Assessing and Comparing Reference Evapotranspiration across Different Climatic Regions of China Using Reanalysis Products

Round 1

Reviewer 1 Report

Paper deals with a comparison of CLDAS abd ERA5 reanalysis data with observed and measured meteorological 689 station in China for climate parameter Tmax, Tmin, rH, Rs and calculated ETo after Penman-Monteith. It´s a huge statistical analysis of data, statistical methods are conventional, well described and used in the right manner (Pbias, R2, RMSE, MAE. Results are documented with ntables and figures well and spatial accuracy is shown (Fig. 4) and described. In Fig. 6 annual course of ETo with the three data bases is shown, higher discrepancy exist for the humid zone in China - why is not clear? Also in Fig. 5 in the caption is written 1st line, 3rd line and 2nd, 4th line - but in the figures only one correlation line!!

It must be more discussed, why wet and semi-humid zones in China differ more between the different ETo-methods/data?

It is necessary to seperate 3rd chapter in results and discussion; because open scientific questions why these discrepancies in the different data-bases  (observed, ERA5, CLDAS) mainly for ETo comes not clear out!

Also it´s of interest, how the analysed years 2017-2019 are different in ETo (spatial regionalization)!

In Fig. 1 it´s absolutely necessary to mark the 7 climatic zones - only shown the elevational differences in China!

 

Author Response

General comments：

Paper deals with a comparison of CLDAS abd ERA5 reanalysis data with observed and measured meteorological 689 station in China for climate parameter Tmax, Tmin, rH, Rs and calculated ETo after Penman-Monteith. It´s a huge statistical analysis of data, statistical methods are conventional, well described and used in the right manner (Pbias, R2, RMSE, MAE. Results are documented with ntables and figures well and spatial accuracy is shown (Fig. 4) and described.

 

Point 1: In Fig. 6 annual course of ET₀ with the three data bases is shown, higher discrepancy exist for the humid zone in China - why is not clear? Also in Fig. 5 in the caption is written 1st line, 3rd line and 2nd, 4th line - but in the figures only one correlation line!!

Response 1: Thank you for pointing out the problem,The purpose of Fig. 6 is to describe the consistency between the monthly time series of ET₀ estimated by the reanalysis and the observed ET₀, not the variability. In this section, the inappropriate content involving variability has been rewritten, and we have modified the caption of Fig. 5 and redrawn it.

 

Figure 5. Comparing ET_{0 Obs} with ET_{0 CLDAS} and ET_{0 ERA5}  on a daily scale, and using CLDAS, ERA5 reanalysis products of 7 different climate zones.

Point 2: It must be more discussed, why wet and semi-humid zones in China differ more between the different ETo-methods/data?

Response 2: Thank you for pointing out this problem,we have discussed this content in Section 4.3.

Comparing the accuracy of ET₀ estimation in different climatic regions, it was found that the variability of ET₀ in humid and semi-humid areas is higher than that in arid and semi-arid areas. Su et al. (2015) [70] also found that reanalysis estimates of ET₀ in arid and extremely cold regions are more consistent with site observations of ET₀ compared. The same trend was found by Zhang et al. (2018) [71] when they used ERA-Interim reanalysis data to study the characteristics of ET₀ in global arid and semi-arid regions. Furthermore, Paredes et al. (2018) [57] using the ERA-Interim reanalysis product for daily estimation of grass reference evapotranspiration also found that the higher values of RMSE were mostly concentrated in coastal areas of Portugal. These findings are consistent with the conclusions of this paper. This phenomenon can be attributed to the meteorological factors related to the estimation of ET₀. In result 2.1, we also reported that the accuracy of reanalysis R_s in arid and semi-arid regions is higher than that in humid regions, which is partly because the calculation of solar radiation is a complex task in itself, and the high uncertainty of atmospheric turbidity in humid climatic conditions which further exacerbates error between reanalysis solar radiation data and station observations [62,63], R_s is the most significant meteorological factor affecting ET₀[72]. The high error of reanalysis R_s leads to high variability of ET₀ in humid and semi-humid climate zones.

