Technical note: Evaporating water is different from bulk soil water in δ2H and δ18O and has implications for evaporation calculation

Soil evaporation is a key process in the water cycle and can be conveniently quantified using δ2H and δ18O in bulk surface soil water (BW). However, recent research shows that soil water in larger pores evaporates first and differs from water in smaller pores in δ2H and δ18O, which disqualifies the quantification of evaporation from BW δ2H and δ18O. We hypothesized that BW had different isotopic compositions from evaporating water (EW). Therefore, our objectives were to test this hypothesis first and then evaluate whether the isotopic difference alters the calculated evaporative water loss. We measured the isotopic composition of soil water during two continuous evaporation periods in a summer maize field. Period I had a duration of 32 d, following a natural precipitation event, and period II lasted 24 d, following an irrigation event with a 2H-enriched water. BW was obtained by cryogenically extracting water from samples of 0–5 cm soil taken every 3 d; EW was derived from condensation water collected every 2 d on a plastic film placed on the soil surface. The results showed that when event water was heavier than pre-event BW, δ2H of BW in period II decreased, with an increase in evaporation time, indicating heavy water evaporation. When event water was lighter than the pre-event BW, δ2H and δ18O of BW in period I and δ18O of BW in period II increased with increasing evaporation time, suggesting light water evaporation. Moreover, relative to BW, EW had significantly smaller δ2H and δ18O in period I and significantly smaller δ18O in period II (p < 0.05). These observations suggest that the evaporating water was close to the event water, both of which differed from the bulk soil water. Furthermore, the event water might be in larger pores from which evaporation takes precedence. The soil evaporative water losses derived from EW isotopes were compared with those from BW. With a small isotopic difference between EW and BW, the evaporative water losses in the soil did not differ significantly (p > 0.05). Our results have important implications for quantifying evaporation processes using water stable isotopes. Future studies are needed to investigate how soil water isotopes partition differently between pores in soils with different pore size distributions and how this might affect soil evaporation estimation.


