Dear William, thank you very much for your review and useful comments.

First, we would like to explain the measurements uncertainty at the validation sites. Indeed, when all input parameters were available, we have computed the energy balance closure (EBC) ratio (i.e., [Rn-G-H]/LE with Rn: net radiation, G: soil heat flux, H: sensible heat flux, LE: latent heat flux) for the EC sites. In total, the EBC values are derived for five flux sites in Italy. The obtained ratios are as follows:

• vineyard at IT-Lsn (1 m a.s.l.): 0.28,
• grassland at IT-MtM (1450 m a.s.l.): 0.08,
• grassland at IT-MtP (1550 m a.s.l.): 0.13,
• evergreen broadleaf forest at IT-SR2 (4 m a.s.l.): 0.73,
• grassland at IT-Tor (2160 m a.s.l.): -0.02.

After contacting providers of daily flux data from University of Ghent in Belgium, we were recommended to include all eight sites in our analysis due to the small number of EC towers over our study areas. We derived the largest ratio over lowland forest at IT-SR2 exceeding 0.7, while EBC values over sloping terrain (i.e., IT-MtM, IT-MTP, and IT-Tor) achieved acceptable scores (i.e., below 0.2). Unfortunately, for the remaining three sites the EBC ratios were not provided (i.e., FR-Pue, IT-Ren, IT-MBo). As can be seen above, in our case, the energy balance closure values do not depend on topographic complexity. On the other side, to make any conclusions in this regard, we truly believe that the impact of complex topography on energy balance closure shall be further investigated by incorporating more flux sites across different landcovers, topographies, and (micro)climates. We contacted PIs of the EC towers to understand if they applied a planar fit method to eddy covariance flux data. We have received their feedback regarding sites over forested landscapes at FR-Pue, IT-Ren and IT-SR2, and high-mountain grasslands located in Aosta Valley (IT-Tor) and Mazia Valley (IT-MtM and IT-MtP). The first two tree covered sites (FR-Pue, IT-Ren) were processed using a double rotation method, while a planar fit method was applied to IT-SR2. Regarding the alpine grasslands, all eddy covariance flux sites were corrected with a planar fit approach. In addition to the research work of Ross and Grant (2015), we have included ancillary studies investigating EC measurement uncertaintities:

1. Castelli, M., Anderson, M.C., Yang, Y., Wohlfahrt, G., Bertoldi, G., Niedrist, G., Hammerle, A., Zhao, P., Zebisch, M. and Notarnicola, C., 2018. Two-source energy balance modeling of evapotranspiration in Alpine grasslands. Remote Sensing of Environment, 209, pp.327-342.
2. Mauder, M., Cuntz, M., Drüe, C., Graf, A., Rebmann, C., Schmid, H.P., Schmidt, M. and Steinbrecher, R., 2013. A strategy for quality and uncertainty assessment of long-term eddy-covariance measurements. Agricultural and Forest Meteorology, 169, pp.122-135.
3. Rannik, Ü., Vesala, T., Peltola, O., Novick, K.A., Aurela, M., Järvi, L., Montagnani, L., Mölder, M., Peichl, M., Pilegaard, K. and Mammarella, I., 2020. Impact of coordinate rotation on eddy covariance fluxes at complex sites. Agricultural and Forest Meteorology, 287, p.107940.

Description on measurements uncertaintities was added in the revised version of the manuscript in the section 2.2 (lines 189-193), and also it has been discussed in the ‘Results and discussion’ (lines 479-482) and in the Conclusions (lines 642-644).

Second, we fully agree with you that more attempts are required to enhance the TSEB performance over heterogenous terrain. This is a part of our current work where we aim to improve the model in the European Alps, including forest sites and areas with complex topography. In this regard, topographically corrected solar radiation and landcover at finer spatial resolutions shall be incorporated to account for heterogenous landscapes and different types of vegetation along with their biophysical characteristics. Moreover, as mentioned by you before, the Priestely-Taylor parameter with a default value of 1.26 needs to be adjusted to take into account variations in green vegetation cover during growing season. All described above improvements are our close-future objectives in order to derive more reasonable results, especially over forests as shown in the suggested reference research studies. To improve our manuscript, in the section 4.1 (lines 485-497) we have added some text where together with the suggested reference papers we describe potential improvements in the TSEB-PT for deriving more robust ET estimates over forests and areas with complex topography. Additionally, in the revised manuscript we included the following research publication on TSEB modelling over boreal forests:

1. Cristóbal, J., Prakash, A., Anderson, M.C., Kustas, W.P., Alfieri, J.G. and Gens, R., 2020. Surface energy flux estimation in two Boreal settings in Alaska using a thermal-based remote sensing model. Remote Sensing, 12(24), p. 4108.