 

Point 3: It is necessary to seperate 3rd chapter in results and discussion; because open scientific questions why these discrepancies in the different data-bases (observed, ERA5, CLDAS) mainly for ETo comes not clear out!

Response 3: Thank you for pointing out this problem, we have separated the results and discussion section.

4.1. Analysis of meteorological variables related to ET₀ estimation

In this paper, the key meteorological factors for calculating ET₀ are evaluated. It is found that Tmin shows an overestimation trend, Tmax shows an underestimation trend, and PBias of Tmin is larger.Similar conclusions were found by Wu et al. (2022) [54] and Qian et al. (2022) [55] when they used reanalysis data to assess air temperature in the Chinese region. In addition, Simmons et al. (2010) [56] also reported a tendency for over-estimation of T_min and under-estimation of T_max by ERA-Interim, the same tendencies were reported for ERA-Interim estimates of T_max and T_min in Continental Portugal [57], those authors hypothesized that CLDAS, ERA5 reanalysis data sets are capturing warming over land more than over sea, which makes that T_{min CLDAS} and T_{min ERA5} larger than T_{min Obs} data, T_{max CLDAS}, T_{max ERA5} showed opposite performance. For reanalysis Rs, both CLDAS and ERA5 slightly overestimate the site-observed Rs. Which is similar to the reports of Bojanowski et al. (2014) [58] and Urraca et al. (2017) [59]. Moreover, Sheffield et al. (2006) [60] also reported a similar conclusion and combined this behavior of overestimation of global radiation for NCEP1 with the fact that the reanalysis dataset does not consider the blocking of shortwave radiation by clouds. The result on solar radiation reported in the literature that Rs from reanalysis is often overestimated when using reanalysis products with smaller timescales [61-63]. these authors hypothesize that the overestimation of Rs using the reanalysis dataset may be due to a lack of con-sideration of the impact of air pollution in some areas or the effects of water vapor when estimating atmospheric transmissivity. For reanalysis relative humidity, CLDAS and ERA5 were underestimated in most climatic regions and overestimated in a few climatic regions. This is consistent with the conclusion of Simmons et al. (2010) [56] that the relative humidity of ERA5 reanalysis is underestimated in the mid-low latitudes. Fu et al. (2015) [64] also concluded that the RH of NCEP1 reanalysis was slightly underestimated in Australia. Differently, Martins et al. (2016) [29] reported for Iberia no tendency for over- or underestimation of RH when using the reanalysis monthly products. The evaluation of reanalysis wind speed showed that its accuracy was low in seven climatic regions. The average R² of U_{2 CLDAS} is 0.25, and the average R2 of U_{2 ERA5} is 0.18, and many scholars have reanalyzed the wind speed and obtained similar results, the low accuracy of wind speed derived from reanalysis was reported by Carvalho et al. (2014) [65] studies applied to Continental Portugal. The same results were reported by Martins et al. (2016) [29] in the Iberian Peninsula, but a reanalysis of wind speed did not cause large variability in ET₀ [66]. For the problem of low accuracy in analyzing wind speed. The two wind speed estimation alternatives proposed by Allen et al (1998) [7]. are to replace the daily wind speed with the annual average wind speed over the local area or to use the world average wind speed value U₂=2m s^-1 as the default wind speed, obtained from over 2000 weather stations worldwide. Jabloun and Sahli (2008) [67] compared the above methods and found that both methods obtained wind speeds with high applicability, but the ET₀ was better estimated using the local area’s annual average wind speed.