Samples collection and measurement 100
In order to determine the water isotopic composition in EW from condensation water of the evaporation vapor, we randomly selected three rectangular areas of 40 cm long and 30 cm wide. A channel of 3 cm deep was dug around the edge of the area (Fig. 1). Subsequently, a piece of plastic film without hole (about 0.2 m 2 , 40 cm by 50 cm) was used to cover the soil surface, with an extra 5 cm at each side. Then the channels were back filled with soil to keep the covered area free of the effect of wind. After 105 equilibrium for two days, the condensation water adhered on the underside of the plastic film was collected using an injection syringe in the early morning at about 7 a.m. to eliminate the secondevaporation of the condensation water (Fig. 1), and transferred into a 1 mL glass vial. We assume that the condensation water is in constant equilibrium with evaporating water in soil and thus the water isotopes of evaporating water in soil can be obtained from that of condensation water on the plastic film. 110 After the collection, the plastic film was removed with little disturbance to the site. Subsequently, three new areas were selected randomly and covered in a similar manner with a new piece of plastic film for https://doi.org/10.5194/hess-2020-648 Preprint. Discussion started: 26 January 2021 c Author(s) 2021. CC BY 4.0 License. the next water collection. In addition, BW was obtained from 0-5 cm surface soil water (Wen et al., 2016). The 0-5 cm soil samples were collected using a soil auger every three days with three replicates, and each was well mixed and separated into two subsamples: one for determining the soil gravimetric water content and the other for water stable isotope analysis. The subsample for soil gravimetric water content was stored in aluminum 120 box and oven dried for 24 h at 105 o C. The other one was stored in 150 mL high density polyethylene bottles, sealed with parafilm®, transported to a freezer at -20 o C in the laboratory until cryogenic liquid water extraction took place. A cryogenic vacuum distillation system (LICA, Li-2000, China) with a pressure about 0.2 Pa and heating temperature at 95 o C was used to extract soil water (Wang et al., 2020).
The extraction time was at least 2 h until all water evaporated from the soil and deposited to the cryogenic 125 tube. In order to calculate the extraction efficiency, samples were weighed before and after extraction, and weighed again after oven-dried 24 h following extraction. Samples with an extraction efficiency less than 98 % were discarded. In terms of weight, the cryogenic vacuum distillation extracts all the water from soil. However, spiking experiments show that the extracted water is depleted in heavy isotope than the spiking water and the depletion is positive with increasing soil clay contents but negative with 130 https://doi.org/10.5194/hess-2020-648 Preprint. Discussion started: 26 January 2021 c Author(s) 2021. CC BY 4.0 License. increasing water contents. Moreover, higher temperature (>200 o C) is suggested to be used for soil water extractions (Gaj et al., 2017a;Gaj et al., 2017b;Orlowski et al., 2018;Orlowski et al., 2016;Orlowski et al., 2013). Therefore, the water isotopic compositions obtained from our distillation system were subsequently corrected by a calibration equation. The equation contains clay and soil water content as factors and was obtained through a spiking experiment with 205 o C oven-dried soils (the related data was 135 submitted to Hydrological Processes, under review).
where  2 H and  18 O are the soil water isotopic compositions; a and b are the slope and intercept of local meteoric water line, respectively.
Precipitation was collected during the whole growth season by three rainfall collectors (Wang et al., 2010) 145 in the experimental field. The rainfall amount was obtained by weighing using an electrical balance.
Subsequently, subsample of these rainfall samples were transferred to 15 mL glass vials and sealed immediately with parafilm® and placed in a refrigerator at 4 o C.
The air and 0-5 cm soil temperature under the newly covered plastic film during 2016/9/10 to 2016/9/28 were measured by E-type thermocouple (OMEGA, USA) with a CR1000 datalogger and 0-5 cm soil 150 temperature in field condition during the whole field season was measured by ibutton (Maxim Integrated, DS1921G, USA) with the frequency of one hour. We estimated 0-5 cm soil temperature under the newly covered plastic film before 2016/9/10 from the temperature of 0-5 cm soil without the plastic film covering through regression. The regression was established using 0-5 cm soil temperature under the newly covered plastic film and soil temperature without plastic film covering between 2016/9/10 to 155 2016/9/28 using ibutton. Similarly, air temperature under the newly covered plastic film before 2016/9/10 was calculated from the temperature of 0-5 cm soil under the newly covered plastic film by regression between air temperature and 0-5 cm soil temperature under the newly covered plastic film. The regression equations were presented in the Supplement File. Moreover, the hourly ambient air relative humidity was https://doi.org/10.5194/hess-2020-648 Preprint. Discussion started: 26 January 2021 c Author(s) 2021. CC BY 4.0 License. recorded by an automatic weather station (HOBO event logger, USA) located nearby at a distance of 3 160 km.
A micro-lysimeter (Ding et al., 2013;Kool et al., 2014) with three replicates, made of high-density polyethylene with 10 cm in depth, 5.2 cm in inner radius, and 3 mm in thick, was used to obtain soil evaporation amount. The micro-lysimeter was pushed into the soil surface between maize rows to retrieve an undisturbed soil sample. Subsequently, we sealed the bottom, weighed the micro-lysimeter, placed it 165 back in the soil with the same level with soil surface, and no other sensor was installed. After two days' evaporation, we weighed it again. The mass difference was the soil evaporation amount. Further, the soil of the inside lysimeter was changed every four days. In addition, after every rainfall or irrigation, the soil inside the micro-lysimeter was changed immediately.
At the end of growing season, stainless rings with the volume of 100 cm 3 were pushed into the soil to 170 obtain the soil samples. Subsequently, the soil samples were oven-dried and weighed. The bulk soil density was obtained by dividing the dry soil mass by volume.
All the water samples were analyzed for  2 H and  18 O using isotopic ratio infrared spectroscopy (Los Gatos Research, IWA (Model)-45EP, USA) at Northwest A&F University, China. The precision of this machine is 1.0‰ and 0.2‰ for  2 H and  18 O, respectively. 175 The results are reported in -notation relative to V-SMOW as detailed in Equation (2).
where Rsample denotes the ratio of the number of heavy isotopes to that of the light one in sample water; Rstandard denotes the ratio in the Vienna Standard Mean Ocean Water (V-SMOW).

Equilibrium fractionation processes 180
Isotopic composition of EW was calculated from that of the condensation water that was adhered on the underside of the newly covered plastic film. We assumed that the water vapor under newly covered plastic film, and above the surface soil constitutes a closed system. Within the system, two equilibrium fractionation processes are temperature-dependent and occur independently: Evaporation from surface soil water to air under the plastic film occurs during the day time (8 a.m. to 8 p.m., Fig. 2 1000×ln where + and * are the equilibrium fractionation factor during condensation and evaporation, 195 respectively; δ liquid is the isotopic composition in the liquid water, δ vapor is the isotopic composition in the vapor, T is temperature presented in Kelvin.

