Next Article in Journal
Assessment of Different Object Detectors for the Maturity Level Classification of Broccoli Crops Using UAV Imagery
Next Article in Special Issue
Prospects for Reconstructing Daily Runoff from Individual Upstream Remotely-Sensed Climatic Variables
Previous Article in Journal
Evaluating Machine Learning and Remote Sensing in Monitoring NO2 Emission of Power Plants
Previous Article in Special Issue
The Influence of Climate Change on Forest Fires in Yunnan Province, Southwest China Detected by GRACE Satellites
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Remote Sensing
Volume 14
Issue 3
10.3390/rs14030730
remotesensing-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Estimation of Evapotranspiration in the Yellow River Basin from 2002 to 2020 Based on GRACE and GRACE Follow-On Observations

by
Wei Qu
1,2,*,
Zehui Jin
1,
Qin Zhang
1,
Yuan Gao
1,
Pufang Zhang
1 and
Peinan Chen
1
1
College of Geology Engineering and Geomatics, Chang’an University, Xi’an 710054, China
2
State Key Laboratory of Geohazard Prevention and Geoenvironment Protection, Chengdu University of Technology, Chengdu 610059, China
*
Author to whom correspondence should be addressed.
Remote Sens. 2022, 14(3), 730; https://doi.org/10.3390/rs14030730
Submission received: 27 December 2021 / Revised: 1 February 2022 / Accepted: 3 February 2022 / Published: 4 February 2022
(This article belongs to the Special Issue Carbon, Water and Climate Monitoring Using Space Geodesy Observations)
Browse Figures
Versions Notes

Abstract

:
Evapotranspiration (ET) plays an important role in the hydrological cycle of river basins. Studying ET in the Yellow River Basin (YRB) is greatly significant for the scientific management of water resources. Here, we made full use of the advantages of the Gravity Recovery and Climate Experiment (GRACE) and GRACE Follow-On (GRACE-FO) gravity satellites for monitoring large-scale hydrological changes to calculate the terrestrial water storage anomaly (TWSA) and terrestrial water flux in the YRB from May 2002 to June 2020. Furthermore, combined with terrestrial water flux, precipitation, and runoff data, ET in the YRB was calculated based on the water budget equation and then compared with other traditional ET products. The mutation of annual mean ET was identified by the Mann–Kendall trend test method, and the seasonal and interannual variations of ET were explored. ET was closely related to precipitation. Annual mean ET exhibited a sudden change in 2011, with an insignificant downward trend from 2003 to 2010, followed by an increasing trend from 2011 to 2019, particularly after 2016. Compared with the traditional ET monitoring methods and products, the ET estimated by GRACE/GRACE-FO observations provides a new way to effectively obtain continuous and reliable ET data in a wide range of river basins.

1. Introduction

Evapotranspiration (ET) includes evaporation from soil and water surfaces and plant transpiration, among others [1]. It is an important part of ecosystems and is a dominant component of the global water and energy cycle [2]. ET has important value when applied to the rational distribution of water resources and drought monitoring [3,4]. Because ET is one of the most important components of the climate system—connecting the water, energy, and carbon cycles—ET changes can be used as an indicator of climate change, especially in areas where the water cycle is accelerated. In recent decades, ET has often been used to evaluate the changes in regional drought characteristics [5,6]. The Yellow River Basin (YRB) is a major river basin in China that is the most seriously affected by drought [7]. Therefore, studying ET has important theoretical and practical significance for understanding the hydrological cycle in the YRB.
The YRB is located between 32–42° N and 96–119° E. The full length of the river basin is 5464 km, and the area is approximately 795,000 km 2 . The terrain of the YRB gradually decreases from west to east [8]. Figure 1, in which the main stream of the Yellow River and the Lijin hydrological station are marked, shows the boundary of the YRB.
Using different techniques and methods, several studies have estimated ET in different scales and regions around the world, such as Illinois [9], the whole territory of United States [10], the Amazon Basin [11], the extended Salado Basin [12], the Colorado River Basin [13], the Haihe River basin [14,15], the Ganga River basin [16], the North China Plain [17], and even at the global scale [18,19,20,21,22,23,24].
For the YRB, researchers also have used different methods and data to study the characteristics of ET in the YRB. The daily ET changes of the YRB was from 0 to 5 mm and it changes with the latitude and topography [25]. The spatial distribution of ET presented a spatial pattern similar to that of precipitation, and a significant decreasing trend was detected in the YRB from 1961 to 2006 [26]. The variation in the precipitation amount might largely affect the spatial and temporal patterns of the actual ET in the YRB [8]. Potential ET showed a significant decreasing trend, but the increasing and decreasing trends were different in different areas of the YRB from 1961 to 2012 [27]. Five ET products in the YRB from 1982 to 2000 obviously exhibit different performances, in terms of either their magnitude or temporal variations, which are mainly due to the quality of precipitation forcing data [28]. The ET in the YRB from 2003 to 2015 based on an operational simplified surface energy balance model and other 8 ET products could explain 23–52% of the variability in the water balance ET for 14 small catchments in the YRB, and the free global ET product derived from the simplified surface energy balance model highly underestimated the annual total ET trend for the YRB [29]. The main factors affecting pan evaporation (1961–2019) in the YRB were daily range of average annual temperature, sunshine duration, annual precipitation, and relative humidity based on observation data from meteorological stations [30].
However, at present, the above-mentioned studies on the ET of the YRB have mainly focused on obtaining the ET data of the YRB using different models based on observation data or using models for the evaporating dish at ground meteorological stations, including surface energy balance system (SEBS) model, two parameter steady-state model, and variable infiltration capacity (VIC) model. The long-term trend in variations in ET in the YRB and related factors affecting this were obtained. The ET data calculated by these models were further compared with those obtained from various other ET products. From the long-term time series, these studies all show that the ET of the YRB is on a downward trend, but the actual ET and evaporating dish evaporation showed a consistent trend in humid areas; however, the actual ET increased, and the evaporating dish evaporation decreased in arid and semi-arid areas in China [31]. In addition, the ground monitoring stations can provide ET observations at each site [32], but the sites are often too sparse for basin-scale studies. The considerable advantages of the GRACE gravity satellites at detecting long-term hydrological change information over large scales [33]—particularly the successful launch and application of the subsequent GRACE Follow-On (GRACE-FO) satellites—provide an important means for the continued use of gravity satellites to measure large-scale changes in regional land water storage [34]. Taking into account the unique nature of GRACE observations, Rodell et al. firstly proposed an equation to estimate ET by using observed data [5]. Compared with the traditional land surface models, the new method [5] was further demonstrated to be well applied to the estimation of ET and water modeling on regional scales, and may help in evaluating modeled ET [35].
GRACE data and other measured data also have been combined to generate estimates of actual ET based on the water budget equation in many regions [36,37,38,39,40,41], such as the Lake Chad Basin [42] and West Xiliao River Basin [6].
Therefore, in this study, we combined GRACE/GRACE-FO RL06 Mascon, precipitation, and runoff data to calculate the long-term time series of ET in the YRB from May 2002 to June 2020 based on the water budget equation, and further compared it with four other ET products: Global Land Data Assimilation Systems (GLDAS), Global Land Evaporation Amsterdam Model (GLEAM), European Centre for Medium-Range Weather Forecasts Reanalysis 5 (ERA5), and Complementary Relationship (CR). The comparison results verified the reliability of combining GRACE gravity satellites and measured data to obtain the actual ET characteristics in the YRB and evaluated the applicability of different ET products. This study provides an effective way to obtain continuous ET data for a wide range of river basins.

2. Data

2.1. GRACE and GRACE-FO Data

We calculated the terrestrial water storage anomaly (TWSA) and terrestrial water flux in the YRB based on GRACE and GRACE-FO RL06 Mascon data (https://grace.jpl.nasa.gov/data/get-data/jpl_global_mascons/, accessed on 12 April 2021), and combined this with the water budget equation to evaluate ET characteristics. GRACE RL06 Mascon data with a spatial resolution of 0.5° for the period April 2002–June 2017 were selected. Some missing data were filled using cubic spline interpolation. GRACE-FO RL06 Mascon data with a spatial resolution of 0.5° for the period October 2018–July 2020 were selected. In addition, the discontinuous data between GRACE and GRACE-FO (June 2017–October 2018) were supplemented by a dataset of land water reserve changes that was reconstructed based on precipitation in China, as released by Zhong et al. [34]. This dataset included GRACE/GRACE-FO RL06 Mascon data released by the Center for Space Research (CSR), real-time analysis system data of daily grid precipitation in China, CN05.1 temperature data, and other datasets. As the dataset was established using a precipitation reconstruction model and considering factors such as seasonal items in the Mascon product and trend items, the quality of the data is generally high [34].