4.2. Analysis of meteorological variables related to ET₀ estimation

Spatial and frequency distribution of the annual mean values of ET₀ calculated from stations observations, ET₀ estimated by CLDAS, and ET₀ estimated by ERA5 in China is analyzed. It is found that the frequency distribution and spatial distribution of different intervals of annual mean evapotranspiration estimated by CLDAS and ERA5 are similar to the observed ET₀, but in most cases, the reanalysis ET₀ is slightly overestimated. The overestimation of reference evapotranspiration derived from reanalysis has been discovered in many relevant studies, Betts et al. (2009) [68] compared ET₀ at the watershed scale using ERA-Interim reanalysis data and observations and found that the reference crop evapotranspiration from reanalysis data was significantly larger than observed value. Moreover, other studies using ERA-Interim reanalysis products also reported an overall over-estimation bias for ET₀ [28]. Dee and Uppala (2009, 2011) [44,69] considered that the overestimation of reanalysis data is mainly related to the utilization of small-time scale (daily or smaller timescale) reanalysis products, water vapor movement processes, and variational bias correction.

4.3 Analysis of the accuracy of ET₀ estimated by CLDAS and ERA5 in different climatic regions

Comparing the accuracy of ET₀ estimation in different climatic regions, it was found that the variability of ET₀ in humid and semi-humid areas is higher than that in arid and semi-arid areas. Su et al. (2015) [70] also found that reanalysis estimates of ET₀ in arid and extremely cold regions are more consistent with site observations of ET₀ compared. The same trend was found by Zhang et al. (2018) [71] when they used ERA-Interim reanalysis data to study the characteristics of ET₀ in global arid and semi-arid regions. Furthermore, Paredes et al. (2018) [57] using the ERA-Interim reanalysis product for daily estimation of grass reference evapotranspiration also found that the higher values of RMSE were mostly concentrated in coastal areas of Portugal. These findings are consistent with the conclusions of this paper. This phenomenon can be attributed to the meteorological factors related to the estimation of ET₀. In result 2.1, we also reported that the accuracy of reanalysis R_s in arid and semi-arid regions is higher than that in humid regions, which is partly because the calculation of solar radiation is a complex task in itself, and the high uncertainty of atmospheric turbidity in humid climatic conditions which further exacerbates error between reanalysis solar radiation data and station observations [62,63], R_s is the most significant meteorological factor affecting ET₀[72]. The high error of reanalysis R_s leads to high variability of ET₀ in humid and semi-humid climate zones.

Point 4: Also it´s of interest, how the analysed years 2017-2019 are different in ETo (spatial regionalization)!

Response 4: Thank you for pointing out this problem. In the study, we found that the spatial distribution of ET₀ between years was highly consistent, considering the length of the article, the content of this part was not added. In addition, the following articles can be used to understand the spatial differences of ET0 between years.

Zheng, R.W.; Zhang, Y.; W, Q.M.; Gui, Y.P. Prediction and attribution analysis of future evolution of ET0 in climatic regions of China. Journal of Yangtze River Scientific Research Institute, 1-10. http://kns.cnki.net/kcms/detail/42.1171.TV.20220718.2237.002.html

Wu, L.; Qian, L.; Huang, G.; Liu, X.G.; Wang, Y.C.; Bai, H.; Wu, S.F. Assessment of Daily of Reference Evapotran-spiration Using CLDAS Product in Different Climate Regions of China. Water. 2022, 14(11):1744. https://doi.org/10.3390/w14111744

Point 5: In Fig. 1 it´s absolutely necessary to mark the 7 climatic zones - only shown the elevational differences in China!

Response 5: Thank you for pointing out this problem, We have modified the graphics to use different colors to distinguish seven climate zones.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript aims at comparing the quality of the derivative reference evapotranspiration ET0 products based on ERA5 and CLDAS reanalyses products against station-based derivative coverage. This direction of regional studies emerges with the increase in precision and resolution of the global climatic data coverages. From this point I am surprised not meeting any references to WorldClim products throughout the manuscripts.

The manuscript will benefit from removing the redundant text sections, and section reordering. The authors could also be more inventive in finding better ways to graphically present their results.

