Fig. 1 Location of the study site (red box).
The evaporation experiment was conducted from May 13, 2018 to June 13, 2019. During the test, surface vegetation was removed, and a rain shed was built to prevent rainfall infiltration, thereby maintaining continuous-evaporation conditions. Every two months, the surface 0-30 cm soil was sampled at 5 cm intervals by a soil auger, and a total of 42 samples were collected six times. All the samples were stored in a frozen state (−15℃). Moreover, a portable time-domain reflectometry soil-water measurement system (MPM-160) was used to measure the volumetric water content in the soil profiles.
Sample processing and determination were performed at the State Key Laboratory of Biogeology and Environmental Geology at the China University of Geosciences (Wuhan). The soil water was extracted using an automatic vacuum condensation extraction system (LI-2100) and pre-processed using a needle filter (filter membrane: 0.22 µm). For the experiment, a vacuum of approximately 2000 Pa was maintained and a heating temperature of 110°C was used. The vacuum level extraction time is 8 min, and the first replenishment vacuum time and cycle replenishment vacuum time are 300 min. The lower limit of the replenishment vacuum was 1200 Pa, and the upper limit of the replenishment vacuum was 2500 Pa. Lastly, the compositions of the stable hydrogen and oxygen isotopes in the samples were measured using a liquid water isotope analyzer (IWA-45-EP). The measurements were obtained by feeding an average of one standard sample for every three samples, and the result obtained for each tested sample and the standard sample was the average of six feed measurements. The standard error of measurement was ±0.2‰ and ±0.6‰ for δ18O and δ2H, respectively. The measured data were expressed in thousandths relative to the Vienna Standard Mean Ocean Water standard sample:
(1)
(2)
2.2 Meteorological data andδ18O and δ2H isotopes in precipitation
During the test period, the daily average temperature, relative humidity, and evaporation data of Wuhan were obtained from the National Oceanic and Atmospheric Administration Climate Prediction Center (https://www.cpc.ncep.noaa.gov/) and NASA Earth Observations (https://neo.sci.gsfc.nasa.gov/).
Rainfall is the only source of soil water recharge in the study area. A comprehensive understanding of the composition of precipitation isotopes in the study area helps us to determine the initial conditions of soil water isotopes. In this study, we collected 115 monthly average precipitation isotope data in the Wuhan area, spanning from 1986 to 2013, with a couple months of missing data, and the available data were divided into three parts: 23 data from January 1986 to October 1992, 27 data from January 1996 to May 1998, and 21 data from September 2011 to May 2013. Data from January 1986 to May 1998 were obtained from the International Atomic Energy Agency (IAEA) (https://nucleus.iaea.org/wiser/index.aspx). The monitoring site is the Wuhan meteorological station (114.13°E, 30.62°N) and the monitoring frequency is measured month by month. The standard used was the Vienna standard average seawater isotope value, and the test precision for δ18O and δ2H was ±0.1‰ and ±0.8‰, respectively. Data from September 2011 to May 2013 were obtained from (Deng et al. , 2016). The monitoring site is located at the Wuhan University Engineering Department and Wuhan Institute of Biological Products dormitory (114.36°E, 30.54°N), and the monitoring frequency is measured month by month. The standard used was the Vienna standard average seawater isotope value, and the test precision for δ18O and δ2H was ±0.2‰ and ±2‰, respectively.
2.3 Quantification of soil water evaporation using the C–G model
When the stable isotope composition of water changes only because of isotopic fractionation during the evaporation process, and information about the water undergoing evaporation (ambient temperature, relative humidity, and isotopic composition) is known, the non-steady-state model in the C–G model (Bennett et al. , 2008; Horita et al. , 2008) is often used to calculate the evaporation losses of open water bodies. Theoretically, this method is also suitable for evaluating soil water evaporation under continuous-evaporation conditions. In our study, soil-water evaporation losses under continuous-evaporation conditions were quantified via three field experiments conducted from May 13, 2018 to July 12, 2018; July 12, 2018 to October 19, 2018; and January 5, 2019 to April 20, 2019.
In this scenario, the evaporation loss fraction of the soil water volume (f) can be calculated using the following equation (Stephen K. Hamilton, 2005). On multiplying f by the thickness of the soil layer and subsequently dividing it by the number of test days, we obtain the corresponding evaporation loss (mm/d) at different depths.
(3)
where δP(soil) is the initial value of the soil-water sample isotope composition, δL(soil) is the final value of the soil-water sample isotope composition, δ* is the limiting factor for the isotope enrichment—this term is also known as the limit isotope composition enrichment (‰) (J. R. Gat, 1978)—and m denotes the correlation between the isotopic composition of evaporated water vapor and the isotopic composition of liquid water—this term is also known as the enrichment slope (G.B. Allison, 1982).
The change in the ratio of the difference between the soil-water isotope composition and the environmental limit isotope composition at the end and the initial time reflects the change in the remaining soil-water isotope composition value. A large value (for example, f = 0.40) indicates that 40% of the initial water content has evaporated.
(4)
where h is the average relative humidity (fraction), ε is the total isotope fractionation (‰), and δA is the stable isotopic composition of water in ambient air (‰).
The most accurate method for determining δA is based on on-site measurements. However, owing to the difficulty associated with direct measurements in the field, the δA value is generally estimated based on the stable isotopic composition of the local rainfall (Gat.J.R 1995; Gibson and Reid, 2014).
(5)
where is the average of the precipitation isotopes at the time of the two sampling periods before and after; ε+ and α+ are the equilibrium enrichment factor and equilibrium fractionation factor that change with temperature, respectively, as proposed by(Juske Horita, 1994):
(6)
2H: (7)
18O: (8)
where temperature (T) is expressed in Kelvin.
(9)
where is the total isotopic fractionation factor (‰) and is the kinetic isotopic fractionation factor (‰).
(10)
where n is a constant, which is related to the correlation between the molecular diffusion resistance and molecular diffusion coefficient and is usually considered to be 1 for non-mobile air layers (e.g., soil water evaporation and plant transpiration), θ is the ratio of the molecular diffusion coefficient to the total diffusion fractionation coefficient and is generally considered to be 1 for soil water evaporation. CD is a parameter that describes the diffusion efficiency of the molecule and has a value of 25.1‰ and 28.5‰ for hydrogen and oxygen, respectively (GONFIANTINI, 1986; John Crusius, 2000).
(11)
In this study, precipitation isotopes are used to reflect the environmental changes during evaporation, i.e., in Eq. (5). In this way, the during the test period can be determined by knowing the precipitation isotope values for both sampling periods. In the absence of precipitation isotope data for Wuhan during the test period, precipitation isotope data for Wuhan collected from network monitoring data from 1986 to 1998 were used to replace precipitation isotope data during the test period in order to calculate rain. As shown in Fig. 2, we conclude that precipitation isotope values for the same period from 1986 to 1998 and from 2018 to 2019 can be considered similar if the past and current monthly averages of precipitation and temperature parameters in the same month were similar in the same study setting. Therefore, we used the following equations to calculate δ18O and δ2H for the months that satisfied this condition.
(12)
(13)
(14)
(15)
In these formulas, n is the total number of data collected in a particular month; j is the month j; i is the data collected in a particular month i; rain is the rainfall in mm; T is the temperature in ℃; 18O, 2H, and T are the oxygen 18 isotope value, deuterium isotope value, and temperature, respectively; and , , , and are the corresponding monthly mean values.