2.2. Precipitation and Runoff Data

The precipitation data used in this study were obtained from the China Meteorological Administration website (http://data.cma.cn/, accessed on 17 November 2020). Monthly data for China’s surface precipitation grid dataset product (V2.0) with a spatial resolution of 0.5° for the period January 2000–July 2020 were selected. The Lijin hydrological station is the last hydrological station before the main stream of the Yellow River flows into the Bohai Sea; therefore, the runoff data from this station were selected to represent the net outflow runoff of the YRB. The runoff data were obtained from the Yellow River Sediment Bulletin (http://www.yrcc.gov.cn/zwzc/gzgb/gb/nsgb/, accessed on 13 October 2021). As the runoff data from the Lijin hydrometric station represented the net outflow runoff of the YRB, this study used monthly runoff data from this station from January 2000 to July 2020.

2.3. ET Products

2.3.1. GLDAS Products

The GLDAS product was jointly established by NASA’s Goddard Space Flight Center and the National Center for Environmental Prediction (NCEP). The product was obtained using a land surface model and data assimilation technology driven by satellite and ground observation data [43]. For this study, GLDAS-NOAH-2.1 data with a spatial resolution of 0.25° from April 2002 to May 2020 were selected (https://hydro1.gesdisc.eosdis.nasa.gov/data/GLDAS/, accessed on 23 July 2021).

2.3.2. GLEAM Products

The GLEAM surface ET product was jointly established by the Free University of Amsterdam in the Netherlands and the National University of Ghent, Belgium. The product used three different types of underlying surfaces: bare soil, high vegetation, and low vegetation. The actual ET of the product was obtained by multiplying the optical depth of the vegetation and the moisture of the soil root system to establish an evaporation forcing factor [44]. In this study, GLEAM v3.5a data with a spatial resolution of 0.25° for the period April 2002–December 2020 were selected (http://www.gleam.eu, accessed on 25 December 2021).

2.3.3. ERA5 Reanalysis Products

The ERA5 reanalysis product was the fifth-generation reanalysis data of the European Centre for Medium-Range Weather Forecasts (ECMWF). The product adopted the 4D data assimilation technology of the ECMWF comprehensive forecasting system, with 137 isobaric surface data with an interval of 0.01 hPa in the vertical direction [45]. In this study, the monthly mean ET data for the surface layer with a spatial resolution of 0.25° for the period April 2002–December 2020 were selected (https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5, accessed on 18 October 2021).

2.3.4. CR Products

The CR ET product was established using the complementary ET method. The input data of the product included computational magneto-fluid dynamics (CMFD) downward shortwave radiation, downward longwave radiation, air temperature, and air pressure, as well as Global Land Surface Satellite (GLASS) surface emissivity and albedo, ERA5-land surface temperature and air humidity, and National Centers for Environmental Prediction (NCEP) scattered emissivity [46,47]. These products were available for the period 1982–2017. Therefore, in this study, monthly data with a spatial resolution of 0.1° for the period 2002–2017 were selected (http://data.tpdc.ac.cn/en/data/b6d9f525-5b76-48b0-82db-bb2963465cac/, accessed on 1 July 2021).

3. Methods

3.1. Water Budget Equation

For a closed watershed system, the actual ET in the watershed can be calculated based on the water budget equation using the land water storage change value estimated by the GRACE satellites, combined with hydrometeorological data, such as precipitation and runoff. The actual ET can be calculated as follows [42,48,49]:
E T = P − R − d s d t
d s d t ≈ T W S A ( t + 1 ) − T W S A ( t − 1 ) 2 Δ t
where ET is the actual ET of the river basin (mm); P is the monthly precipitation (mm); R is the monthly runoff depth (mm); ds/dt is the change in terrestrial water storage for a specific time period; Δt is the time resolution of the GRACE/GRACE-FO; TWSA(t + 1) and TWSA(t − 1) are the terrestrial water storage anomalies corresponding to the river basin in the (t + 1) month and (t − 1) month, respectively. It should be noted that the rigorous form of Equation (2) was proposed by Rodell et al. [5]. The Equation (2) used in this study is a linearly approximation to the water mass variations for month t [42], which is corresponding to monthly accumulated precipitation and runoff data.

3.2. Mann–Kendall (MK) Trend Test Method

In this study, the trend change of actual ET in the YRB was revealed using the Mann–Kendall (MK) trend test method [50,51,52]. This method has the advantage of not requiring data to follow a normal distribution and is not affected by a few outliers; however, it does require sequences to be random and independent, with the same probability distribution, and can relatively objectively determine the trend of long-term series data and test mutation points [53]. Therefore, this method has been widely used in trend testing and mutation analysis of precipitation, temperature, and ET [54,55,56,57].
For the time series x with n samples, the order column S k can be calculated as follows:
S k = ∑ i = 1 k r i ,       k = 2 ,   3 , ⋯ , n ,
where r i is as follows:
r i = { + 1    x i > x j + 0    x i ≤ x j      j = 1 ,   2 , ⋯ , i .
The order column S k is the cumulative value of the number of values at time i that are greater than that at time j. The statistic U F k can be defined under the condition of random independence of the time series, as follows:
U F k = [ S k − E ( S k ) ] v a r ( S k )      k = 1 , 2 , ⋯ , n ,  
where U F 1 = 0 , E ( S k ) and v a r ( S k ) are the mean and variance of the S k , respectively.
When the x 1 , x 2 , ⋯ , x n are independent and have the same continuous distribution, E ( S k ) and v a r ( S k ) can be calculated as follows:
{ E ( S k ) = k ( k − 1 ) 4 v a r ( S k ) = k ( k − 1 ) ( 2 k + 5 ) 72       k = 2 , 3 , ⋯ , n ,
where U F k is the standard normal distribution that can be calculated according to the time series order x 1 , x 2 , ⋯ , x n . Then, the above process was repeated in the reverse order of time series x n , x n − 1 , ⋯ , x 1 , and we made U B k = − U F k ( k = n , n − 1 , ⋯ , 1 ) ,   U B 1 = 0 . Given the significance level α , for example α = 0.05 , the critical value u 0.05 = ± 1.96 , the U F k and U B k statistics curves, and ± 1.96 straight lines can be drawn on the same graph. This study used the significance level α = 0.05 to detect the annual ET trend.
The principle of time series change trend and mutation point test are as follows [58,59]. If the U F k changes within the range of the critical value, the change trend and mutation are not significant. If the U F k value is >0, the sequence shows an upward trend; otherwise, it shows a downward trend. When the U F k value exceeds the critical line, it indicates a significant upward or downward trend. If the U F k and the U B k curves intersect within the critical line, the time corresponding to the intersection is the time at which the mutation begins.

4. Results and Analysis

4.1. TWSA, Precipitation, and Runoff Characteristics in the YRB

The TWSA of the YRB for the period May 2002–June 2020 can be calculated based on GRACE and GRACE-FO RL06 Mascon data [60] (Figure 2). Monitoring results from GRACE and GRACE-FO showed that the TWSA presented characteristics of significant interannual variation. The terrestrial water flux for the period May 2002–June 2020 could be further calculated according to Equation (2). A comparison of the TWSA and terrestrial water fluxes is shown in Figure 2. The monthly runoff depth data of the YRB were obtained by dividing the runoff at the Lijin hydrological station by the area of the YRB. A comparison of the net outflow runoff and monthly runoff depth data is shown in Figure 3.
The TWSA of the YRB showed a continuous decline, at a rate of 6.47 mm/month from May 2002 to June 2020 (Figure 2), which may be related to the reduction in groundwater in the North China Plain in the middle and lower reaches of the river basin, because the TWSA of the river basin will certainly be affected by the strong changes around it, particularly in large-scale regions [61]. Precipitation and runoff in the YRB exhibited significant seasonal changes from April 2002 to July 2020 (Figure 3). Based on the change in monthly precipitation and runoff and considering that the precipitation data and runoff data of the station are normally maintained, the uncertainty of precipitation and measured runoff is usually estimated as 10% and 5%, respectively [5,15]. The mean annual precipitation of the YRB from 2003 to 2019 was up to 490.29 ± 49.0 mm, and the interannual change in precipitation was low. The mean annual runoff depth of the YRB from 2003 to 2019 was 23.93 ± 1.2 mm. Overall, the runoff of the YRB increased with an increase in precipitation. For example, the precipitation of the YRB reached a multi-year maximum in 2013. The rapid increase in precipitation led to a significant increase in runoff during the same period, and the increase in precipitation in 2018 also caused a significant increase in runoff during the same period (Figure 3).