The time period of the study is from 2017 (the launch of the CLDAS) to 2019, while in fact it can be extended to 2022 since all the data are available for this time period. Not using the most recent data significantly decreases the value of this study. 

Multiple suggestions can be found in the attached pdf as comments to the main text. The authors should be as clear and concise as possible in conveying their results to the readership, for now the textflow is messy and in many points redundant; which overall I quality as moderate revision.

Comments for author File: Comments.pdf

I have provided sevaral suggestions for the correct wording in the attached document. The English is mostly fine but the phrasing at certain points drives the reading experience down. The manuscript will benefit from more concise writing. The information already presented in Tables must not be re-presented in the main text and only addressed by reference to a particular table or figure.

Author Response

General comments：

I have provided sevaral suggestions for the correct wording in the attached document. The English is mostly fine but the phrasing at certain points drives the reading experience down. The manuscript will benefit from more concise writing. The information already presented in Tables must not be re-presented in the main text and only addressed by reference to a particular table or figure.

 

Point 1: The manuscript aims at comparing the quality of the derivative reference evapotranspiration ET0 products based on ERA5 and CLDAS reanalyses products against station-based derivative coverage. This direction of regional studies emerges with the increase in precision and resolution of the global climatic data coverages. From this point I am surprised not meeting any references to WorldClim products throughout the manuscripts.

 Response 1: Thank you for pointing out the problem, we have added the corresponding part in the introduction.

Several global reanalysis datasets are available for ET₀ estimation, such as World Clim provides global historical and future climate data and elevation data [19]; The World Meteorological Organization's Climate Explorer weather data retrieval platform with a wealth of global or regional climate data [20]. 

 

Point 2: The manuscript will benefit from removing the redundant text sections, and section reordering. The authors could also be more inventive in finding better ways to graphically present their results.

Response 2: Thank you for pointing out the problem, We have simplified and reordered each part of the article and modified the graphics to make the article more readable.

Point 3: The time period of the study is from 2017 (the launch of the CLDAS) to 2019, while in fact it can be extended to 2022 since all the data are available for this time period. Not using the most recent data significantly decreases the value of this study.

Response 3: Thank you for pointing out this problem, but due to the impact of the novel coronavirus, only site data for 2017-2019 were collected, and there were many missing sites after 2000. In the later research work, we will consider using the existing findings to correct the data set after 2000.

 

Point 4: Multiple suggestions can be found in the attached pdf as comments to the main text. The authors should be as clear and concise as possible in conveying their results to the readership, for now the textflow is messy and in many points redundant; which overall I quality as moderate revision.

Response 4: Thank you for your valuable suggestions for this article. For the comments in pdf, we have carefully read each article, made detailed modifications, and adjusted the structure of the entire text. In addition, in order to make the results as clear and concise as possible, we have also made a lot of rewriting of the conclusions and discussions.

 

Specific comments

 

Point 5: In Abstract, aim to avoid acronyms or at least introduce them appropriately before the first use. I suggest rewriting this Abstract more concisely, with shorter phrasing and more direct logic.

 Response 5: Thank you for pointing out the problem,We have rewrited the content of this part, streamlined the redundant part, and especially modified the use of acronyms.

Abstract:This study aims to assess the accuracy of the reference evapotranspiration （ET0）estimated by CLDAS, ERA5 reanalysis products, and the quality of reanalysis weather variables required to calculate PM-ET0. For this purpose, the applicability of surface meteorological elements from the ERA5 reanalysis datasets provided by the European Centre for Medium-Range Weather Forecasts (ECMWF), and the second‐generation China Meteorological Administration Land Data Assimilation System (CLDASV2.0) datasets are evaluated in China by comparison with local observations from 689 stations reported by the Chinese Meteorological Administration (CMA). Statistical statistics including percent bias (PBias), coefficient of determination (R2), root mean square error (RMSE) and mean absolute error (MAE) are used to check the accuracy. The results show the highest correlation between reanalysis temperature and observations, with a mean R2 of 0.96,0.90 for CLDAS maximum and minimum air temperatures and 0.87,0.84 for ERA5, for the reanalysis solar radiation (Rs) and relative humidity (RH), an overestimation trend is shown for Rs, an underestimation trend is shown for RH, for reanalysis wind speed, it shows relatively low accuracy. The accuracy of ET0 estimated by the two reanalysis products is acceptable in China, but the spatial and temporal consistency between CLDAS estimates and site observations is higher, with mean RMSE, R2 of 0.91,0.82 for CLDAS and 1.42, 0.70 for ERA5, respectively, moreover, CLDAS reanalysis products are more effective in describing the boundary details of the study area.

