals in Lawston et al. (2017) and provide more reliable data than ASCAT (Kumar et al., 2018)?

References:

Kumar, S.V., Dirmeyer, P.A., Peters-Lidard, C.D., Bindlish, R., Bolten, J., 2018. Information theoretic evaluation of satellite soil moisture retrievals. Remote Sens. Environ. 204, 392–400. https://doi.org/10.1016/j.rse.2017.10.016

Kumar, S.V., Peters-Lidard, C.D., Santanello, J.A., Reichle, R.H., Draper, C.S., Koster, R.D., Nearing, G., Jasinski, M.F., 2015. Evaluating the utility of satellite soil moisture retrievals over irrigated areas and the ability of land data assimilation methods to correct for unmodeled processes. Hydrol. Earth Syst. Sci. 19, 4463–4478. https://doi.org/10.5194/hess-19-4463-2015

Lawston, P.M., Santanello, J.A., Kumar, S.V., 2017. Irrigation Signals Detected From SMAP Soil Moisture Retrievals: Irrigation Signals Detected From SMAP. Geophys. Res. Lett. 44, 11,860-11,867. https://doi.org/10.1002/2017GL075733

C7: The paragraph from lines 75-86 (or the following paragraph) could benefit from discussion of Abolafia-Rosenzweig et al. (2019) which designed a system to assimilation remotely sensed soil moisture with land surface models to quantify irrigation water use as well as Jalilvand et al. (2021) which compliments Lawston et al. (2017) by evaluating irrigation signals from SMAP-S1 soil moisture retrievals (i.e., from Das et al., 2019).

References:

Abolafia‐Rosenzweig, R., Livneh, B., Small, E.E., Kumar, S.V., 2019. Soil Moisture Data Assimilation to Estimate Irrigation Water Use. J. Adv. Model. Earth Syst. 11, 3670–3690. https://doi.org/10.1029/2019MS001797

Das, N.N., Entekhabi, D., Dunbar, R.S., Chaubell, M.J., Colliander, A., Yueh, S., Jagdhuber, T., Chen, F., Crow, W., O’Neill, P.E., Walker, J.P., Berg, A., Bosch, D.D., Caldwell, T., Cosh, M.H., Collins, C.H., Lopez-Baeza, E., Thibeault, M., 2019. The SMAP and Copernicus Sentinel 1A/B microwave active-passive high resolution surface soil moisture product. Remote Sens. Environ. 233, 111380. https://doi.org/10.1016/j.rse.2019.111380

Jalilvand, E., Abolafia-Rosenzweig, R., Tajrishy, M., Das, N.N., 2021. Evaluation of SMAP-Sentinel1 High-Resolution Soil Moisture Data to Detect Irrigation over Agricultural Domain. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 1–1. https://doi.org/10.1109/JSTARS.2021.3119228

C8: Please reference the following when introducing NASA’s LIS:

Kumar, S., Peters-Lidard, C., Tian, Y., Houser, P., Geiger, J., Olden, S., Lighty, L., Eastman, J., Doty, B., Dirmeyer, P. Land information system: An interoperable framework for high resolution land surface modeling. Environmental Modelling & Software 21, 1402–1415. https://doi.org/10.1016/j.envsoft.2005.07.004 (2006).

Peters-Lidard, C.D., Houser, P.R., Tian, Y., Kumar, S.V., Geiger, J., Olden, S., Lighty, L., Doty, B., Dirmeyer, P., Adams, J., Mitchell, K., Wood, E.F., Sheffield, J. High-performance Earth system modeling with NASA/GSFC’s Land Information System. Innov. Syst. Softw. Eng. 3, 157–165. https://doi.org/10.1007/s11334-007-0028-x (2007).

General comments:

The study idea is to explore the possibility of improving the irrigation water use simulation by direct assimilation of sentinel1 backscatter in co- or cross-polarization, which contains both soil moisture and vegetation information, with the Noah MP land surface model. The results suggested that assimilating Sentinel 1 backscatters data can slightly improve irrigation simulation over some test sites (especially the VH polarization DA). Still, poor parametrization of the Noah-MP irrigation module does not allow the DA to improve the irrigation simulation significantly. This study and the previous study (Modanesi et al., 2021) provide valuable insights into the limitations and benefits of assimilating Sentinel-1 backscatter with the land surface model for improving irrigation simulation. However, I have some concerns regarding the improvement in accumulated irrigation after DA, the spatial mismatch between the model and the test sites scale, and the accuracy of the benchmark datasets used for the validation. Please see my comments for details.

Specific comments:

1- L65: I think studies that focused on calculating the Evapotranspiration through the energy balance algorithm should also be mentioned here as examples for consumptive water use estimation using optical and thermal sensors.

2- L77: Consider the following study along with Lawston et al., 2017 that shows the more recent and high-resolution SMAP-Sentinel1 SM product also contains the irrigation signal.

Jalilvand, R. Abolafia-Rosenzweig, M. Tajrishy, and N. N. Das, "Evaluation of SMAP/Sentinel 1 High-Resolution Soil Moisture Data to Detect Irrigation Over Agricultural Domain," in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 14, pp. 10733-10747, 2021, DOI: 10.1109/JSTARS.2021.3119228.

3- L97: The irrigation module of the Noah-MP model calculates the ideal IRR needed for the crop to avoid water stress which is different from the actual irrigation (the farmer might over or under irrigate the fields). How do you account for that?

4- L179: Here, you are talking about the time and location of irrigation. I think GVF looks a little out of context here; some explanation regarding where GVF is used in the Noah MP model is needed.

5- L250: typo, de Kalman should be changed to the Kalman

6- L378: [Major] I can't entirely agree with this statement that the accumulated irrigation has improved compared to the OL run. Looking closely at Figure 6e, the DA underestimation of irrigation during the mid-summer months of 2015 and 2017 resulted in the overall lower accumulated irrigation (the OL run simulation during the same period closely matched the observed irrigation). In other words, the underestimation during these months compensated for overestimations in other months (e.g., the late summer months of 2016 and 2017), and the right result is obtained here for the wrong reasons! Please comment on this.

7- L383: The most considerable overestimation by the DA run relative to the observed irrigation occurred in July 2016 (Figure 6e), which is right after a significant precipitation underestimation by MERRA2. This contrasts with what is mentioned at the end of this paragraph.

8- L385: The size of the Budrio site is much smaller than your benchmark soil moisture product spatial resolution (ASCAT 12.5 km); the other Italian site or the German site would be a better choice for the SSM time series comparison shown in this figure.

9- Figure 6) It is difficult to compare the 3 time series in Figure 6 as it shows 3 years of data. As the study focuses on irrigation, adding an inset (or possibly another figure) that focuses on one irrigation season can give the readers a better idea of how DA improves or degrades different parameter simulations during the irrigation season.

10 - L425 and L478: The same result is reported on the benefits of LAI DA relative to the SSM DA in this very recent study by Nie et al. 2022, which can be discussed here.

Nie, W., Kumar, S. V., Arsenault, K. R., Peters-Lidard, C. D., Mladenova, I. E., Bergaoui, K., Hazra, A., Zaitchik, . F., Mahanama, S. P., McDonnell, R., Mocko, D. M., and Navari, M.: Towards Effective Drought Monitoring in the Middle East and North Africa (MENA) Region: Implications from Assimilating Leaf Area Index and Soil Moisture into the Noah-MP Land Surface Model for Morocco, Hydrol. Earth Syst. Sci. Discuss. [preprint], https://doi.org/10.5194/hess-2021-263, in review, 2021.