4.2. Comparison of ET Estimated by GRACE/GRACE-FO with ET from Other Products

To obtain the time series characteristics of the actual ET estimated by GRACE/GRACE-FO and the other four ET products, the ET of the YRB from May 2002 to June 2020 was calculated based on the water budget equation combined with terrestrial water flux, precipitation, and runoff data. This is because in Equation (1), the monthly precipitation (P), monthly runoff depth (R), and ds/dt are observed independently of each other. Assuming that they are normally distributed, the ET error can be calculated based on the error propagation law [6,62]. The error estimate of terrestrial water storage changes is twice the root of the error estimation of TWSA [6,62]. The error estimation of TWSA can be obtained from the JPL Mason data. The ET and its uncertainty, estimated by GRACE/GRACE-FO, are shown in Figure 4. The GRACE/GRACE-FO ET, hereafter, will be referred to as GRACE ET. A comparison of the five ET time series in the YRB is shown in Figure 5.
From the overall trend of the five ET time series shown in Figure 5, the variation trend of ET estimated by GRACE/GRACE-FO is in good agreement with that of the other four ET products, showing obvious seasonal variation characteristics.
The five ET time series were highly consistent in the low-value area (0–60 mm), as shown in Figure 5. In the high-value area (>60 mm), the GLDAS, CR, and ERA5 ET products were basically consistent with the ET estimated by GRACE/GRACE-FO. For the period 2003–2019, the multi-year mean of ET estimated by GRACE/GRACE-FO was 469.05 ± 28.05 mm, whereas the multi-year mean values of GLDAS, GLEAM, and ERA5 were 448.78 ± 16.56 mm, 425.54 ± 16.41 mm, and 500.41 ± 22.85 mm, respectively. The annual cumulative ET of GLEAM was approximately 43.5 mm lower than the results estimated by GRACE/GRACE-FO. This indicates that the ET products of GLEAM underestimate the actual ET of the YRB.
To clearly obtain the characteristics of the different numerical intervals between the ET estimated by GRACE/GRACE-FO and the other four ET products, the scatter points and a fitting diagram for the comparisons are given in Figure 6.
As shown in Figure 6, the estimated ET based on the water budget equation (GRACE ET) and the other four ET products (GLDAS ET, ERA5 ET, GLEAM ET, and CR ET) maintained good correlation and consistency from 2003 to 2019. Among them, the highest correlation observed between GRACE ET and GLDAS ET (Figure 6b) (r = 0.90; RMSE = 13.44 mm; relative BIAS = −3.5%) followed by the correlation between GRACE ET and ERA5 ET (Figure 6d) (r = 0.89; RMSE = 14.70 mm; relative BIAS = 8.0%) and the correlation between GRACE ET and GLEAM ET (Figure 6a) (r = 0.87; RMSE = 16.03 mm; relative BIAS = −8.6%). While the relative lower correlation was found between GRACE ET and CR ET (Figure 6c) (r = 0.86; RMSE = 16.74 mm; relative BIAS = −6.9%).
To obtain the relationship between precipitation and ET on a monthly scale and interannually, the monthly mean ET and monthly mean precipitation of the five ET results (Figure 7a), and the annual cumulative ET and annual cumulative precipitation the five ET products (Figure 7b) were calculated.
It can be seen from Figure 7a that the five ET results and precipitation gradually increased from January to July, reached a peak in July, and gradually decreased from July to December. The GLEAM products underestimated ET from July to August, when ET was strong. Figure 7b shows that the annual cumulative ET estimated by GRACE/GRACE-FO was more consistent with the GLDAS and CR ET products from 2003 to 2011, whereas it was more consistent with ERA5 ET products from 2012 to 2019. Annual mean precipitation of the YRB was 490.29 ± 49.03 mm from 2003 to 2019. During this period, the multi-year mean of ET of GRACE/GRACE-FO was 469.05 ± 28.05 mm, and the multi-year mean values of GLDAS, ERA5, and GLEAM were 448.78 ± 16.56 mm, 500.41 ± 22.85 mm, 425.54 ± 16.41 mm, respectively. The estimated annual mean ET of CR was 435.61 ± 17.48 mm from 2003 to 2017.
It can be clearly seen that, the difference between the ET estimated by GRACE/GRACE-FO and the other 4 ET products (GLDAS, ERA5, GLEAM, and CR) and precipitation was no more than 70 mm. Therefore, it can be concluded that ET in the YRB was closely related to precipitation.

4.3. Interannual and Seasonal ET Variations in the YRB

The MK trend test method was used to test the annual cumulative ET estimated by GRACE/GRACE-FO in the YRB from 2003 to 2019, and the mutation point was obtained in approximately 2011 (Figure 8). Therefore, the time series of the annual cumulative ET was divided into two parts, 2003–2010 and 2011–2019, and piecewise linear fitting was carried out (Figure 8). Comparisons of mean annual precipitation and annual cumulative ET of different products in different time periods are summarized in Table 1, where the uncertainty is the standard deviation between the multi-year accumulated ET.
The U F k curve values did not exceed the significance level ( α = 0.05 ) before 2016, indicating that the ET of the YRB showed an insignificant downward trend from 2003 to 2016. As shown in Figure 8, the intersection of curve U F k and curve U B k occurred around 2011, indicating a mutation point in 2011. The corresponding values of the U F k curve exceeded the significant level ( α = 0.05 ) after 2016; i.e., the ET of the YRB increased significantly from 2016 to 2019.
According to the mutation point that appeared in 2011, we further calculated the changes in ET in the YRB during 2003–2010 and 2011–2019. The mean ET of the YRB was 450.26 ± 26.92 mm from 2003 to 2010 (Table 1). The linear regression results showed that the ET presented a downward trend from 2003 to 2010 with a decline rate of 0.76 mm/a (Figure 8); however, it failed to pass the significance level of 0.05, indicating that the ET presented an insignificant linear downward trend from 2003 to 2010. This conclusion is consistent with the ET change trend of the YRB obtained by Tong et al. based on the VIC model [30].
The mean ET of the YRB was 485.75 ± 29.05 mm from 2011 to 2019 (Table 1). The linear regression results showed that the ET increased overall from 2011 to 2019 at a rate of 3.25 mm/a (Figure 9); however, it failed to pass the significance level (0.05) before 2016 and passed after 2016. This indicates that the ET presented an insignificant linear increasing trend from 2011 to 2016 and a significant increasing trend from 2016 to 2019.
It can also be seen from Table 1 that the mean annual precipitation in the YRB in 2003–2010 was less than that in 2011–2019, and the corresponding GRACE/GRACE-FO ET in 2003–2010 was less than that in 2011–2019. The GRACE/GRACE-FO ET, GLDAS ET, and ERA5 ET were approximately equivalent to precipitation, which also indicated that ET in the YRB was closely related to precipitation.
Vegetation cover types vary in different seasons, and ET also presents different characteristics in different seasons [63]. Under climate and vegetation cover change conditions, studying the seasonal variation of ET has important reference value for reasonably adjusting structures of agricultural production and optimizing planting structures. Therefore, it was essential to determine the ET of the YRB during different seasons. Then, the time series of the mean ET of the YRB in spring, summer, autumn, and winter were calculated (Figure 10).
It can be seen from Figure 10 that the ET of the YRB in spring estimated by GRACE/GRACE-FO was between 76.53 mm and 114.25 mm from 2003 to 2019, and that the of multi-year mean ET value was 92.64 mm. The ET of the YRB in summer was between 178.56 mm and 298.20 mm from 2003 to 2019, and the multi-year mean ET value was 230.53 mm. The ET of the YRB in autumn was between 97.51 mm and 167.81 mm from 2003 to 2019, and the multi-year mean ET value was 123.96 mm. The ET of the YRB in winter was between 3.34 mm and 41.95 mm from 2003 to 2019, and the multi-year mean ET value was 21.92 mm.
The multi-year mean ET values in the four seasons of the YRB decreased in the following order: summer > autumn > spring > winter (Figure 10). This is because the low temperature and low precipitation in winter led to the minimum ET. In spring, the temperature and precipitation increased gradually, and the ET of vegetation and the evaporation of soil water were also strengthened, resulting in greater ET in spring than in winter. In summer, the temperature was the highest, precipitation was abundant, and vegetation transpiration and soil water evaporation were the strongest. Therefore, ET was the highest in summer. The decrease in temperature and precipitation in autumn, the reduction in vegetation transpiration, and the weakening of soil water evaporation resulted in lower ET in autumn than in summer.
As can be seen from Figure 10, the change in the ET trend in spring, autumn, and winter of the YRB from 2003 to 2019 was not obvious; furthermore, the change in the ET trend in summer was not obvious before 2015 but showed an increasing trend after 2015, reaching a maximum in summer, 2018.