 

Point 6: Allen et al. is not an appropriate reference for this point, better refer to Holdridge life zones system or the like

Response 6: Thank you for pointing out the problem, we replaced the corresponding references.

1.Holdridge, L.R. Life Zone Ecology. Tropical Science. 1967

 

Point 7: Correct reference might be the FAO Irrigation and drainage paper 56 itself, this Allen et al. used before

Response 7: Thank you for pointing out the problem, We have modified this reference.

Allen, R.G.; Pereira, L.S.; Raes, D.; Smith, M. Crop evapotranspiration: guidelines for computing crop water requirements. Food and Agriculture Organization of the United Nations, Rome. 1998

 

Point 8: This section should stand before the ETc presentation, as ET0 is a more general concept, and ETc is based on ET₀

Response 8: Thank you for pointing out the problem, we have modified it

 

Point 9: Not all empirical models estimate PM-ET0, there are Hargreaves, Thornthwaite, Blaney-Criddle, to name a few.

Response 9: Thank you for pointing out the problem, we have modified this description.

At present, ET₀ can also be estimated from pan evaporation, but such methods are expensive, technically complex, and not practical. Therefore, Penman-Monteith model is usually used to estimate ET₀

 

Point 10: 'coarse'; again, WorldClim has roughly 1km resolution at the equator which is quite fine

Response 10: Thank you for pointing out the problem, We have modified this inappropriate term.

 

Point 11:reference

Response 11: Thank you for pointing out the problem, we have supplemented this part.

ERA5_land reanalysis data were applied to observations of Peruvian glaciers verified by Martin et al, and found to be appropriate for characterizing 2 m air temperature and relative humidity, particularly in the wet outer tropics [26]. Song et al. (2020) [27] compared the applicability of various soil moisture data in Inner Mongolia and found that the ERA5 simulation capability is optimal.

26.Martí, B.; Jose, U.; Giovanni, L.; Philipp, K.; álvaro, N.; Rolando. C. Validation of ERA5-Land temperature and relative humidity on four Peruvian glaciers using on-glacier observations. Journal of Mountain Science. 2022, 19, 1849-1873.https://doi.org/10.1007/s11629-022-7388-4

27.Song, H.Q.; Sun, X.L.; Li, Y.P. Evaluation of EＲA5 reanalysis soil moisture over inner mongolia. Science Technology and Engineering. 2020, 20, 2161-2168.

 

Point 12: Line 129 :repetitive with the rest of the phrase

Response 12: Thank you for pointing out the problem. We have modified it. Using GLDAS and CLDAS forcing data.

 

Point 13: Line 147: more topic-specific references are needed, i.e., Vanella et al., 2022 https://doi.org/10.1016/j.ejrh.2022.101182 and the like

Response 13: Thank you for pointing out the problem, We have modified the references of this part.

