Improved Soil Evaporation Remote Sensing Retrieval Algorithms and Associated Uncertainty Analysis on the Tibetan Plateau

Feng, Jin; Zhang, Ke; Zhan, Huijie; Chao, Lijun

doi:https://doi.org/10.5194/hess-2022-210

Preprints

https://doi.org/10.5194/hess-2022-210

Preprints

10 Jun 2022

Status: a revised version of this preprint is currently under review for the journal HESS.

Improved Soil Evaporation Remote Sensing Retrieval Algorithms and Associated Uncertainty Analysis on the Tibetan Plateau

Jin Feng^1,2,3, Ke Zhang^1,2,3,4, Huijie Zhan¹, and Lijun Chao^1,2,3 Jin Feng et al.

Show author details

¹State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, and College of Hydrology and Water Resources, Hohai University, Nanjing, Jiangsu, 210098, China
²Yangtze Institute for Conservation and Development, Hohai University, Nanjing, Jiangsu, 210098, China
³CMA-HHU Joint Laboratory for Hydro-Meteorological Studies, Hohai University, Nanjing, Jiangsu, 210098, China
⁴Key Laboratory of Water Big Data Technology of Ministry of Water Resources, Hohai University, Nanjing, Jiangsu, 210098, China

Received: 01 Jun 2022 – Discussion started: 10 Jun 2022

Abstract. Actual evapotranspiration (ET) is the key link between water and energy cycles. However, accurate evaporation estimation in alpine barren areas remains understudied. In this study, we aimed to improve the satellite-driven Process-based Land Surface ET/Heat fluxes algorithm (P-LSH) by introducing two frameworks for quantifying moisture constraints to ET, and to test the applicability of satellite soil moisture and precipitation data for improving ET retrieval. As a result, it formed two improved P-LSH algorithms. The first framework normalizes the surface soil moisture to represent moisture stress, while the second framework takes the ratio of cumulative precipitation to cumulative equilibrium evaporation to quantify soil water stress. We systematically assessed the performances of the two improved P-LSH algorithms and six existing remote sensing ET retrieval algorithms on two barrens-dominated basins of the Tibetan Plateau using reconstructed ET estimates derived from the terrestrial water balance method as a benchmark. The two frameworks largely improved the performance of the P-LSH algorithm and showed better performance in both basins (root mean square error (RMSE) = 7.36 and 7.76 mm month^-1; R² = 0.86 and 0.87), resulting in a higher simulation accuracy than all six existing algorithms. We used five soil moisture and five precipitation datasets to investigate the impact of moisture constraint uncertainty on the improved P-LSH algorithm. The ET estimates of the improved P-LSH algorithm, driven by the GLDAS_Noah soil moisture, performed best compared with those driven by other soil moisture and precipitation datasets, while ET estimates driven by various precipitation datasets generally showed a high and stable accuracy. These results suggest that high-quality soil moisture can optimally express moisture supply to ET, and that more accessible precipitation data can serve as a substitute for soil moisture as an indicator of moisture status for its robust performance in barren evaporation.

Jin Feng et al.

Status: final response (author comments only)

RC1:
'Comment on hess-2022-210', Anonymous Referee #1, 04 Jul 2022
General Comment:

This manuscript tried to improve the satellite-based land surface ET algorithm by introducing soil moisture. Using satellite data to simulate the water cycle or calibrate models are very attractive considering the growing availability of remote sensing data. The method is reasonable, the findings are useful, and the article was well written. However, there are still some minor issues that need to be addressed.

Specific Comments:

L15: ‘… introducing two frameworks …’ there are no any information about these ‘two’ in this sentence. It would be better combine this and the following sentence.

L20: ‘two improved P-LSH algorithm’ seems not clear. What are they?

It is better to highlight the significant point of the study further.

Table 1: why use 30’’

Figure 3 and 4: More information (A1, …; ETrecon) is needed in the caption to make it be understandable.

I am not quite clear how the authors evaluate the ‘uncertainty’.
Citation: https://doi.org/10.5194/hess-2022-210-RC1
- AC1:
  'Reply on RC1', Ke Zhang, 30 Jul 2022
  We greatly appreciate the anonymous referee for providing valuable and constructive comments that are of great help for us to improve the quality of the manuscript. We have fully considered the comments and will revise the manuscript accordingly. The point-to-point responses to the comments and our plans for revision are listed below.
  Replies to the General Comments:This manuscript tried to improve the satellite-based land surface ET algorithm by introducing soil moisture. Using satellite data to simulate the water cycle or calibrate models are very attractive considering the growing availability of remote sensing data. The method is reasonable, the findings are useful, and the article was well written. However, there are still some minor issues that need to be addressed.
  Response:
  Thanks for your positive evaluation and encouraging comments on our manuscript. Your individual comments are replied below.
  