200
Based on Eqs. (3-6) and Fig. 1c, the fractionation factors for two processes under the newly covered plastic film are expressed using equations 7 and 8.
where Vp represents isotope values of vapor water under the newly covered plastic film, EW represents 205 the isotope value in evaporating water, and CW represents the isotope value in condensation water.
Combining equations (7) and (8), we obtained the isotopic composition in the EW:
(2 ) = 25.115(1 − ℎ) , (17) 220 where f represents the ratio of evaporative water loss to the total water source; 1 * is the equilibrium fractionation factor in the soil; + is the equilibrium fractionation factor in the ambient air; ℎ is the average ambient air relative humidity over 30 days prior to each soil water sampling (Sprenger et al., 2017); is the amount weighted isotopic composition in precipitation from 2016/7/11 to 2016/9/16; 225 is the isotopic signal of 0-5 cm bulk soil water; is defined as the isotopic signal of the original water source by calculating the intercept between the evaporation line of the 0-5 cm bulk soil water isotope data in Period Ⅰ in the dual-isotope plot and the LMWL (Fig. 3).  In Period Ⅱ, the initial value was calculated from the amount weighted average of the isotope values of 235 irrigation water and Period Ⅰ original water described above. In order to calculate evaporative water loss from EW  18 O, we used BW to express EW and obtained the following formulas (Eqs. 19-20) for evaporation fraction.   Figure 4 shows that the soil water content in 0-5 cm was close to saturation right after the first large precipitation event (2016/7/24) and then decreased with evaporation time (grey bars in Fig. 4c). Similarly, after the irrigation event (2016/8/26), 0-5 cm soil water content jumped to a high value and then decreased 265 with the increase of evaporation time (Fig. 4c). In total, there were 12.73±0.58 mm and 7.51±1.24 mm reduction of soil water storage in 0-5 cm during Period Ⅰ and Period Ⅱ, respectively. However, from the micro-lysimeters, we obtained the total evaporation amount of 20.45±0.95 mm in Period Ⅰ, and 9.56±1.18 mm in Period Ⅱ. Therefore, the evaporation amount in each of the two periods was larger than the soil water storage reduction in 0-5 cm, suggesting soil water from below 5 cm moved up and was participated 270 in evaporation in each of the two periods, especially in Period I.

 2 H and  18 O in evaporating water and bulk soil water
The precipitation on 2016/7/24 had a  18 O value of -8.11±0.05‰ and  2 H value of -62.97±0.20‰, which were smaller than the respective values of pre-event BW (-1.22±0.91‰ for  18 O and -37.79±2.81‰ for  2 H) (Fig. 4). The irrigation water -with a  18 O of -9.40±0.05‰ and  2 H of 51.12±2.7‰ on 2016/8/26 -275 had a lower  18 O, but a much higher  2 H than the pre-irrigation BW (-0.22± 0 .59‰ for  18 O and -39.21± 2 .81‰ for  2 H). Therefore, in Period I, the newly added water was more depleted in heavy isotopes relatively to pre-event BW (p<0.05). In Period II, the newly added water had a lower  18 O, but a higher  2 H than pre-event BW (p<0.05).  In Period II, BW  18 O also increased as evaporation occurred (p<0.05). The increase of BW  18 O also 290 had a significant linear relationship with evaporation time (p<0.05; Fig. 5). On the contrary,  2 H of BW, surprisingly decreased linearly with evaporation (p<0.01). The slope and intercept were both significantly different from zero (p<0.01). This suggests that in Period II evaporation favors the lighter isotope for O, but heavier isotope for H.
The change of water isotopes in EW is very similar to that in BW. For example, in Period II, water 295 isotopes in EW showed a similar trend as in BW:  18 O increased with evaporation time (Fig. 5d) and the slope and intercept were significantly different from zero (p<0.05). And  18 O was consistently more depleted in EW than in BW in the period with same slope but significantly smaller intercept (p<0.01).
Also similar to that in BW,  2 H in EW decreased with evaporation time but did not differ from that in https://doi.org/10.5194/hess-2020-648 Preprint. Discussion started: 26 January 2021 c Author(s) 2021. CC BY 4.0 License. BW (p>0.05,Figs. 4,5), therefore the two lines had the similar slope and intercept (Fig. 5b). Therefore, 300 the linear relationship in  18 O between EW and BW was given as: (Fig. 5) While the slopes represent the evaporative demand of the atmosphere, regardless of the source of water, the intercept represents the initial condition of the source of water for evaporation. Therefore, the initial water source in Period II had a  18 O value of -1.67‰ for BW, but of -3.80‰ for EW. Therefore, the 305 sources of water for BW and EW had different isotopic compositions in Period II.