5. Discussion

5.1. Analysis of the Differences in the ET Estimated by Different ET Products

Differences in the ET estimated by different ET products (Figure 5) might be because there are differences in the data types and methods used for the calculation of different ET products and because these models use different assumptions and structures. In this study, we exclusively used three types of ET products for comparisons with the GRACE ET. One type is the diagnostic model ET product, which is uses remote sensing vegetation information as important input data and is calculated using traditional estimation methods. For this type, GLEAM v3.5a ET products and the Complementary Relationship (CR) ET were used in this study. GLEAM uses a set of algorithms to separately estimate the different components (transpiration, bare-soil evaporation, interception loss, open-water evaporation, and sublimation) of land ET. The Priestley and Taylor equation were used in GLEAM to calculate potential evaporation based on observations of surface net radiation and near-surface air temperature. The rationale of GLEAM is to maximize the recovery of information on evaporation contained in current satellite observations of climatic and environmental variables [62]. CR ET is derived from the CR method, with the spatial coverage being the Chinese land area [46,47]. This ET dataset is only for land surfaces. The estimation model based on the CR principle does not require complex underlying surface data and can estimate the actual ET with high accuracy only with conventional meteorological data.
The second type is the land mode ET product. For this type, the GLDAS-NOAH-2.1 ET product with a spatial resolution of 0.25° was used in this study. The goal of the GLDAS is to generate optimal fields of land surface states and fluxes, by utilizing satellite- and ground-based observational data products, using advanced land surface modeling and data assimilation techniques [5]. GLDAS drives multiple, offline (not coupled to the atmosphere) land surface models, integrates a huge quantity of observation-based data, and executes globally at high resolutions, enabled by the Land Information System (LIS) [64]. The Noah model is one of the four land surface models driven by GLDAS. The GLDAS-2.1 model simulation requires National Oceanic and Atmospheric Administration (NOAA)/Global Data Assimilation System (GDAS) atmospheric analysis fields [65], the disaggregated Global Precipitation Climatology Project (GPCP) precipitation fields [66], as well as the Air Force Weather Agency’s AGRicultural METeorological modeling system (AGRMET) radiation fields.
The third type is the reanalysis ET product. The ERA5 ET reanalysis product was used in this study. ERA5 benefits from a decade of developments in model physics, core dynamics, and data assimilation. ERA 5 contains estimates of atmospheric variables, such as air temperature, pressure, and wind at different altitudes, as well as surface variables, such as precipitation, soil moisture content, and ocean wave heights. The feature of this product is the use of more historical observation data, especially satellite data, with advanced data assimilation and model systems to more accurately estimate atmospheric conditions.
It can be seen from this analysis that the data types, data structures, assumptions, or calculation methods of different ET products are different to a certain extent. For example, the ET products of GLDAS and GLEAM do not consider the input of artificial irrigation water. The CR ET products consider only the actual ET on land and do not include the differences between different ET products caused by the ET from the water surface. Furthermore, even with stringent data quality control and a high-performing model and data assimilation system, different ET products are associated with varying degrees of uncertainty. It should be noted that there are also some connections between the different ET products. For example, both GLDAS ET and ERA5 ET are related to satellite data and assimilate GRACE data. However, quantifying ET remains a challenging task since it is controlled by complicated interactions among atmospheric demand, soil water status, and typically heterogeneous vegetation and soil properties [67].

5.2. Trend in Interannual ET Variations Revealed by the MK Mutation Method in the YRB

The MK test method is a rank-based nonparametric test. Its advantage is that it can test linear or nonlinear trends [68,69] and it is suitable for nonlinear distribution data such as meteorology and hydrology data. The mutation point indicates that the changing trend before and after the time series is different. According to the MK mutation test results, the ET of the YRB mutated in 2011 (Figure 8). The main reasons for this might be as follows.
It can be seen from Figure 7b that the annual cumulative precipitation of the YRB remained at a low level before 2011 and a high level after 2011, which is basically consistent with the variation in the ET in the YRB (Figure 9). Therefore, it can be concluded that the long-term mean change in precipitation before and after 2011 is an important factor for the change in the ET in the YRB. Relevant evidence can be found in some references. For example, Wang et al. analyzed the correlation coefficient between the ET and precipitation calculated by different ET products in the YRB, indicating that the change in the ET is closely related to precipitation [26]. Liu et al. obtained the conclusion that precipitation dominates the change in actual ET in the YRB by analyzing the changing trend of the actual ET and potential ET and precipitation in the YRB [28].
According to the literature, the annual mean temperature in the YRB remained stable before 2011 and increased significantly after 2011 [28]. Therefore, it is inferred that the long-term mean change in temperature is also an important factor for the sudden change in the ET in the YRB around 2011. Similar conclusions can be found in the related references. For example, Edoga et al. analyzed the relationship between temperature and ET in Minna from 1996 to 2006, and found that temperature has an important effect on the ET [70]. Liu et al. analyzed the changing trend of reference ET and surface temperature in Northwest China from 1960 to 2010 and found that surface temperature will have an effect on the reference ET [71].
Another possibly reason for the ET of the YRB mutating in 2011 might be related to the changes of annual vegetation net primary productivity (NPP) and vegetation coverage. The NPP refers to the total amount of organic matter accumulated by green plants through photosynthesis in a unit time and unit area [72]. The long-term changes in annual vegetation NPP and vegetation coverage can reflect climate change to a certain extent. Ji et al. selected a time series of annual vegetation NPP and vegetation coverage from satellite data to analyze the vegetation greening from 2000 to 2019 in the YRB [73]. The results showed that the YRB experienced a rapid increase in temperature and precipitation during this period, and the annual NPP and coverage also showed an obviously increasing trend from 2000 to 2019; in particular, the annual NPP showed a significant oscillating acceleration trend after 2011. The rapid increases in temperature and precipitation allowed vegetation to flourish, as evidenced by significant positive correlations between climate variables and vegetation variables [73]. Therefore, it is inferred that the change in NPP and vegetation coverage might also be a factor underlying the mutated ET of the YRB in 2011.
It should be noted that the runoff of the Lijin hydrological station was used to represent the overall net outflow runoff of the YRB. Moreover, we can see from Figure 1 that the region from the Zhengzhou to Lijin is long and narrow, which could be less than the grid GRACE Mascon solutions and might have led to low accuracy in extracting the regional mean TWSA. However, our study area still retains this narrow area. The main consideration is to better estimate the ET of the entire YRB. It is necessary to select one that can best represent the outbound runoff of the entire YRB, and the Lijin hydrological station is the most representative hydrological station. In addition, this narrow area accounts for a small proportion relative to the entire YRB, which will have little impact on the results.
Additionally, the study area in this manuscript (within the red wireframe in Figure 1) contains a small part of the North China Plain (only part of the Shanxi basin), whereas the main part of the North China Plain is still outside of our study area (outside the red wireframe in Figure 1). In fact, the leakage effect will significantly affect the GWS signal, especially in the North China Plain [33]. In this study, we did not adopt the data outside the red wireframe when calculating the ET in the YRB. Furthermore, the TWSA of the YRB calculated based on the GRACE/GRACE-FO Mason data provided by JPL, which has its own leakage factor correction coefficient in the Shanxi basin.