We would like to thank the reviewer for revising the manuscript and providing helpful suggestions and comments to improve this work. Below we provide answers to the reviewer’s comments.

Response 1: Thank you very much for this useful comment. Indeed, it is true that our data paper focuses on 100-m evapotranspiration estimation over selected parts of the Mediterranean region. In order to avoid confusion regarding dataset extent, we have modified the title and the Conclusions section. In the abstract we directly indicate our study areas (i.e., Ebro, Po, Herault, and Medjerda basins) where ET is produced.

Response 2: We fully agree with your point that the TSEB performance driven by both original Sentinel-3 LST data and its 100-m downscaled product shall be carried out in order to evaluate the effectiveness of thermal sharpening on ET retrievals. In this work, we have exploited data mining sharpener (DMS) of Gao et al. (2012) successfully used in many research studies for estimating high spatial resolution TIR-ET (Anderson et al., 2021; Guzinski and Nieto, 2019; Yang et al., 2021; Guzinski et al., 2023). As presented by Guzinski and Nieto (2019), TSEB-PT driven by downscaled DMS-based surface temperatures is more performant compared to ET estimates driven by original 1-km LST data with around a 13% increase in Pearson correlation coefficient (r) between in-situ ETs and their corresponding modelled observations. Furthermore, the authors of the ESA Sen-ET report estimated evapotranspiration using METRIC, ESVEP, and TSEB-PT algorithms at 11 flux tower sites across different vegetation types and climate zones, and derived the best accuracy scores from the latter model when data mining LST sharpener (either based on Artificial Neural Networks or Decision Trees regressors) was applied (https://www.esa-sen4et.org/downloads/prototype_evaluation_v1.3.pdf). According to the report, the Priestley-Taylor Two-Source Energy Balance of ET was ranked as the most robust approach with consistently lower Root Mean Square Error (RMSE) and higher correlation for latent flux yielding an average RMSE of 90 W m^-2 and r exceeding 0.7, which largely outperforms METRIC and ESVEP by more than 11% and 30% for RMSE and Pearson correlation, respectively. Furthermore, the TSEB-PT has been constantly updated in order to improve the modelling scheme for thermal sharpening, and as reported in Guzinski et al. (2023) enhanced DMS-driven TSEB-PT at field scale achieved accuracy of 0.8 mm per day, which is our next-future goal to be implemented. Moreover, Sánchez et al. (2023) conducted extensive study on the performance of LST downscaling in Spain, and based on their validation results with in-situ measurements the DMS approach gave nearly two times smaller RMSE error compared to the 1-km S3 LST. In addition to the abovementioned literature review, in our co-authored paper we compared Sen-ET outcomes with other evapotranspiration products, including 3-km MSG SEVIRI and 70-m ECOSTRESS ET which on average gave less robust accuracy metrics than our 100-m retrievals (De Santis et al., 2022). These results and other authors’ findings moved us towards generation of 100-m ET dataset. Therefore, we have decided to apply the LST sharpening strategy in our ET workflow assuming its better performance in different land covers and climates compared to original 1-km S3-driven TSEB-PT. In the section 3.2 of the revised manuscript together with relevant research papers we provide more information on the performance of DMS procedure for estimating high spatial resolution evapotranspiration (lines 327-340; 358-363).

Considering the scarcity of eddy covariance towers over the Mediterranean catchments and time of interest (2017-2021) for our analysis, together with University of Ghent we have managed to gather in-situ observations at only eight eddy covariance sites (i.e., seven locations in Italy and one site in France) that provide long time-series of latent heat flux. In order to get more general conclusions on the results, we fully agree that the validation shall be performed including more in-situ EC towers represented by wider variety of landcover types, climate zones, and topography which is our future priority objective. This might be done by extending the spatial coverage of the ET data in order to increase number of available local flux measurements.