  Replies to the Specific Comments:
  Replies to the Specific Comments:
  L15: ‘… introducing two frameworks …’ there are no any information about these ‘two’ in this sentence. It would be better combine this and the following sentence.
  
  Response:
  Thanks for your suggestion. We proposed to modified the descriptions to “…remains understudied. In this study, we aimed to improve the satellite-driven Process-based Land Surface ET/Heat fluxes algorithm (P-LSH) by introducing two frameworks for quantifying moisture constraints to ET, with the first framework normalizing the surface soil moisture, and the second framework taking the ratio of cumulative precipitation to cumulative equilibrium evaporation. As a result, it formed two improved P-LSH algorithms. We systematically assessed the performances of the two improved P-LSH algorithms and …”.
  
  L20: ‘two improved P-LSH algorithm’ seems not clear. What are they?
  
  Response:
  Thanks for your suggestion. “the two improved P-LSH algorithms” refers to the improved algorithms composed of the two moisture constraint frameworks. We have restated the description and make the sentence clearer, as shown in Comment #1.
  
  It is better to highlight the significant point of the study further.
  
  Response:
  Thanks for your suggestion. We conducted this study based on the following backgrounds. On one hand, the Tibetan Plateau is crucial for Asian monsoon development and concurrent water and energy cycles, but relatively few studies have been carried out on its hinterland because of the difficulty of surveying. Remote sensing retrieval can conveniently estimate ET in this region, but the accuracy needs to be evaluated. On the other hand, some studies (Zhang et al., 2015; Pan et al., 2020) have shown that water supply is the main control factor of ET in arid and semi-arid regions, whose structures are rarely systematically assessed or discussed in existing studies.
  In this paper, the feasibility of various moisture constraint equations in existing ET algorithms in typical arid/semi-arid basins was analyzed, and then the soil moisture and precipitation were used to improve the P-LSH algorithm. Finally, the uncertainties of key inputs are assessed. Thanks again. We will further highlight the significant point in the revised manuscript.
  
  Table 1: why use 30’’
  
  Response:
  As we mentioned in the caption of Table 1, we listed the original resolution of the datasets. The soil properties dataset is in a raster format with a resolution of 30 arc seconds. To match other inputs in the ET algorithm, we aggregated the dataset from the original 30'' resolution to 1/12° using the arithmetic averaging method.
  
  Figure 3 and 4: More information (A1, …; ET_recon) is needed in the caption to make it be understandable.
  
  Response:
  Thanks for your suggestion. The A1, A2, A3, A4, A5, A6 are the coupling algorithms that coupled the vegetation evapotranspiration scheme and water evaporation scheme from the P-LSH algorithm with the six existing soil evaporation algorithms (see Table 2). The ET_recon item represents the reconstructed ET estimates derived from the terrestrial water balance method. We will add more information in the caption of Figures 3 and 4 in the revised manuscript.
  
  I am not quite clear how the authors evaluate the ‘uncertainty’.
  
  Response:
  In the improved algorithms, the precipitation and soil moisture data are used to express the moisture constraint on ET. We investigated the impact of various precipitation and soil moisture datasets on the ET to determine the impacts of key inputs uncertainty on model outputs. Taking the P-LSH_θ algorithm as an example, we investigated the variation between multiple ET estimates derived from multiple soil moisture datasets, and as a comparison, we also investigated the variation of multiple soil moisture datasets. The same method is also applied to the uncertainty evaluation of P-LSH_Palgorithm, as shown in Figures 10-12. By quantifying the variations between soil moisture/precipitation and their responding ET estimates, the characteristics and uncertainties of the improved algorithms are discussed in Section 4.3. Thanks for your question. We will add the clear description to the revised manuscript.
  
  Citation: https://doi.org/10.5194/hess-2022-210-AC1
RC2:
'Comment on hess-2022-210', Anonymous Referee #2, 29 Jul 2022
This manuscript improves satellite-based algorithm for estimating soil evaporation by adding the frameworks for quantifying moisture constraints to ET into P-LSH model, and assesses the impact of moisture constraint uncertainty on the estimated ET. Mechanism studies about ET and their components (e.g., transpiration, soil evaporation etc) in alpine barren areas, especially for Tibetan Plateau (TP), are very limited, and this study of ET mechanism in TP region is quite necessary. There are still some issues should be addressed before a publication.

The main comments:

The authors have paidmore attention to soil evaporation, and neglected the vegetation transpiration. For example, in line 538-539 of page 23, “On the barrens of the TP, vegetation is sparse, and only soil evaporation exists”. This addressing is not very rigorous. Grasslands account for 20.2% in Qaidam basin and 39.7% in Qiangtang Plateau, respectively, and Nelson et al (2020) indicate that transpiration in grasslands accounts for 40%-60% ET during growing reasons. I think that the authors should pay some attentions to transpiration estimation by considering the uncertainties of some others vegetation canopy conductance models. And, it is not clear the canopy conductance is calculated by which model; is it the empirical relationship between conductance and climatic variables, or the Jarvis-Stewart model? If the later, Jarvis model has poor performances in capturing the responses of conductance to climatic variables (e.g. air temperature), compared to other models such as Ball-Berry model, Ball-Berry-Leuning model and Mdelyn model. The uncertainties caused by choice of conductance model on ET may result in 32%-53% errors (Zhao et al., 2020). Therefore, I suggest the authors can also consider the influences of vegetation conductance model on estimated ET in TP.

The methoddescription for estimating soil evaporation is not clear in section 3. The authors introduced five existing soil evaporation algorithms and then proposed two improvements. In each algorithm, descriptions of main parameters are needed. For example, how the biome-specific constants are determined in the PM-Brust soil evaporation algorithm?

What is the difference between P-LSH soil evaporation algorithm (P-LSHp) and PML soil evaporation algorithm? How thepotential evaporation was calculated? Actually, it is not fair to compare soil evaporation algorithms with different potential evaporation equations. If the authors use the same equations, it is reasonable to compare soil evaporation algorithms. And, the difference between P-LSH soil evaporation algorithm (PLSθ) and PML soil evaporation algorithm is fwet. Why the authors do not add the fwet into PLSθ ?

Figure 3 and 4 have showedthe results of A1-A6 for five existing soil evaporation algorithms. I suggest that two improvement soil evaporation algorithms proposed by the authors should be added into the comparisons.

Some specific comments:

Line 68: “32 days”is right?

Sometimes, the logical relationshipbetween some context sentences is not strong. For example, line 110-111: “Saline lakes and deserts cover approximately one-quarter 110 and one-third of the Qaidam Basin, respectively. This region is thus very dry.”.

Figure 1 should include scale bar and compass.

Line 301: the description “vegetation evapotranspiration”is not right.

References

Nelson JA, Pérez-Priego O, Zhou S, et al. Ecosystem transpiration and evaporation: Insights from three water flux partitioning methods across FLUXNET sites. Glob Change Biol. 2020;00:1–15. https://doi.org/10.1111/gcb.15314.

Zhao, Wen Li; Qiu, Guo Yu; Xiong, Yu Jiu; Paw U, Kyaw Tha; Gentine, Pierre; Chen, Bao Yu (2020). Uncertainties caused by resistances in evapotranspiration estimation using high-density eddy covariance measurements. Journal of Hydrometeorology, doi:10.1175/JHM-D-19-0191.1.
Citation: https://doi.org/10.5194/hess-2022-210-RC2
- AC2: 'Reply on RC2', Ke Zhang, 20 Aug 2022
  
  Please see the attached PDF file for the detailed responses.
  
  Citation: https://doi.org/10.5194/hess-2022-210-AC2

Jin Feng et al.

Viewed

Total article views: 595 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
480	98	17	595	2	1

HTML: 480
PDF: 98
XML: 17
Total: 595
BibTeX: 2
EndNote: 1

Views and downloads (calculated since 10 Jun 2022)

Month	HTML	PDF	XML	Total
Jun 2022	218	40	7	265
Jul 2022	153	23	7	183
Aug 2022	84	30	3	117
Sep 2022	25	5	0	30

Cumulative views and downloads (calculated since 10 Jun 2022)

Month	HTML	PDF	XML	Total
Jun 2022	218	40	7	265
Jul 2022	153	23	7	183
Aug 2022	84	30	3	117
Sep 2022	25	5	0	30

Viewed (geographical distribution)

Total article views: 569 (including HTML, PDF, and XML) Thereof 569 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 28 Sep 2022

Short summary

Here we improved a satellite-driven evaporation algorithm by introducing two frameworks for quantifying moisture constraints to ET. The two frameworks greatly improved the evaporation estimation on two barrens-dominated basins of the Tibetan Plateau. Moisture constraint uncertainty showed that high-quality soil moisture can optimally express moisture, and more accessible precipitation data generally have robust performance in barren evaporation.


Total:	0
HTML:	0
PDF:	0
XML:	0