6. Conclusions

In this study, the TWSA and terrestrial water flux of the YRB were calculated using GRACE and GRACE-FO RL06 Mascon data from April 2002 to July 2020. Combined with precipitation and runoff data, the change in actual ET in the YRB was estimated based on the water budget equation, which was compared with four other ET products. The time series of ET was further tested using the MK trend test method, and the following important results were obtained:
  • The actual ET in the YRB estimated by GRACE/GRACE-FO was in good agreement with the GLDAS, GLEAM, ERA5, and CR ET products. Among them, the GLDAS ET products had the highest consistency with the ET calculated by GRACE/GRACE-FO. The ET estimated by GLEAM in the seasons with low precipitation was in good agreement with the GRACE/GRACE-FO results; however, the difference was relatively large in the seasons with high precipitation.
  • ET in the YRB was closely related to precipitation. ET levels increased from January to July, with the increase in precipitation, and decreased from August to December with the decrease in precipitation. There was little difference between ET and precipitation in the YRB.
  • The annual mean ET of the YRB exhibited a sudden change around 2011; it showed an insignificant downward trend from 2003 to 2010 and an increasing trend from 2011 to 2019, especially after 2016. The multi-year mean ET values in the four seasons in the YRB decreased from summer > autumn > spring > winter.
This study emphasizes the unique ability of gravity satellites (GRACE and GRACE-FO) to monitor TWSA on a large scale. The variation characteristics of the ET in the YRB from 2002 to 2020 were effectively revealed using the GRACE/GRACE-FO data combined with precipitation and runoff data. A comparative analysis with other traditional ET products showed that this study provides a new way to effectively obtain continuous actual ET data in a large-scale river basin.

Author Contributions

Conceptualization, W.Q. and Q.Z.; methodology, W.Q., Z.J. and Y.G.; writing—original draft preparation, W.Q. and Z.J.; analyzed and interpreted the results, W.Q., Z.J., Y.G., P.Z. and P.C. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the national natural science foundation of China (NSFC) (42174006, 42090055), the opening fund of state key laboratory of geohazard prevention and geoenvironment protection of Chengdu university of technology (SKLGP2021K022), the Fundamental Research Funds for the Central Universities, CHD (300102261404).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

GRACE and GRACE-FO RL06 Mascon data can be downloaded from https://grace.jpl.nasa.gov/data/get-data/jpl_global_mascons/, accessed on 12 April 2021 [74]. Precipitation data can be downloaded from http://data.cma.cn/, accessed on 17 November 2020. Runoff data can be downloaded from http://www.yrcc.gov.cn/zwzc/gzgb/gb/nsgb/, accessed on 13 October 2021. GLDAS-NOAH-2.1 data can be downloaded from https://hydro1.gesdisc.eosdis.nasa.gov/data/GLDAS/, accessed on 23 July 2021. GLEAMv3.3a data can be downloaded from http://www.gleam.eu, accessed on 25 December 2021. ERA5 reanalysis products can be downloaded from https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5, accessed on 18 October 2021. CR products can be downloaded from http://data.tpdc.ac.cn/en/data/b6d9f525-5b76-48b0-82db-bb2963465cac/, accessed on 1 July 2021.