1. Gao, F., Kustas, W. P. and Anderson, M. C.: A data mining approach for sharpening thermal satellite imagery over land. Remote Sensing, 4(11), pp.3287-3319, 2012.
2. Anderson, M. C., Yang, Y., Xue, J., Knipper, K. R., Yang, Y., Gao, F., Hain, C. R., Kustas, W. P., Cawse-Nicholson, K., Hulley, G. and Fisher, J. B.: Interoperability of ECOSTRESS and Landsat for mapping evapotranspiration time series at sub-field scales. Remote Sensing of Environment, 252, p. 112189, 2021.
3. Guzinski, R. and Nieto, H.: Evaluating the feasibility of using Sentinel-2 and Sentinel-3 satellites for high-resolution evapotranspiration estimations. Remote sensing of Environment, 221, pp.157-172, 2019.
4. Yang, Y., Anderson, M. C., Gao, F., Wood, J. D., Gu, L. and Hain, C.: Studying drought-induced forest mortality using high spatiotemporal resolution evapotranspiration data from thermal satellite imaging. Remote Sensing of Environment, 265, p.112640, 2021.
5. Guzinski, R., Nieto, H., Sánchez, R. R., Sánchez, J. M., Jomaa, I., Zitouna-Chebbi, R., Roupsard, O. and López-Urrea, R.: Improving field-scale crop actual evapotranspiration monitoring with Sentinel-3, Sentinel-2, and Landsat data fusion. International Journal of Applied Earth Observation and Geoinformation, 125, p.103587, 2023.
6. Sánchez, J. M., Galve, J. M., Nieto, H. and Guzinski, R.: Assessment of High-Resolution LST Derived From the Synergy of Sentinel-2 and Sentinel-3 in Agricultural Areas. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 17, pp. 916-928, 2023.
7. De Santis, D., D'Amato, C., Bartkowiak, P., Azimi, S., Castelli, M., Rigon, R. and Massari, C.: Evaluation of remotely-sensed evapotranspiration datasets at different spatial and temporal scales at forest and grassland sites in Italy. In 2022 IEEE Workshop on Metrology for Agriculture and Forestry (MetroAgriFor) (pp. 356-361). IEEE, November 2022.

Response 3a: Thank you for your comment. We have included the contextual methods for evapotranspiration retrieval. You can find our modifications in the second paragraph of the Introduction section (lines 46-49). In the revised version of the manuscript, we included some research papers on the contextual ET methods. They are as follows:

1. Bastiaanssen, W. G. M., Noordman, E. J. M., Pelgrum, H., Davids, G., Thoreson, B. P. and Allen, R. G.: SEBAL model with remotely sensed data to improve water-resources management under actual field conditions. Journal of irrigation and drainage engineering, 131(1), pp. 85-93, 2005.
2. Chirouze, J., Boulet, G., Jarlan, L., Fieuzal, R., Rodriguez, J. C., Ezzahar, J., Er-Raki, S., Bigeard, G., Merlin, O., Garatuza-Payan, J. and Watts, C.: Intercomparison of four remote-sensing-based energy balance methods to retrieve surface evapotranspiration and water stress of irrigated fields in semi-arid climate. Hydrology and earth system sciences, 18(3), pp. 1165-1188, 2014.
3. Sobrino, J. A., Souza da Rocha, N., Skoković, D., Suélen Käfer, P., López-Urrea, R., Jiménez-Muñoz, J. C. and Alves Rolim, S. B.: Evapotranspiration Estimation with the S-SEBI Method from Landsat 8 Data against Lysimeter Measurements at the Barrax Site, Spain. Remote Sensing, 13(18), p. 3686, 2021.
4. Trezza, R., Allen, R. G. and Tasumi, M.: Estimation of actual evapotranspiration along the Middle Rio Grande of New Mexico using MODIS and landsat imagery with the METRIC model. Remote Sensing, 5(10), pp. 5397-5423, 2013.

Response 3b: Thank you very much for this suggestion. We have included this interesting research paper as a reference in the manuscript (line 102).

Response 4: To be more precise with the description for daily ET aggregation procedure from 30-min ground observations at available EC towers, in the third paragraph of the section 2.2 (starting from line 184) we provide more detailed information on daily ET retrieval according to the strategy developed by the project partners from the University of Ghent working in hydrology domain. Additionally, we provide a reference paper of Hulsman et al. (2023) that explains the preprocessing procedure for the in-situ EC observations:

1. Hulsman, P., Keune, J., Koppa, A., Schellekens, J. and Miralles, D. G.: Incorporating Plant Access to Groundwater in Existing Global, Satellite‐Based Evaporation Estimates. Water Resources Research, 59(8), p. e2022WR033731, 2023.

Response 5: As described in the Methodology section (lines 248-253; 306-309), we have generated daily ET maps with spatio-temporal coverage corresponding to daily Sentinel-3 (S3) LST data and 10-day Sentinel-2 composite product specially adjusted to S3 acquisition days. Indeed, in case of Sentinel-2 we minimized the cloud occurrence by means of temporal compositing, while S3 LST is more affected by overcast conditions. Sentinel-3 LST datasets were not composited and this is the main reason for relatively large spatiotemporal gaps in the daily ET product.

Response 6: Thank you for this comment. Canopy aerodynamic resistance (Rx) and green vegetation cover (fg) are expressed as follows:

1. Rx = (C´/F) * [l_w/(u_do+z_om)]^1/2

where C´ is derived from weighting a coefficient in the formulation for leaf boundary layer resistance over the height of canopy (it assumed to be equal to 90 s^1/2 m^-1), F is local leaf area index (i.e., LAI/fc with fc: fractional vegetation cover), lw is the effective leaf width, and u_do+z_om corresponds to the wind speed within the canopy-air interface (Norman et al., 1995).

1. 𝑓𝑔=𝐹𝐴𝑃𝐴𝑅/𝐹𝐼𝑃𝐴𝑅

where FAPAR is the fraction of absorbed photosynthetically active radiation obtained from the ESA Snap biophysical processor, and FIPAR corresponds to the fraction of photosynthetically active radiation intercepted by green and brown vegetation and it is expressed using following formula:

1. FIPAR = 1 – exp[-0.5*PAI/cosϴ]

where PAI = LAI / fg.

As shown above, the retrieval of fg is an iterative procedure that re-calculates FIPAR parameter until fg converges (Guzinski et al., 2020).

Green vegetation cover is driven by plant area index (PAI), FAPAR, and sun zenith angle (ϴ) derived from Sentinel-2 reflectance imagery. It means that fg values are estimated for each S2 pixel in space and time ranging from 0 to 1. Similarly to fg product, aerodynamic resistance at the canopy boundary layer is based on Sentinel-2 grid and changes in time since it is derived from Earth Observation inputs, such as LAI, ERA5 wind speed, and landcover information derived from ESA CCI LUT (e.g. lw) following Guzinski et al. (2019).

1. Norman, J. M., Kustas, W. P. and Humes, K. S., 1995. Source approach for estimating soil and vegetation energy fluxes in observations of directional radiometric surface temperature. Agricultural and Forest Meteorology, 77(3-4), pp.263-293.
2. Guzinski, R. and Nieto, H., 2019. Evaluating the feasibility of using Sentinel-2 and Sentinel-3 satellites for high-resolution evapotranspiration estimations. Remote sensing of Environment, 221, pp.157-172.
3. Guzinski, R., Nieto, H., Sandholt, I. and Karamitilios, G., 2020. Modeling high-resolution current evapotranspiration through Sentinel-2 and Sentinel-3 data fusion. Remote Sensing, 12 (9), p.1433.

Response 7: Indeed, air temperature and solar radiation from ERA5 data were enhanced with SRTM DEM. While shortwave radiation was corrected for illumination conditions using elevation, air temperature (TA) originally derived at 2-m height was recalculated to the blending height of 100 m using the elevation product and standard lapse rate of 6.5 K/1000 m. The blending height for low-spatial resolution air temperature is assumed to be more representative rather than TA at the height of 2 m due to weaker impact of land-atmosphere interactions at 100-m above ground surface. We expect a strong impact of those variables over complex areas, such as mountains where most EC towers are located in this study. To be more consistent, we have added some text (lines 234-235; 365-370) to explain the utility of DEM in the ERA5 input data adjustments together with a relevant research paper:

1. Guzinski, R., Nieto, H., Sánchez, J. M., López-Urrea, R., Boujnah, D. M. and Boulet, G.: Utility of copernicus-based inputs for actual evapotranspiration modeling in support of sustainable water use in agriculture. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 14, pp.11466-11484, 2021.

Response 8: Thank you very much for this suggestion. We included this publication in our manuscript (line 485).

Response 9: We fully agree with the reviewer’s comment. We have included this aspect in the indicated paragraph of the revised manuscript (lines 510-514). We also added some reference papers to explain limitations of reflectance bands as predictors for thermal downscaling as depicted below:

1. Hu, Y., Tang, R., Jiang, X., Li, Z.L., Jiang, Y., Liu, M., Gao, C. and Zhou, X., 2023. A physical method for downscaling land surface temperatures using surface energy balance theory. Remote Sensing of Environment, 286, p. 113421.
2. Merlin, O., Duchemin, B., Hagolle, O., Jacob, F., Coudert, B., Chehbouni, G., Dedieu, G., Garatuza, J. and Kerr, Y.: Disaggregation of MODIS surface temperature over an agricultural area using a time series of Formosat-2 images. Remote Sensing of Environment, 114(11), pp. 2500-2512, 2010.

Response to edits: Thank you. We have corrected the manuscript considering the abovementioned issues. In addition, we updated the table with a list of abbreviations and acronyms (page 27-28).