Variation of deep soil water content,  2 H,  18 O, and lc-excess
The precipitation on 2016/7/24 increased the soil water content in the top 60 cm, but decreased soil water  2 H and  18 O in the top 20 cm (Fig. 6, upper panel). Therefore, the top 20 cm lc-excess on 2016/8/3 increased. However, the precipitation had no influence on deeper soil  2 H,  18 O, and lc-excess. At the 310 end of evaporation Period Ⅰ, soil water content decreased in the top 60 cm. And in the top 10 cm, soil water  2 H and  18 O increased and lc-excess decreased.

Why evaporating and bulk soil water isotopic compositions differ 340
During evaporation, light isotopes are favored to the vapor making the residual liquid water enriched in heavy isotopes (Mook and De Vries, 2000). This can explain why, with increasing evaporation time, both  2 H and  18 O in BW experienced increasing trend in Period Ⅰ. In Period Ⅱ,  18 O (Fig. 5) displayed a similar, increasing trend, but  2 H had an opposite, decreasing trend. This is inconsistent with the trend, of  18 O in the same period and, of both  2 H and  18 O in Period I (Fig. 5). The progressive decrease in 345  2 H with increasing evaporation time cannot be explained by the general notion that with evaporation, residual soil water becomes more enriched with heavy water isotopes. Therefore, there must be a mechanism that either preferentially removes 2 H or dilutes 2 H by 2 H-depleted water.
For the latter, because there is negligible input of water from the atmosphere (both in vapor and liquid form), the only input of water could be from the soil below 5 cm. Indeed, because the evaporation amount, 350 derived by lysimeters, was larger than 0-5 cm soil water storage reduction (Sect. 3.1), the water below 5 cm must have moved upward as evaporation occurred. Consequently, due to evaporation, the order of  2 H value should be 0-5 cm > the mixture of pre-evaporation 0-5 cm and 5-10 cm soil water > 5-10 cm.
However, 0-5 cm  2 H at the end of evaporation period, i.e. on 2016/9/16, is similar to 5-10 cm  2 H (Fig.   6f). Moreover, if dilution occurred, the  18 O would also be diluted, which is not supported by the 355 progressive increase of BW  18 O during evaporation in the same period and of both  2 H and  18 O in BW of Period I, which should have more deep soil water contribution (Sect. 3.1). Therefore, dilution should not have substantially affected the isotopic signature of BW. This is further supported by the larger  18 O in BW in Period Ⅱ than that in EW (Figs. 4,5). By deduction, the possible cause of the depletion in 2 H would be a preferential removal of 2 H from the top 5 cm of soil. 360 We did not detect significant  2 H differences in EW from that in BW in Period Ⅱ (Fig. 5) Different isotopic signatures of BW and EW indicates that the sources of water for BW and EW were different. The source of EW is closer to new water than that of BW. This could be explained by a conceptual model of new water and old water partition in soil (Fig. 8). 365 As pointed out abundantly in the recent literature, there could be isotopic separation in water isotopes https://doi.org/10.5194/hess-2020-648 Preprint.  Wang, P., Song, X., Han, D., Zhang, Y., and Liu, X.: A study of root water uptake of crops indicated by hydrogen and oxygen stable isotopes: A case in Shanxi Province, China, Agric Water Manag, 97, 475-482, doi:10.1016/j.agwat.2009.11.008, 2010 Weiler, M. and Naef, F.: An experimental tracer study of the role of macropores in infiltration in grassland soils, Hydrol Process, 17, 477-493, doi:10.1002Process, 17, 477-493, doi:10. /hyp.1136Process, 17, 477-493, doi:10. , 2003 Wen, X., Yang, B., Sun, X., and Lee, X.: Evapotranspiration partitioning through in-situ oxygen isotope  , 41, 3148-3156, doi:10.13227/j.hjkx.201911063, 2020.