Acknowledgments

Some figures were prepared using the public domain Generic Mapping Tools—GMT. The authors thank Yulong Zhong, Jianlin Zhao, and Hailu Chen for their insightful suggestion and discussions. We also express our thanks to the two anonymous reviewers and academic editor for their constructive comments and suggestions that improved the manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Huo, Z.; Dai, X.; Feng, S.; Kang, S.; Huang, G. Effect of climate change on reference evapotranspiration and aridity index in arid region of China. J. Hydrol. 2017, 492, 24–34. [Google Scholar] [CrossRef]
  2. Alsdorf, D.E.; Lettenmaier, D.P. Tracking Fresh Water from Space. Science 2003, 301, 1491–1494. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  3. Abbasi, A.; Khalili, K.; Behmanesh, J.; Shirzad, A. Drought monitoring and prediction using SPEI index and gene expression programming model in the west of Urmia Lake. Theor. Appl. Climatol. 2019, 138, 553–567. [Google Scholar] [CrossRef]
  4. Miah, M.G.; Abdullah, H.M.; Jeong, C. Exploring standardized precipitation evapotranspiration index for drought assessment in Bangladesh. Environ. Monit. Assess. 2017, 189, 547. [Google Scholar] [CrossRef] [PubMed]
  5. Rodell, M.; Famiglietti, J.S.; Chen, J.; Seneviratne, S.I.; Viterbo, P.; Holl, S.; Wilson, C.R. Basin scale estimates of evapotranspiration using GRACE and other observations. Geophys. Res. Lett. 2004, 31, L20504. [Google Scholar] [CrossRef] [Green Version]
  6. Zhong, Y.; Zhong, M.; Feng, W.; Yan, H.; Li, C. Evaluation of the Evapotranspiration in the West Liaohe River Basin Based on GRACE Satellite and in Situ Measurements. J. Wuhan Univ. Inf. Sci. 2020, 45, 173–178, (In Chinese with English Abstract). [Google Scholar]
  7. Wang, F.; Wang, Z.; Yang, H.; Zhao, Y. Study of the temporal and spatial patterns of drought in the Yellow River basin based on SPEI. Sci. China Earth Sci. 2018, 61, 1098–1111. [Google Scholar] [CrossRef]
  8. Zhang, Q.; Xu, C.Y.; Chen, Y.D.; Ren, L. Comparison of evapotranspiration variations between the Yellow River and Pearl River basin, China. Stoch. Environ. Res. Risk Assess. 2011, 25, 139–150. [Google Scholar] [CrossRef]
  9. Yeh, P.J.F.; Famiglietti, J.S. Regional terrestrial water storage change and evapotranspiration from terrestrial and atmospheric water balance computations. J. Geophys. Res. 2008, 113, D09108. [Google Scholar] [CrossRef]
  10. Ferguson, C.R.; Sheffield, J.; Wood, E.F.; Gao, H. Quantifying uncertainty in a remote sensing-based estimate of evapotranspiration over the continental United States. Int. J. Remote Sens. 2010, 31, 3821–3865. [Google Scholar] [CrossRef]
  11. Karam, H.N.; Bras, R.L. Climatological basin-scale Amazonian evapotranspiration estimated through a water budget analysis. J. Hydrometeorol. 2008, 9, 1048–1060. [Google Scholar] [CrossRef]
  12. Cesanelli, A.; Guarracino, L. Estimation of regional evapotranspiration in the extended Salado Basin (Argentina) from satellite gravity measurements. Hydrogeol. J. 2011, 19, 629–639. [Google Scholar] [CrossRef]
  13. Castle, S.L.; Reager, J.T.; Thomas, B.F.; Purdy, A.J.; Lo, M.-H.; Famiglietti, J.S.; Tang, Q. Remote detection of water management impacts on evapotranspiration in the Colorado River basin. Geophys. Res. Lett. 2016, 43, 5089–5097. [Google Scholar] [CrossRef] [Green Version]
  14. Sun, C.; Ren, L. Assessment of surface water resources and evapotranspiration in the Haihe River basin of China using SWAT model. Hydrol. Processes 2013, 27, 1200–1222. [Google Scholar] [CrossRef]
  15. Pan, Y.; Zhang, C.; Gong, H.; Yeh, P.J.F.; Shen, Y.; Guo, Y.; Huang, Z.; Li, X. Detection of human-induced evapotranspiration using GRACE satellite observations in the Haihe River basin of China. Geophys. Res. Lett. 2017, 44, 190–199. [Google Scholar] [CrossRef]
  16. Syed, T.H.; Webster, P.J.; Famiglietti, J.S. Assessing variability of evapotranspiration over the ganga river basin using water balance computations. Water Resour. Res. 2014, 50, 2551–2565. [Google Scholar] [CrossRef] [Green Version]
  17. Cao, G.; Han, D.; Song, X. Evaluating actual evapotranspiration and impacts of groundwater storage change in the North China Plain. Hydrol. Processes 2014, 28, 1797–1808. [Google Scholar] [CrossRef]
  18. Fisher, J.B.; Tu, K.P.; Baldocchi, D.D. Global estimates of the land-atmosphere water flux based on monthly AVHRR and ISLSCP-II data, validated at 16 FLUXNET sites. Remote Sens. Environ. 2008, 112, 901–919. [Google Scholar] [CrossRef]
  19. Mu, Q.; Heinsch, F.A.; Zhao, M.; Running, S.W. Development of a global evapotranspiration algorithm based on MODIS and global meteorology data. Remote Sens. Environ. 2007, 111, 519–536. [Google Scholar] [CrossRef]
  20. Zhang, K.; Kimball, J.S.; Nemani, R.R.; Running, S.W. A continuous satellite-derived global record of land surface evapotranspiration from 1983 to 2006. Water Resour. Res. 2010, 46, W09522. [Google Scholar] [CrossRef] [Green Version]
  21. Vinukollu, R.K.; Meynadier, R.; Sheffield, J.; Wood, E.F. Multi-model, multi-sensor estimates of global evapotranspiration: Climatology, uncertainties and trends. Hydrol. Processes 2011, 25, 3993–4010. [Google Scholar] [CrossRef]
  22. Vinukollu, R.K.; Wood, E.F.; Ferguson, C.R.; Fisher, J.B. Global estimates of evapotranspiration for climate studies using multisensor remote sensing data: Evaluation of three process-based approaches. Remote Sens. Environ. 2011, 115, 801–823. [Google Scholar] [CrossRef]
  23. Wang, K.; Dickinson, R.E. A review of global terrestrial evapotranspiration: Observation, modeling, climatology, and climatic variability. Rev. Geophys. 2012, 50, RG2005. [Google Scholar] [CrossRef]
  24. Zeng, Z.; Wang, T.; Zhou, F.; Ciais, P.; Mao, J.; Shi, X.; Piao, S. A worldwide analysis of spatiotemporal changes in water balance-based evapotranspiration from 1982 to 2009. J. Geophys. Res. Atmos. 2014, 119, 1186–1202. [Google Scholar] [CrossRef]
  25. Yang, S.; Rui, S.; Mao, D.; Zhang, X.; Cui, J. Estimation of daily evapotranspiration in the Yellow River Basin by using MODIS data. IEEE Int. Geosci. Remote Sens. Symp. 2004, 2, 1343–1345. [Google Scholar]
  26. Liu, Q.; Yang, Z. Quantitative estimation of the impact of climate change on actual evapotranspiration in the Yellow River Basin, China. J. Hydrol. 2010, 395, 226–234. [Google Scholar] [CrossRef]
  27. Tong, R.; Yang, X.; Ren, L.; Liu, Y.; Ma, M. Temporal and spatial characteristics of evapotranspiration in the Yellow River Basin during 1961–2012 and analysis of its influence factors. Water Resour. Prot. 2015, 31, 16–21, (In Chinese with English Abstract). [Google Scholar]
  28. Wang, G.; Pan, J.; Shen, C.; Li, S.; Lu, J.; Lou, D.; Hagan, D. Evaluation of Evapotranspiration Estimates in the Yellow River Basin against the Water Balance Method. Water 2018, 10, 1884. [Google Scholar] [CrossRef] [Green Version]
  29. Yin, L.; Wang, X.; Feng, X.; Fu, B.; Chen, Y. A Comparison of SSEBop-Model-Based Evapotranspiration with Eight Evapotranspiration Products in the Yellow River Basin, China. Remote Sens. 2020, 12, 2528. [Google Scholar] [CrossRef]
  30. Zhuo, Y.; Zhao, H.; Wei, M.; Sui, X.; Liu, H. Temporal and Spatial Changes of Pan Evaporation and Its Influence Factors in the Yellow River Basin in Recent 59 Years. Yellow River 2021, 43, 28–34, (In Chinese with English Abstract). [Google Scholar]
  31. Liu, B.; Xiao, Z.; Ma, Z. Relationship between Pan Evaporation and Actual Evaporation in Different Humid and Arid Regions of China. Plateau Meteor. 2010, 29, 629–636, (In Chinese with English Abstract). [Google Scholar]
  32. Wang, Y.; Liu, B.; Zhai, J.; Su, B.; Luo, Y.; Zhang, Z. Relationship Between Potential and Actual Evaporation in Yangtze River Basin. Adv. Clim. Change Res. 2011, 7, 393–399. [Google Scholar]
  33. Feng, W.; Shum, C.; Zhong, M.; Pan, Y. Groundwater Storage Changes in China from Satellite Gravity: An Overview. Remote Sens. 2018, 10, 674. [Google Scholar] [CrossRef] [Green Version]
  34. Zhong, Y.; Feng, W.; Humphrey, V.; Zhong, M. Human-Induced and Climate-Driven Contributions to Water Storage Variations in the Haihe River Basin, China. Remote Sens. 2019, 11, 3050. [Google Scholar] [CrossRef] [Green Version]