Vanella, D.; Longo-Minnolo, G.; Belfiore, O.R.; Ramírez-Cuesta, J.M.; Pappalardo, S.E.; Consoli, S.; D’Urso, G.; Chirico, G.B.; Coppola, A.; Comegna, A.; Toscano, A.; Quarta, R.; Provenzano, G.; Ippolito, M.; Castagna, A.; Gandolfi, C. Comparing the use of ERA5 reanalysis dataset and ground-based agrometeorological data under different climates and topography in Italy. J. Hydrol. 2022, 101182. https://doi.org/10.1016/j.ejrh.2022.101182

Shea, J.G.; Worley, J.; Stern, R.N. An introduction to atmospheric and oceanographic datasets. (No.NCAR/TN-404+IA). University Corporation for Atmospheric Research. 1994, https://doi.org/10.5065/D6NP22DP

 

Point 14: Line 150: applicability

Response 14: Thank you for pointing out the problem, we have modified it and replaced it with region-specific applicability.

 

Point 15: Line 167: This section is in part repetitive to the Introduction, so consider regrouping the text to eliminate redundancy.

Response 15: Thank you for pointing out the problem, we have deleted the repetitive part of the introduction and reintegrated the content of this part to make it more streamlined.

CLDAS uses fusion and assimilation technology to fuse data from various sources such as ground observation, satellite observation, numerical model products, and numerical model products, then output land surface driven products with high spatial and temporal resolution including maximum and minimum air temperatures, 2m specific humidity, 10m wind speed, et. It covers the Asian region (60°- 160°E, 0°- 65°N) with a spatial resolution of 0.0625°× 0.0625° and a temporal resolution of 1h. The dataset is formed by using the ECMWF numerical analysis/forecasting product as the background field and the topographic adjustment, Multigrid variational assimilation, optimal interpolation and other techniques in the Chinese region to fuse the ground-based automatic station observations and interpolate them to the reanalysis grid points [33,43].

ERA5 is the fifth generation of ECMWF's atmospheric reanalysis of the global climate, created by the EU-funded Copernicus Climate Change Service (C3S). It assimilates remote sensing information, upper atmosphere, and near-surface conventional observation data including different regions and sources on a global scale, spanning from 1979 to the present, achieving real-time updates with a spatial resolution of 0.25°0.25° latitude-longitude [44]. ERA5 reanalysis data provide many kinds of meteorological elements, including 2m air temperature, 2m dew point temperature, and 10m wind speed, et. Data sources at http://cdds.climate.copernicus.eu/.

 

Point 16: Line 193-194: A list of the retrieved variables is defined by the Penman-Monteith FAO56 Framework, so I suggest presenting the methodology (your Section 2.3) before you describe the retrieved data.

Response 16: Thank you for pointing out the problem. We changed Section 2.3 and modified the formula.

 

Point 17: Line 214-215, 218-219: methodology for treating the observed data is unclear from this description

Response 17: Thank you for pointing out the problem,we have rewritten the paragraph to make it clearer and more accurate.

The meteorological variables related to the calculation of PM-ET0 have been extracted from the CLDAS2.0, ERA5 reanalysis grid points, and then these grid points’ meteorological data are interpolated to the same latitude and longitude positions as the 689 ground observation sites, and the meteorological data obtained from the interpolation are evaluated as true values for the test. Grid data from four grid points around it were selected and interpolated to the station by the inverse distance weight (IDW) method. The formula is as follows:

 

where  is the true value, is the value of the control point, and  is the weight coefficient.

 

Point 18: Line 261: The following two paragraphs explain the same point twice, so they might be combined to facilitate reading and save space. Thse references that support the points made above should be moved above accordingly to avoid excess citations in the Results and discussion section.

Response 18: Thank you for pointing out the problem, we have rewritten it, combined them together, and separated the discussion section, increasing the readability of the article.

 

Point 19: Line 294: For Tables 1 and 2 better group by variable, not by reanalysis, this enhances the comparison and understanding of your data by the general readership.

 

Response 19: Thank you for pointing out the problem, we have modified Tables 1 and 2 by grouping variables to make the article easier to understand.

 

Point 20: Line 351, 365: reanalysis monthly,

 

Response 20: Thank you for pointing out the problem, we have modified the phrase and the caption of the graph.

 

Point 21: Line 416: shorten the subsection title

Response 21: Thank you for pointing out the problem, we have shortened the subtitle title. Comparison of annual mean of observations calculated ET0 and reanalysis estimated ET₀ 

 

Point 22: Line 493:Figures like this suggest using concordance instead or in addition to correlation as a significant and useful metric to estimate the closeness to a 1:1 ratio line.

 

Response 22: Thank you for pointing out the problem, We have redrawn this graph.

 

Figure 5. Comparing ET0 Obs with ET0 CLDAS and ET0 ERA5  on a daily scale, and using CLDAS, ERA5 reanalysis products of 7 different climate zones.

 

Point 23: Line 530: Estimators such as Nash-Sutcliffe efficiency can be used to assess the time series accuracy in general. Otherwise, generalized interannual distributions can be used to describe the assessment quality, since we would normally nothing about how one year differs from another.

 

Response 23: Thank you for your valuable suggestions, we will use Nash-Sutcliffe efficiency and other methods to estimate the accuracy of time series in the following study, but the main purpose of this figure is to describe the consistency of ET0 between reanalysis estimation and site observation in time series, not to express its variability. We also modify the inappropriate content in this section.

 

Point 24: Line 534: This might come before the previous section, as sson as general ET₀ accuracy estimates are involved

Response 24: Thank you for pointing out the problem, we have reordered the section to make the article more continuous.

 

 

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The reviewer questions are solved well - Fig. 5 new; responses for 2 - 5 are all ok; now better with the separation between results and discussion.

Author Response

The reviewer questions are solved well - Fig. 5 new; responses for 2 - 5 are all ok; now better with the separation between results and discussion.

 Response 1:  Thank you for your affirmation of the revised part of my manuscript, and we have revised the discussion

4.1. Analysis of meteorological variables related to ET₀ estimation

In this paper, the key meteorological factors for calculating ET₀ are evaluated. It is found that T_min shows an overestimation trend and T_max showes an underestimation trend, and the consistency of the reanalysis minimum temperature of the two data sets ( average R² is 0.96 and 0.87 ) is higher than that of the maximum temperature ( 0.90 and 0.84 ), but the PBias of the former is larger. Similar conclusions were found by Wu et al. (2022) [54] and Qian et al. (2022) [55] when they used reanalysis data to assess air temperature in the Chinese region. In addition, Simmons et al. (2010) [56] and Paredes et al [57] also reported a tendency for over-estimation of T_min and under-estimation of T_max by ERA-Interim in Continental Portugal [57], those authors hypothesized that CLDAS, ERA5 reanalysis data sets are capturing warming over land more than over sea, which makes that T_min CLDAS and T_min ERA5 larger than T_min Obs data, T_max CLDAS, T_max ERA5 showed opposite performance. For reanalysis Rs, both CLDAS and ERA5 slightly overestimate the site-observed Rs, the PBias of R_{s CLDAS} and R_{s ERA5} range from 0.02-0.17 and 0.06-0.24 Which is similar to the reports of Bojanowski et al. (2014) [58] and Urraca et al. (2017) [59]. Sheffield et al. (2006) [60] also reported a similar conclusion and combined this behavior of overestimation of global radiation for NCEP1 with the fact that the reanalysis dataset does not consider the blocking of shortwave radiation by clouds. As for reanalysis relative humidity, the PBias of RH_CLDAS are negative (ranging from -0.07 to -0.30) in the 1-6 climate zones,  a slight underestimation, the PBias of climate zone 7 is positive, and value is 0.35.. This is consistent with the conclusion of Simmons et al. (2010) [56] that the relative humidity of ERA5 reanalysis is underestimated in the mid-low latitudes. Fu et al. (2015) [61] also concluded that the RH of NCEP1 reanalysis was slightly underestimated in Australia. Differently, Martins et al. (2016) [29] reported for Iberia no tendency for over- or underestimation of RH when using the reanalysis monthly products, which can be attributed to the different time scales of the study. Different from the previous four meteorological factors, the analysis wind speed showed lower accuracy in seven climatic regions, the average R² of U_{2 CLDAS} is 0.25, and the average R2 of U_{2 ERA5} is 0.18. Many scholars have reanalyzed the wind speed and obtained similar results, the low accuracy of wind speed derived from reanalysis was reported by Carvalho et al. (2014) [62] studies applied to Continental Portugal. The same results were reported by Martins et al. (2016) [29] in the Iberian Peninsula, For the problem of low accuracy in analyzing wind speed. The two wind speed estimation alternatives proposed by Allen et al (1998) [7]. are to replace the daily wind speed with the annual average wind speed over the local area or to use the world average wind speed value U₂=2m s^-1 as the default wind speed, obtained from over 2000 weather stations worldwide. Jabloun and Sahli (2008) [63] compared the above methods and found that both methods obtained wind speeds with high applicability, but the ET₀ was better estimated using the local area’s annual average wind speed.

4.2. Analysis of the spatial distribution of observations calculated ET₀ and reanalysis estimated ET₀ 

Spatial and frequency distribution of the annual mean values of ET₀ calculated from stations observations, ET₀ estimated by CLDAS, and ET₀ estimated by ERA5 in China is analyzed. It is found that the frequency distribution and spatial distribution of different intervals of annual mean evapotranspiration estimated by CLDAS and ERA5 are similar to the observed ET₀, but in most cases, the reanalysis ET₀ is slightly overestimated. The overestimation of reference evapotranspiration derived from reanalysis has been discovered in many relevant studie., Betts et al. (2009) [64] compared ET₀ at the watershed scale using ERA-Interim reanalysis data and observations and found that the reference crop evapotranspiration from reanalysis data was significantly larger than observed value. Moreover, other studies using ERA-Interim reanalysis products also reported an overall over-estimation bias for ET₀ [28]. Dee and Uppala (2009, 2011) [44,65] considered that the overestimation of reanalysis data is mainly related to the utilization of small-time scale (daily or smaller timescale) reanalysis products, water vapor movement processes, and variational bias correction.

4.3 Analysis of the accuracy of ET₀ estimated by CLDAS and ERA5 in different climatic regions

Comparing the accuracy of ET₀ estimation in different climatic regions, it was found that the variability of ET₀ in humid and semi-humid areas is higher than that in arid and semi-arid areas, the RMSE of ET_{0 CLDAS} ranged from 0.78-0.82 in climate zones 1-3, RMSE ranged from 0.89-0.91 in climate zones 4-6, and RMSE of ET_0ERA5 ranged from 1.12-1.33 in climate zones 1-3 and RMSE ranged from 1.42-1.59 in climate zones 4-6. Su et al. (2015) [66] also found that reanalysis estimates of ET₀ in arid and extremely cold regions are more consistent with site observations of ET₀ compared. The same trend was found by Zhang et al. (2018) [67] when they used ERA-Interim reanalysis data to study the characteristics of ET₀ in global arid and semi-arid regions. Furthermore, Paredes et al. (2018) [57] using the ERA-Interim reanalysis product for daily estimation of grass reference evapotranspiration also found that the higher values of RMSE were mostly concentrated in coastal areas of Portugal. These findings are consistent with the conclusions of this paper. This can be attributed to the meteorological factors related to the estimation of ET_0, in result 2.1, we also reported that the accuracy of reanalysis R_s in arid and semi-arid regions is higher than that in humid regions, which is partly because the calculation of solar radiation is a complex task in itself, and the high uncertainty of atmospheric turbidity in humid climatic conditions which further exacerbates error between reanalysis solar radiation data and station observations, R_s is the most significant meteorological factor affecting ET₀[68]. The high error of reanalysis R_s leads to high variability of ET₀ in humid and semi-humid climate zones.

Author Response File: Author Response.docx