  35. Rodell, M.; McWilliams, E.B.; Famiglietti, J.S.; Beaudoing, H.K.; Nigro, J. Estimating evapotranspiration using an observation based terrestrial water budget. Hydrol. Processes 2011, 25, 4082–4092. [Google Scholar] [CrossRef]
  36. Billah, M.M.; Goodall, J.L.; Narayan, U.; Reager, J.T.; Lakshmi, V.; Famiglietti, J.S. A methodology for evaluating evapotranspiration estimates at the watershed-scale using GRACE. J. Hydrol. 2015, 523, 574–586. [Google Scholar] [CrossRef] [Green Version]
  37. Madeleine, A.; Pascolini, C.; Reager, J.T.; Fisher, J.B. GRACE-based Mass Conservation as a Validation Target for Basin-Scale Evapotranspiration in the Contiguous United States. Water Resour. Res. 2020, 56, e2019WR026594. [Google Scholar]
  38. Liu, Y.; Yue, Q.; Wang, Q.; Yu, J.; Zheng, Y.; Yao, X.; Xu, S. A Framework for Actual Evapotranspiration Assessment and Projection Based on Meteorological, Vegetation and Hydrological Remote Sensing Products. Remote Sens. 2021, 13, 3643. [Google Scholar] [CrossRef]
  39. Chao, L.; Zhang, K.; Wang, J.; Feng, J.; Zhang, M. A Comprehensive Evaluation of Five Evapotranspiration Datasets Based on Ground and GRACE Satellite Observations: Implications for Improvement of Evapotranspiration Retrieval Algorithm. Remote Sens. 2021, 13, 2414. [Google Scholar] [CrossRef]
  40. Boronina, A.; Ramillien, G. Application of AVHRR imagery and GRACE measurements for calculation of actual evapotranspiration over the Quaternary aquifer (Lake Chad basin) and validation of groundwater models. J. Hydrol. 2008, 348, 98–109. [Google Scholar] [CrossRef]
  41. Liu, Y.; Mo, X.; Hu, S.; Chen, X.; Liu, S. Assessment of human-induced evapotranspiration with GRACE satellites in the Ziya-Daqing Basins, China. Hydrol. Sci. J. 2020, 65, 2577–2589. [Google Scholar] [CrossRef]
  42. Ramillien, G.; Frappart, F.; Güntner, A.; Ngo-Duc, T.; Cazenave, A.; Laval, K. Time variations of the regional evapotranspiration rate from Gravity Recovery and Climate Experiment (GRACE) satellite gravimetry. Water Resour. Res. 2006, 42, 1–8. [Google Scholar] [CrossRef] [Green Version]
  43. Rodell, M.; Houser, P.R.; Jambor, U. The global land data assimilation system. Bull. Am. Meteorol. Soc. 2004, 85, 381–394. [Google Scholar] [CrossRef] [Green Version]
  44. Martens, B.; Gonzalez Miralles, D.; Lievens, H.; Van Der Schalie, R.; De Jeu, R.A.M.; Fernández-Prieto, D.; Beck, H.E.; Dorigo, W.A.; Verhoest, N. GLEAM v3: Satellite-based land evaporation and root-zone soil moisture. Geosci. Model Dev. 2017, 10, 1903–1925. [Google Scholar] [CrossRef] [Green Version]
  45. Hersbach, H.; Bell, B.; Berrisford, P.; Hirahara, S.; Horányi, A.; Muñoz-Sabater, J.; Nicolas, J.; Peubey, C.; Radu, R.; Schepers, D.; et al. The ERA5 global reanalysis. Q. J. R. Meteorol. Soc. 2020, 146, 1999–2049. [Google Scholar] [CrossRef]
  46. Ma, N.; Szilagyi, J.; Zhang, Y.; Liu, W. Complementary-relationship-based modeling of terrestrial evapotranspiration across China during 1982–2012: Validations and spatiotemporal analyses. Geophys. Res. Atmos. 2019, 124, 4326–4351. [Google Scholar] [CrossRef]
  47. Ma, N.; Szilagyi, J. The CR of evaporation: A calibration-free diagnostic and benchmarking tool for large-scale terrestrial evapotranspiration modeling. Water Resour. Res. 2019, 55, 7246–7274. [Google Scholar] [CrossRef] [Green Version]
  48. Ferreira, V.G.; Gong, Z.; He, X.; Zhang, Y.; Andam-Akorful, S.A. Estimating total discharge in the Yangtze River Basin using satellite-based observations. Remote Sens. 2013, 5, 3415–3430. [Google Scholar] [CrossRef] [Green Version]
  49. Long, D.; Longuevergne, L.; Scanlon, B.R. Uncertainty in evapotranspiration from land surface modeling, remote sensing, and GRACE satellites. Water Resour. Res. 2014, 50, 1131–1151. [Google Scholar] [CrossRef] [Green Version]
  50. Kendall, M.G. A new measure of rank correlation. Biometrika 1938, 30, 81–93. [Google Scholar] [CrossRef]
  51. Hamed, K.H.; Rao, A.R. A modified Mann-Kendall trend test for autocorrelated data. J. Hydrol. 1998, 204, 182–196. [Google Scholar] [CrossRef]
  52. Kendall, M.G. Rank Correlation Methods. Br. J. Psychol. 1990, 25, 86–91. [Google Scholar] [CrossRef]
  53. Ketema, A.; Siddaramaiah, D.G. Trend and variability of hydrometeorological variables of Tikur Wuha watershed in Ethiopia. Arab. J. Geosci. 2020, 13, 142. [Google Scholar] [CrossRef]
  54. Guo, M.; Li, J.; He, H.; Xu, J.; Jin, Y. Detecting Global Vegetation Changes Using Mann-Kendall (MK) Trend Test for 1982–2015 Time Period. Chin. Geogr. Sci. 2018, 28, 907–919. [Google Scholar] [CrossRef] [Green Version]
  55. Shadmani, M.; Marofi, S.; Roknian, M. Trend Analysis in Reference Evapotranspiration Using Mann-Kendall and Spearman’s Rho Tests in Arid Regions of Iran. Water Resour. Manag. 2012, 26, 211–224. [Google Scholar] [CrossRef] [Green Version]
  56. Feng, W.; Lu, H.; Yao, T.; Yu, Q. Drought characteristics and its elevation dependence in the Qinghai-Tibet plateau during the last half-century. Sci. Rep. 2020, 10, 14323. [Google Scholar] [CrossRef] [PubMed]
  57. Wang, W.; Zhu, Y.; Xu, R.; Liu, J. Drought severity change in China during 1961–2012 indicated by SPI and SPEI. Nat. Hazards 2015, 75, 2437–2451. [Google Scholar] [CrossRef]
  58. Dinpashoh, Y.; Babamiri, O. Trends in reference crop evapotranspiration in Urmia Lake basin. Arab. J. Geosci. 2020, 13, 372. [Google Scholar] [CrossRef]
  59. Dey, S.; Bhatt, D.; Haq, S.; Mall, R.K. Potential impact of rainfall variability on groundwater resources: A case study in Uttar Pradesh, India. Arab. J. Geosci. 2020, 13, 114. [Google Scholar] [CrossRef]
  60. Yang, X.; Tian, S.; Feng, W.; Ran, J.; You, W.; Jiang, Z.; Gong, X. Spatio-Temporal Evaluation of Water Storage Trends from Hydrological Models over Australia Using GRACE Mascon Solutions. Remote Sens. 2020, 12, 3578. [Google Scholar] [CrossRef]
  61. Jing, W.; Yao, L.; Zhao, X.; Zhang, P.; Liu, Y.; Xia, X.; Song, J.; Yang, J.; Li, Y.; Zhou, C. Understanding terrestrial water storage declining trends in the Yellow River Basin. Geophys. Res. Atmos. 2019, 124, 12963–12984. [Google Scholar] [CrossRef]
  62. Zhong, Y.; Zhong, M. Evaluation of Evapotranspiration for Exorheic Catchments of China during the GRACE Era: From a Water Balance Perspective. Remote Sens. 2020, 12, 511. [Google Scholar] [CrossRef] [Green Version]
  63. Cheng, M.; Jiao, X.; Jin, X.; Li, B.; Liu, K.; Shi, L. Satellite time series data reveal interannual and seasonal spatiotemporal evapotranspiration patterns in China in response to effect factors. Agric. Water Manag. 2021, 255, 107046. [Google Scholar] [CrossRef]
  64. Kumar, S.V.; Peters-Lidard, C.D.; Tian, Y.; Houser, P.R.; Geiger, J.; Olden, S.; Lighty, L.; Eastman, J.L.; Doty, B.; Dirmeyer, P.; et al. Land information system: An interoperable framework for high resolution land surface modeling. Environ. Model Softw. 2006, 21, 1402–1415. [Google Scholar] [CrossRef]
  65. Derber, J.C.; Parrish, D.F.; Lord, S.J. The new global operational analysis system at the National Meteorological Center. Weather Forecast. 1991, 6, 538–547. [Google Scholar] [CrossRef] [Green Version]
  66. Adler, R.F.; Huffman, G.J.; Chang, A.; Ferraro, R.; Xie, P.; Janowiak, J.; Rudolf, B.; Schneider, U.; Curtis, S.; Bolvin, D.; et al. The Version 2 Global Precipitation Climatology Project (GPCP) Monthly Precipitation Analysis (1979–Present). J. Hydrometeor. 2003, 4, 1147–1167. [Google Scholar] [CrossRef]
  67. Or, D.; Lehmann, P. Surface Evaporative Capacitance: How Soil Type and Rainfall Characteristics Affect Global-Scale Surface Evaporation. Water Resour. Res. 2019, 55, 519–539. [Google Scholar] [CrossRef] [Green Version]
  68. Hisdal, H.; Stahl, K. Have stream flow droughts in Europe become more severe or frequent. Int. J. Climatol. 2001, 21, 317–333. [Google Scholar] [CrossRef]
  69. Wu, H.; Soh, L.K.; Samal, A.; Chen, X.H. Trend analysis of stream flow drought events in Nebraska. Water Resour. Manag. 2008, 22, 145–164. [Google Scholar] [CrossRef]
  70. Edoga, R.N.; Suzzy, A.B.U. Effect of Temperature Changes on Evapotranspiration in Minna, Niger State. J. Eng. Appl. Sci. 2012, 3, 482–486. [Google Scholar]
  71. Liu, X.; Zhang, D. Trend analysis of reference evapotranspiration in Northwest China: The roles of changing wind speed and surface air temperature. Hydrol. Processes 2013, 27, 3941–3948. [Google Scholar] [CrossRef]
  72. Wang, X.; Tan, K.; Chen, B.; Du, P. Assessing the Spatiotemporal Variation and Impact Factors of Net Primary Productivity in China. Sci. Rep. 2017, 7, 44415. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  73. Ji, Y.; Zhou, G.S.; Wang, S.; Wang, L. Prominent vegetation greening and its correlation with climatic variables in northern China. Environ. Monit. Assess. 2020, 192, 636. [Google Scholar] [CrossRef] [PubMed]
  74. Landerer, F.W.; Flechtner, F.M.; Save, H.; Webb, F.H.; Bandikova, T.; Bertiger, W.I.; Bettadpur, S.V.; Byun, S.H.; Dahle, C.; Dobslaw, H.; et al. Extending the global mass change data record: GRACE Follow-On instrument and science data performance. Geophys. Res. Lett. 2020, 47, e2020GL088306. [Google Scholar] [CrossRef]
Remotesensing 14 00730 g001 550
Figure 1. Boundary of the Yellow River Basin and the location of important hydrological stations.
Figure 1. Boundary of the Yellow River Basin and the location of important hydrological stations.
Remotesensing 14 00730 g001
Remotesensing 14 00730 g002 550
Figure 2. Time series of terrestrial water flux and terrestrial water storage anomaly in the Yellow River Basin. The light blue and light green shaded parts represent uncertainty for the ds/dt and TWSA, respectively.
Figure 2. Time series of terrestrial water flux and terrestrial water storage anomaly in the Yellow River Basin. The light blue and light green shaded parts represent uncertainty for the ds/dt and TWSA, respectively.
Remotesensing 14 00730 g002
Remotesensing 14 00730 g003 550
Figure 3. Time series of monthly precipitation and runoff depth in the Yellow River Basin.
Figure 3. Time series of monthly precipitation and runoff depth in the Yellow River Basin.
Remotesensing 14 00730 g003
Remotesensing 14 00730 g004 550
Figure 4. The ET and its uncertainty estimated by Gravity Recovery and Climate Experiment (GRACE)/GRACE Follow-On (GRACE/GRACE-FO) based on the water budget equation. Orange shadow represents the uncertainty of the ET. GRACE—Gravity Recovery and Climate Experiment.
Figure 4. The ET and its uncertainty estimated by Gravity Recovery and Climate Experiment (GRACE)/GRACE Follow-On (GRACE/GRACE-FO) based on the water budget equation. Orange shadow represents the uncertainty of the ET. GRACE—Gravity Recovery and Climate Experiment.
Remotesensing 14 00730 g004
Remotesensing 14 00730 g005 550
Figure 5. Comparison of five evapotranspiration (ET) time series in the Yellow River Basin. GLEAM—Global Land Evaporation Amsterdam Model; GLDAS—Global Land Data Assimilation Systems; CR—Complementary Relationship; ERA5—European Centre for Medium-Range Weather Forecasts Reanalysis 5; GRACE—Gravity Recovery and Climate Experiment.
Figure 5. Comparison of five evapotranspiration (ET) time series in the Yellow River Basin. GLEAM—Global Land Evaporation Amsterdam Model; GLDAS—Global Land Data Assimilation Systems; CR—Complementary Relationship; ERA5—European Centre for Medium-Range Weather Forecasts Reanalysis 5; GRACE—Gravity Recovery and Climate Experiment.
Remotesensing 14 00730 g005
Remotesensing 14 00730 g006a 550Remotesensing 14 00730 g006b 550
Figure 6. Comparison scatter plots among the Gravity Recovery and Climate Experiment (GRACE) evapotranspiration (ET) and other products in the Yellow River Basin. GRACE—Gravity Recovery and Climate Experiment; (a) GLEAM—Global Land Evaporation Amsterdam Model; (b) GLDAS—Global Land Data Assimilation Systems; (c) CR—Complementary Relationship; (d) European Centre for Medium-Range Weather Forecasts Reanalysis 5 (ERA5).
Figure 6. Comparison scatter plots among the Gravity Recovery and Climate Experiment (GRACE) evapotranspiration (ET) and other products in the Yellow River Basin. GRACE—Gravity Recovery and Climate Experiment; (a) GLEAM—Global Land Evaporation Amsterdam Model; (b) GLDAS—Global Land Data Assimilation Systems; (c) CR—Complementary Relationship; (d) European Centre for Medium-Range Weather Forecasts Reanalysis 5 (ERA5).
Remotesensing 14 00730 g006aRemotesensing 14 00730 g006b
Remotesensing 14 00730 g007 550
Figure 7. (a) Monthly mean and (b) annual cumulative evapotranspiration (ET) and precipitation in the Yellow River Basin. GRACE—Gravity Recovery and Climate Experiment; GLEAM—Global Land Evaporation Amsterdam Model; GLDAS—Global Land Data Assimilation Systems; CR—Complementary Relationship; ERA5—European Centre for Medium-Range Weather Forecasts Reanalysis 5.
Figure 7. (a) Monthly mean and (b) annual cumulative evapotranspiration (ET) and precipitation in the Yellow River Basin. GRACE—Gravity Recovery and Climate Experiment; GLEAM—Global Land Evaporation Amsterdam Model; GLDAS—Global Land Data Assimilation Systems; CR—Complementary Relationship; ERA5—European Centre for Medium-Range Weather Forecasts Reanalysis 5.
Remotesensing 14 00730 g007
Remotesensing 14 00730 g008 550
Figure 8. Mutation points detected by the Mann–Kendall (MK) trend test method. The U F k and U B k are two statistics calculated based on the MK trend test method.
Figure 8. Mutation points detected by the Mann–Kendall (MK) trend test method. The U F k and U B k are two statistics calculated based on the MK trend test method.
Remotesensing 14 00730 g008
Remotesensing 14 00730 g009 550
Figure 9. Time series of evapotranspiration (ET) calculated by Gravity Recovery and Climate Experiment (GRACE)/GRACE Follow-On (GRACE/GRACE-FO) in the Yellow River Basin from 2003 to 2019. GRACE—Gravity Recovery and Climate Experiment.
Figure 9. Time series of evapotranspiration (ET) calculated by Gravity Recovery and Climate Experiment (GRACE)/GRACE Follow-On (GRACE/GRACE-FO) in the Yellow River Basin from 2003 to 2019. GRACE—Gravity Recovery and Climate Experiment.
Remotesensing 14 00730 g009
Remotesensing 14 00730 g010 550
Figure 10. Time series of evapotranspiration (ET) calculated by Gravity Recovery and Climate Experiment (GRACE)/GRACE Follow-On (GRACE/GRACE-FO) in the Yellow River Basin in different seasons.
Figure 10. Time series of evapotranspiration (ET) calculated by Gravity Recovery and Climate Experiment (GRACE)/GRACE Follow-On (GRACE/GRACE-FO) in the Yellow River Basin in different seasons.
Remotesensing 14 00730 g010
Table
Table 1. Annual mean precipitation and annual cumulative evapotranspiration (ET) of different products in different periods in the Yellow River Basin (unit: mm). GRACE—Gravity Recovery and Climate Experiment; GLEAM—Global Land Evaporation Amsterdam Model; GLDAS—Global Land Data Assimilation Systems.
Table 1. Annual mean precipitation and annual cumulative evapotranspiration (ET) of different products in different periods in the Yellow River Basin (unit: mm). GRACE—Gravity Recovery and Climate Experiment; GLEAM—Global Land Evaporation Amsterdam Model; GLDAS—Global Land Data Assimilation Systems.
Time PeriodAnnual PrecipitationGRACE-ETGLEAM-ETGLDAS-ETERA5-ET
2003–2010 471.42   ±   47.14 450.26   ±   26.92 422.68   ±   14.26 442.09   ±   11.68 504.82   ±   24.21
2011–2019 507.07   ±   50.71 485.75   ±   29.05 428.08   ±   17.73 454.73   ±   18.57 496.49   ±   22.24
2003–2019 490.29   ±   49.03 469.05   ±   28.05 425.54   ±   16.41 448.78   ±   16.56 500.41   ±   22.85
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Qu, W.; Jin, Z.; Zhang, Q.; Gao, Y.; Zhang, P.; Chen, P. Estimation of Evapotranspiration in the Yellow River Basin from 2002 to 2020 Based on GRACE and GRACE Follow-On Observations. Remote Sens. 2022, 14, 730. https://doi.org/10.3390/rs14030730

AMA Style

Qu W, Jin Z, Zhang Q, Gao Y, Zhang P, Chen P. Estimation of Evapotranspiration in the Yellow River Basin from 2002 to 2020 Based on GRACE and GRACE Follow-On Observations. Remote Sensing. 2022; 14(3):730. https://doi.org/10.3390/rs14030730

Chicago/Turabian Style

Qu, Wei, Zehui Jin, Qin Zhang, Yuan Gao, Pufang Zhang, and Peinan Chen. 2022. "Estimation of Evapotranspiration in the Yellow River Basin from 2002 to 2020 Based on GRACE and GRACE Follow-On Observations" Remote Sensing 14, no. 3: 730. https://doi.org/10.3390/rs14030730

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop