Climatic and environmental controls on stable isotopes in atmospheric water vapor near the surface observed in Changsha, China

Accepted Manuscript Climatic and environmental controls on stable isotopes in atmospheric water vapor near the surface observed in Changsha, China Tianci Yao, Xinping Zhang, Huade Guan, Hui Zhou, Mingquan Hua, Xuejie Wang PII:

S1352-2310(18)30455-2

DOI:

10.1016/j.atmosenv.2018.07.008

Reference:

AEA 16114

To appear in:

Atmospheric Environment

Received Date: 13 March 2018 Revised Date:

29 June 2018

Accepted Date: 3 July 2018

Please cite this article as: Yao, T., Zhang, X., Guan, H., Zhou, H., Hua, M., Wang, X., Climatic and environmental controls on stable isotopes in atmospheric water vapor near the surface observed in Changsha, China, Atmospheric Environment (2018), doi: 10.1016/j.atmosenv.2018.07.008. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Climatic and environmental controls on stable isotopes in

2

atmospheric water vapor near the surface observed in Changsha,

3

China

4

Tianci Yaoa,b, Xinping Zhanga,b,*, Huade Guanc, Hui Zhoua,b, Mingquan Huaa,b, Xuejie Wanga,b

RI PT

1

5 6

a

7

China

8

b

9

China c

SC

Key Laboratory of Geospatial Big Data Mining and Application, Hunan Province, Changsha, 410081,

M AN U

10

College of Resources and Environmental Sciences, Hunan Normal University, Changsha, 410081,

School of the Environment, Flinders University, Adelaide, 5001, Australia

11 * Corresponding author.

13

E-mail address: [email protected] (X. Zhang).

14

Tel: +86 73188871420.

AC C

EP

TE D

12

1

ACCEPTED MANUSCRIPT

15

Abstract Continuous measurements were made of isotopic compositions in atmospheric water vapor (δv)

17

and in atmospheric precipitation (δp) near the surface in the subtropical monsoon region (Changsha,

18

China) from October 2014 to March 2017 to investigate the change characteristics of vapor isotopes

19

and their interactions with precipitation isotopes. On a diurnal time scale, the δ18Ov values were more

20

negative in the daytime and more positive at night, and the change in dv was reversed during non-rainy

21

days; however, the δ18Ov values were more positive in the daytime and more negative at night, and the

22

variation in dv was not significant during rainy days. On a seasonal time scale, affected by distinctive

23

moisture sources, the vapor isotopes had obvious seasonal variations, with higher δ18Ov and dv values

24

occurring in cold seasons than in warm seasons. There were generally consistent and significant

25

negative linear relationships of dv with temperature and absolute humidity, but the linear relationships

26

between δ18Ov and meteorological factors were variable on the different time scales, which is closely

27

associated with the local weather conditions on the short time scales from diurnal to intra-seasonal and

28

movements of air masses with distinct thermodynamic properties on seasonal and sometimes

29

intra-seasonal time scales. Although the change trend of daily δ18Ov closely approximated that of δ18Op,

30

δ18Op experienced significant relative enrichment compared to δ18Ov, with an average enrichment of

31

9.18 ‰. Overall, the isotopic compositions of atmospheric water vapor were in equilibrium with those

32

of precipitation while rainfall events occurred, except for those in the cold season in 2014, and when

33

tracing short-duration weather events, δ18Ov provided more details and information related to weather

34

processes than δ18Op.

35

Keywords: Changsha; atmospheric water vapor; precipitation; stable isotopes; isotope effect

AC C

EP

TE D

M AN U

SC

RI PT

16

2

ACCEPTED MANUSCRIPT 36 37

1. Introduction Stable isotopes in precipitation (e.g., 2H and

18

O) are useful environmental tracers that are used to

investigate climate change and hydrological processes. Since the investigation of the stable isotopic

39

compositions of precipitation in the early 1950s, the change characteristics of stable precipitation isotopes

40

on different temporal and spatial scales and their applications have been widely reported, and a great deal

41

of meaningful information has been obtained (Araguás, Froehlich, & Rozanski, 1998; Dansgaard, 1964;

42

Wu, Zhang, Li, Li, & Huang, 2015; Yao et al., 2013; Yoshimura, 2015), such as the mechanisms

43

controlling natural stable isotopic fractionation (Dansgaard, 1953, 1964; Jouzel, 1986; Stewart, 1975;

44

Zhang, Guan, Zhang, Zhang, & Yao, 2016), the relationships between stable isotopes and meteorological

45

variables (known as the “isotope effect”; Bowen & Wilkinson, 2002; Dansgaard, 1964), and the

46

interpretation of the paleoclimate from stable isotopes records in tree cellulose, ice cores, and carbonates

47

(Cheng et al., 2016; Johnsen et al., 2001; Sidorova et al., 2013; Thompson et al., 2000; Yao et al., 1996).

48

Precipitation, however, only represents the terminal unit of the processes of evaporation, mixing, and

49

subsequent condensation, and it exhibits large spatiotemporal variability. Accordingly, there are certain

50

roles-constrained for stable precipitation isotopes in tracing the water cycle process in a specific region

51

and examining the patterns of atmospheric circulation (Frankenberg et al., 2009; Zhang, Zhang, Guan,

52

Huang, & Wu, 2012). Water vapor is one of the most active atmospheric components. Compared with

53

precipitation, water vapor not only has a continuous spatial-temporal distribution but also responds more

54

sensitively to changes in atmospheric conditions. As a consequence, studies of stable vapor isotopes can

55

complement the deficiency of understanding caused by discrete and brief precipitation observations.

AC C

EP

TE D

M AN U

SC

RI PT

38

56

The systematic sampling survey of stable isotopes in atmospheric water vapor dates back to the late

57

1960s, when Talor (1968) reported measurements of deuterium, the water vapor mixing ratio, and 3

ACCEPTED MANUSCRIPT temperature in the atmosphere at altitudes between 0.5 and 5000 m from June 1967 to June 1968 and

59

delineated 20 average vertical profiles of δ2Hv in Heidelberg. After that, in order to evaluate the influences

60

of thermonuclear explosions on tritium content in the atmosphere and investigate the characteristics of

61

variations of stable vapor isotopes, the monitoring programs of vapor isotopes were carried out

62

successively by several countries. For example, the National Center for Atmospheric Research in the

63

United States investigated the vertical distributions of deuterium, tritium, temperature, and humidity from

64

the surface to an altitude of 9.2 km in Scottsbluff, Santa Barbara, Palestine, and Death Valley, United

65

States, from November 1965 to September 1973, using aircraft and radiosondes (Ehhalt, 1974). In

66

addition, a comprehensive sampling program of 2H,

67

surface in Heidelberg, Paris, Hannover, Krakow, and Miami was sponsored by Germany in June 1980

68

(Schoch-Fischer et al., 1984). All of these abovementioned observations have provided rather valuable

69

data for the studies of vapor isotopes. However, due to instrumental and technical limitations, there are

70

just a few stations for observations of vapor isotopes around the world.

M AN U

SC

RI PT

58

O, and 3H in atmospheric water vapor near the

TE D

18

In recent years, fortunately, with advances in spectral technology, various vapor isotope

72

spectrometers have emerged. Using these instruments, we can monitor the stable isotopic compositions in

73

the atmospheric water vapor of different layers over the ground with automatic, high-frequency, and

74

continuous sampling. The previous studies show that the change characteristics and influencing factors of

75

vapor isotopes near the surface may vary geographically. Lee, Smith, & Williams (2006) studied vapor

76

isotopes in New Haven using a 1-year record observed by a tunable diode laser (TDL) trace gas analyser

77

and found that the changes of vapor isotopes were in good agreement with the variations in the water

78

vapor mixing ratio and temperature, with higher values occurring in warm seasons (May to October) and

79

lower values occurring in cold seasons (November to April). On a short time scale, such as that of a

AC C

EP

71

4

ACCEPTED MANUSCRIPT weather process, affected by the distinct moisture origins and degree of condensation of an advection air

81

mass, the scatter of δv versus the water vapor mixing ratio showed an interesting looping. Tremoy et al.

82

(2012) used the high-temporal-resolution measurements of stable vapor isotopes near the ground to

83

investigate seasonal and diurnal vapor isotopic variations in Niamey. They found that on a seasonal time

84

scale, the change of δ18Ov showed a W-shape, affected by the regional convective activity during the

85

monsoon season and large-scale subsidence north of Niamey during the dry season. However, the

86

pronounced diurnal variation of δ18Ov recorded the propagation of density currents related to mesoscale

87

convective systems during the monsoon season and a mixing process between the boundary layer and the

88

free atmosphere during the dry season. Laskar et al. (2014) showed, on a seasonal time scale, the

89

variations of δ18O in atmospheric water vapor and precipitation were not consistent, but they found that

90

the vapor and precipitation were nearly in isotopic equilibrium during a typhoon in the case study.

M AN U

SC

RI PT

80

In mainland China, most stable vapor isotope observation campaigns are in the Qinghai-Tibet area

92

(Cui, Tian, Liu, & Wen, 2014; Hu et al., 2009; Yin et al., 2008; Yu, Tian, Ma, Xu, & Qu, 2015; Yu, Yao,

93

Tian, Wang, & Yin, 2005) and North China (Wen et al., 2008; Wen, Zhang, Sun, Yu, & Lee, 2010; Yuan,

94

Zhang, Sun, Wen, & Zhang, 2010; Zhang, Sun, Wang, Yu, & Wen, 2011; Zhang, Cai, et al., 2011). For

95

example, in the Qinghai-Tibet area, Yin et al. (2008) sampled atmospheric water vapor near the surface in

96

Delingha with a cold trap. They noted that distinct moisture origins had significant impacts on the vapor

97

isotopes. Overall, higher δ18Ov values were associated with water vapor transported by the westerlies, and

98

lower δ18Ov values were associated with that transported by the southwest monsoon. Wen et al. (2010)

99

measured the in situ δ2H and δ18O of atmospheric water vapor in surface air with a TDL from December

100

2006 to December 2007, and in their observations, much fewer daily and diurnal variations of vapor

101

isotopes were recorded during the summer monsoon season than during other times of the year;

AC C

EP

TE D

91

5

ACCEPTED MANUSCRIPT additionally, during the non-monsoon season, the δ2Hv and δ18Ov values showed a positive log-linear

103

dependence on the water vapor mixing ratio, and the deuterium excess (dv), as calculated by the stable

104

oxygen and hydrogen isotopic compositions of atmospheric water vapor (dv = δ2Hv – 8δ18Ov), revealed a

105

remarkable negative correlation with relative humidity.

RI PT

102

Comparatively speaking, because it is difficult to capture atmospheric water vapor, the related

107

studies to date are deficient, and the observations of vapor isotopes are limited in specific regions,

108

especially in the subtropical and tropical monsoon regions. Consequently, the change patterns of vapor

109

isotopes in these regions are uncertain. To fill this gap, we carried out a field experiment to collect

110

real-time isotope data in atmospheric water vapor and in-event precipitation near the surface from

111

October 2014 to March 2017 in Changsha, a subtropical monsoon region. In the present work, our

112

objectives are to analyse the influences of moisture sources and local meteorological variables on

113

variations in vapor isotopes on diurnal, a few days, seasonal, and intra-seasonal time scales and to detect

114

the interactions between falling raindrops and surrounding water vapor during rainfall events. This will

115

enhance our understanding of the characteristics of stable isotopic compositions in atmospheric water

116

vapor and precipitation near the surface in the subtropical monsoon region and provide insights into

117

distinct moisture sources and local atmospheric and hydrological processes. Furthermore, the results will

118

also have important implications for validation of the isotopic observations of the satellite, as well as the

119

analyses of paleoclimate reconstructions and atmospheric circulation patterns.

120

2. Methodology

121

2.1. Study area

AC C

EP

TE D

M AN U

SC

106

122

Changsha (27.68° - 28.85° N and 111.88° - 114.25° E; Fig. 1) is located in the northeastern Hunan

123

Province of China, i.e., in the lower reach of the Xiangjiang River and western Changsha-Liuyang Basin. 6

ACCEPTED MANUSCRIPT The elevation of Changsha varies greatly, with the Mufu-Luoxiao Mountains to the northeast and

125

Xuefeng Mountains to the northwest. And Changsha is situated in the transition zone from the

126

Changsha-Hengyang Hills to Lake Dongting Plain, with the elevation generally decreasing from the south

127

to the north of the study area.

M AN U

SC

RI PT

124

128

Fig. 1. The map showing the location of sampling point in Changsha, China.

TE D

129

This sampling program was performed at the meteorological observation station (28.19° N, 112.94°

131

E) on the Hunan Normal University campus, which is situated in Changsha City. The study area is in a

132

typical subtropical monsoon moist climate, with a cold and dry winter influenced primarily by air masses

133

from the north, and a warm and wet summer influenced mainly by air masses from the south. Overall, the

134

mean annual temperature and mean annual total precipitation in Changsha are about 17.4 °C and 1428.1

135

mm, respectively, based on the meteorological data between 1981 and 2010 provided by the China

136

Meteorological Administration. The mean annual temperature and mean annual total precipitation are

137

approximately 6.5 °C and 215.7 mm, respectively, in winter (December to February), and 27.8 °C and

138

473.3 mm, respectively, in summer (June to August). Soil types are characterized by red soil and

139

yellow-brown soil, with loose soil texture and sufficient soil water content. Vegetation is dominated by

AC C

EP

130

7

ACCEPTED MANUSCRIPT subtropical evergreen tree species, such as broad-leaf tree species Cinnamomum camphora, Osmanthus

141

fragrans, and Ulmus pumila L. The shrubs are Rhododendron simsii Planch and Loropetalum chinense

142

var. rubrum.

143

2.2. Measurements and data

RI PT

140

The stable isotopic compositions of atmospheric water vapor, at approximately 10 m above the

145

ground, were measured every ~10 s from October 2014 to March 2017 using an infrared laser

146

spectrometer (Model 35EP, Los Gatos Research, Inc., USA) with off-axis integrated cavity output

147

spectroscopy. The spectrometer consists of two subsystems, namely, a standard water vapor generator and

148

a water isotope analyser. The former is used to generate the calibration standard gases at different

149

concentrations with known stable isotope values, and the latter is used to determine the concentrations of

150

calibration standard gases and ambient air samples, as well as their stable isotopic compositions. The two

151

subsystems operate in an alternating fashion in a measurement cycle, which is 1 h in the standard water

152

vapor generator and 1.5 h in the water isotope analyser. Due to the significant concentration dependence

153

on stable isotopic compositions, we are required to make a correction for the initial ambient vapor isotope

154

measurements based on the relationships between the isotope measurements of calibration standard gases

155

and their concentrations. Additionally, memory effects often occur due to residual water vapor that cannot

156

be completely pumped out from the pipeline before the next measurement. Therefore, we excluded the

157

data observed during the first one minute of each measurement.

M AN U

TE D

EP

AC C

158

SC

144

Along with regular water vapor sampling, precipitation (greater than or equal to 0.1 mm) samples

159

were collected at 08:00 A.M. and 20:00 P.M. (China Standard Time, hereafter CST) on rainy days. A total

160

of 435 precipitation samples were obtained during the whole period. The stable isotopic compositions of

161

all precipitation samples were measured using the LGR 35EP. 8

ACCEPTED MANUSCRIPT The isotopic compositions of atmospheric water vapor (or precipitation) were expressed using the

162 163

conventional delta notation: or

164

⁄

=

− 1 × 1000 ‰,

(1)

where Rsample and RV-SMOW are the isotope molar ratios of the heavy to the light isotope (e.g., 2H/1H,

166

18

167

measurement accuracies of δ2Hv and δ18Ov are better than 1.2 ‰ and 0.4 ‰, respectively. The

168

measurement accuracies for δ2Hp and δ18Op are better than 0.6 ‰ and 0.2 ‰, respectively.

RI PT

165

SC

O/16O) in samples and in Vienna Standard Mean Ocean Water (V-SMOW), respectively. The

Some meteorological elements, including surface air temperature (T), precipitation (P), and humidity

170

(Q), were measured using automatic weather survey equipment (Model Weather Hawk-232, USA) at a

171

30-min interval. Additionally, the atmospheric precipitable water (PW) was obtained from the

172

ERA-Interim dataset provided by the European Centre for Medium-Range Weather Forecasts (ECMWF),

173

with a spatial resolution of 1.5º ×1.5º.

TE D

M AN U

169

It should be noted that the calculations of the mean isotopic compositions in atmospheric water

175

vapor and in precipitation were different in a given period. The arithmetic mean value was calculated for

176

the vapor isotopes and the amount-weighted mean value was calculated for the precipitation isotopes. To

177

analyse the relationships between the stable isotopic compositions in atmospheric water vapor and in

178

precipitation, the stable isotopic ratio δe (defined as the predicted value and assumed to be in an

179

equilibrium state with that of precipitation near the surface) was calculated using the classic Rayleigh

180

distillation equation (Saxena, & Eriksson, 1985):

181

AC C

EP

174

δe = (1+δp)/α – 1,

(2)

182

where α is a temperature-dependent equilibrium fractionation factor (Majoube, 1971) and δp is the

183

isotopic ratio in precipitation. Based on the criterion of surface meteorological observations, all daily data 9

ACCEPTED MANUSCRIPT 184

in this study were calculated for the daily period of 8:00 A.M. to 8:00 A.M. on the next day.

185

3. Results and discussion

186

3.1. Overall variations of vapor isotopic compositions Fig. 2 shows the daily variations in δ18Ov, dv, temperature, absolute humidity (Q), and precipitation

188

observed during the whole study period. In addition, the frequency distributions of δ18Ov and dv are

189

shown in Fig. 3.

SC

-10 -15 -20

M AN U

18

δ Ov (‰)

-5

-25 -30

(b)

80

P

40 20 0 9

190

1

3

5 7 2015

9

1

3

5 7 2016

9

11

Q

60 50 40 30 20 10 0 -10 -20 35 30 25 20 15 10 5 0 -5

1 3 2017

Fig. 2. Variations of daily mean δ18Ov, dv, temperature, and absolute humidity, as well as daily total precipitation near

192

the surface during the whole sampling period in Changsha.

(a)

(b) 0.35

0.30

0.30

0.25

0.25 Frequency

Frequency

AC C

0.35

0.20 0.15

0.20 0.15

0.10

0.10

0.05

0.05

0.00 -30

193

11

EP

191

11 2014

TE D

P (mm)

60

T

dv

dv (‰)

δ18Ov

(a)

Q (g kg-1) | T (°C)

0

RI PT

187

-25

-20

-15

-10

-5

0

δ18Ov (‰)

0.00 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 45 50 55 dv (‰)

194

Fig. 3. The frequency distributions of daily mean δ18Ov (a) and dv (b) near the surface during the whole sampling

195

period in Changsha. The sum of the frequency of the slashed bar is greater than 0.8. 10

ACCEPTED MANUSCRIPT As shown in Figs. 2 and 3, the δ18Ov values varied from -28.15 ‰ to -1.18 ‰, with a mean value of

197

-13.94 ‰, and 80 % of the values were distributed in the range from -20 ‰ to -10 ‰ in the whole survey

198

period in Changsha. Comparatively, the mean annual δ18Ov value was more positive than the yearly

199

average of -20.7 ‰ reported in Beijing at 40°00′ N and 116°23′ E (Wen et al., 2010) and the 1-year

200

average of -17.1 ‰ observed in Tsukuba at 36°03′ N and 140°02′ E (Wei et al., 2016), but it is very

201

similar to the yearly average of -14.4 ‰ recorded in Taipei at 25°1′ N and 121°30′ E (Laskar et al., 2014).

202

Considering the similar climatic characteristics of four stations, the difference between the four, in a sense,

203

reflects the impact of the latitudinal effect in precipitation isotopes, which refers to that precipitation δ18O

204

in low latitudes is greater than in high latitudes. The dv values varied from -17.47 ‰ to 52.49 ‰, with a

205

mean value of 18.18 ‰, and 80 % of the values were distributed in the range from 10 ‰ to 25 ‰. By

206

comparison, the mean dv value was also more positive than that observed in Beijing, where the mean

207

annual value was 11.8 ‰, but very close to those observed in Tsukuba and Taipei, with mean values of

208

19.28 ‰ and 19.69 ‰, respectively.

TE D

M AN U

SC

RI PT

196

We defined the warm season as the months from April to September and the cold season as the

210

months from October to March based on the seasonal distribution patterns of precipitation and

211

temperature in the study area (Fig. 2). In general, accompanying the remarkable seasonal variations, in the

212

warm season, the δ18Ov values varied from -22.45 ‰ to -2.72 ‰, with a mean value of -14.14 ‰, and the

213

dv values ranged from -5.02 ‰ to 30.56 ‰, with a mean value of 15.30 ‰; in the cold season, the δ18Ov

214

values varied from -28.15 ‰ to -1.18 ‰, with a mean value of -13.70 ‰, and the dv values varied from

215

-17.47 ‰ to 47.91 ‰, with a mean value of 19.72 ‰. The seasonality patterns of vapor isotopes are

216

connected with the difference of moisture sources between the warm season and the cold season, as well

217

as with the air masses controlling this area (Srivastava, Ramesh, Gandhi, Jani, & Singh, 2015; Strong,

AC C

EP

209

11

ACCEPTED MANUSCRIPT Sharp, & Gutzler, 2007; Yu et al., 2005). Generally, the δ18Op and dp in Changsha were more negative in

219

the warm season, controlled by warm and wet marine air mass transported by the southwest and southeast

220

monsoon, and thus the δ18Ov and dv were also lower, impacted by the recycling of precipitation and under

221

the evaporation fractionation of surface water with relatively negative isotope values; however, the δ18Op

222

and dp in Changsha were more positive in the cold season, controlled by the cold and dry continental air

223

mass transported by westerly circulation, and thus the δ18Ov and dv were also more positive, impacted by

224

the recycling of precipitation and under the evaporation fractionation of surface water with relatively

225

positive isotope values (Zhang, Yang, Niu, & Wang, 2009). For the most positive δ18Ov in the period

226

from December 2014 to April 2015, the large positive deviation is probably related to the abnormally

227

continuous drought over East Asia since the summer of 2014. Using the study area as an example, the

228

total precipitation in autumn and winter of 2014 was much less than the normal precipitation for the

229

climatological period 1961 - 2014; it was only 181.6 mm. Due to limited time series of vapor isotope

230

observations, further study is needed for this inference.

231

3.2. The relationships between the vapor isotopes and meteorological factors on different time scales

232

3.2.1. Diurnal variations of vapor isotopic compositions

EP

TE D

M AN U

SC

RI PT

218

Generally, the movements of air masses with different thermodynamic properties have minimum

234

effects on stable isotopic compositions in atmospheric water vapor on a diurnal time scale. The variations

235

in vapor isotopes are predominantly controlled by short-term weather processes. The full data set was

236

divided into non-rainy and rainy days, with the latter day total precipitation greater than or equal to 0.1

237

mm. By averaging hourly mean δ18Ov and dv, as well as the related meteorological factors for period of

238

non-rainy days and rainy days, the ensemble mean diurnal variations of these variables were obtained

239

(Fig. 4).

AC C

233

12

ACCEPTED MANUSCRIPT (a)

-13.6

(a)

δ18Ov

(b)

dv

T

Q

21

-13.8 20

-14.0

19

-14.4 -14.6

dv (‰)

δ18Ov (‰)

-14.2

18

-14.8 17

-15.2 22

10.8

RI PT

-15.0

21

10.6

20

10.0

17

9.8

Hours (CST)

07:00

05:00

03:00

01:00

23:00

19:00

17:00

15:00

13:00

9.6 9.4

M AN U

240

11:00

09:00

07:00

05:00

03:00

01:00

23:00

21:00

19:00

17:00

15:00

13:00

11:00

14

09:00

15

SC

16

21:00

T (° C)

10.2

18

Q (g kg-1)

10.4

19

241

Fig. 4. Ensemble mean diurnal variations of δ18Ov, dv, temperature and absolute humidity near the surface on

242

non-rainy days (a) and rainy days (b) in Changsha.

During non-rainy days, the mean value of δ18Ov was -13.94 ‰, the mean value of dv was 18.73 ‰,

244

the mean temperature was 18.07 °C, and the mean absolute humidity was 9.69 g kg-1. All of these values

245

were higher than the mean values for the whole sampling period. In response to the diurnal variations of

246

temperature and atmospheric humidity, both δ18Ov and dv had notable diurnal cycles, but with the opposite

247

trends: the maximum δ18Ov value was recorded when the minimum dv value was observed therein (Fig.

248

4a). This diurnal variation in the stable isotopic composition of atmospheric water vapor was likely

249

associated with near-surface atmospheric turbulence (Lai, Ehleringer, Bond, & Paw, 2006; Welp et al.,

250

2008; Zhang et al., 2011). The intensity of turbulence is closely related to the temperature (Monin &

251

Obukhov, 1954), has a direct impact on the variations of atmospheric humidity (McBean, 1971), and

252

thereby affects the stable isotopic compositions of the atmospheric water vapor. Atmospheric turbulence

253

strengthened gradually in the morning due to the increasing near-surface temperature, and the water vapor

254

exchange between the upper and lower ground layers was enhanced in the morning. This resulted in

AC C

EP

TE D

243

13

ACCEPTED MANUSCRIPT decreased atmospheric humidity near the surface and correspondingly depleted stable isotopic signatures

256

in atmospheric water vapor; in particular, the δ18Ov value reduced by about 0.53 ‰ from sunrise to noon.

257

In contrast, atmospheric turbulence gradually weakened with decreasing temperature in the afternoon and

258

was accompanied by the decreased water vapor exchange between the upper and lower ground layers.

259

The increased near-surface atmospheric humidity resulted in enriched stable isotopic signatures. At night,

260

however, the atmospheric lamination was relatively stable and humidity was less variable, with

261

correspondingly negligible variations in the stable isotopic compositions of atmospheric water vapor

262

being recorded. Despite the fact that the diurnal amplitude of air temperature can reach up to 6.37 °C, the

263

low potential evaporation only resulted in variabilities of absolute humidity and δ18Ov of 0.35 g kg-1 and

264

0.71 ‰, respectively. In comparison, dv varied notably, exhibiting daily variability of up to 3.99 ‰. The

265

diurnal variations of dv and δ18Ov had opposite trends because both atmospheric turbulence and

266

evapotranspiration have a positive impact on dv (Welp et al., 2012). Comparatively, the diurnal patterns of

267

δ18Ov and dv are in line with the Welp et al. (2012) meta-study on six sites in mid-latitudes, all of which

268

support the concept that robust diurnal patterns of vapor isotopes may result from the combined effects of

269

local atmospheric turbulence and evapotranspiration on non-rainy days.

EP

TE D

M AN U

SC

RI PT

255

During rainy days, the mean value of δ18O in atmospheric water vapor was -14.66 ‰, the mean

271

value of dv was 17.16 ‰, the mean temperature was 16.35 °C, and the mean absolute humidity was 10.52

272

g kg-1. The mean values of all the above-mentioned variables, except for absolute humidity, were lower

273

than the mean values of those when pooling all the data together during the whole sampling period.

274

Compared with the results on non-rainy days, the stable isotopic signatures on rainy days were notably

275

depleted. The atmospheric humidity was largely controlled by the air temperature, given that the

276

underlying surface had enough water for evapotranspiration. Therefore, diurnal variations in the

AC C

270

14

ACCEPTED MANUSCRIPT temperature, atmospheric humidity, and stable isotopic signatures in atmospheric water vapor exhibited

278

similar trends (Fig. 4b). During the daytime, the atmospheric humidity remained at a high level because

279

of the high temperature and strong underlying surface evapotranspiration, and thereby producing enriched

280

stable isotopic signatures in atmospheric water vapor. At night, however, the atmospheric humidity

281

continuously declined as a result of the low temperature and decreased underlying evapotranspiration,

282

thereby resulting in the continuous depletion of stable isotopic signatures. Since the underlying surface

283

evapotranspiration has no impact on the stable isotopic fractionation, the dv within a day varied

284

minimally.

285

3.2.2. Seasonal and intraseasonal variations of daily vapor isotopic compositions

M AN U

SC

RI PT

277

The above analyses indicated that δ18Ov was proportional to the absolute humidity regardless of

287

whether it was rainy or not. This suggests that the water vapor derived from underlying surface

288

evapotranspiration will probably result in the enrichment of stable isotopic signatures during its diurnal

289

variation. The relationship between the stable isotopic signatures in atmospheric water vapor and the

290

atmospheric humidity was unchanged, although rainfall generally depleted the stable vapor isotopic

291

signatures. Correspondingly, dv had a negative correlation with atmospheric humidity on both non-rainy

292

days and rainy days. However, the relationships between δ18Ov and temperature, as well as those between

293

dv and temperature, showed differences between non-rainy days and rainy days. From Fig. 2, seasonally,

294

the variations of vapor isotopes displayed opposite trends with temperature, precipitation, and absolute

295

humidity. In the warm and rainy summer, vapor isotopic signatures were notably depleted; however, in

296

the cold and dry winter, vapor isotopic signatures were notably enriched. Our results showed that δ18Ov

297

had significant (p < 0.01, similarly hereinafter) negative relationships with temperature, precipitation, and

298

absolute humidity during the whole sampling period (Table 1). Significant negative relationships were

AC C

EP

TE D

286

15

ACCEPTED MANUSCRIPT 299

observed between dv and temperature and between dv and absolute humidity; the negative relationship

300

between dv and precipitation was not significant (p > 0.01, similarly hereinafter) and could be explained

301

by the effect of evapotranspiration during rainfall events (Zhang, Liu, Nakawo, & Xie, 2009). Table 1 The linear relationships between daily mean δ18Ov, dv and temperature, precipitation and absolute humidity

303

near the surface in Changsha.

RI PT

302

Relationships Period

δ18Ov/Q

The whole period

-0.09*(880)

-0.18*(340)

Non-rainy days

-0.09(540)

—

Rainy days

-0.11(340)

-0.18*(340)

-0.19*(332)

dv/T

dv/P

dv/Q

-0.17*(865)

-0.42*(880)

-0.13(340)

-0.46*(865)

-0.16*(533)

-0.44*(540)

—

-0.47*(533)

-0.42*(340)

-0.13(340)

-0.44*(332)

Note. The number in parentheses represents the sample size; asterisk indicates p < 0.01, the same below.

M AN U

304

δ18Ov/P

SC

2014/10-2017/03

δ18Ov/T

The negative correlations between δ18Ov and temperature, absolute humidity, as well as those of

306

between dv and temperature, absolute humidity, did not notably change between non-rainy days and rainy

307

days, as shown in Table 1. This result indicates that air masses with distinct thermodynamic properties

308

and their interactions are the most predominant factors controlling the seasonal variations of vapor

309

isotopic signatures. On a seasonal time scale, the variations of stable isotopes in atmospheric water vapor

310

depended largely on moisture sources.

EP

TE D

305

Intra-seasonal analyses are mainly used to investigate the linear relationships of vapor isotopes with

312

temperature, precipitation, and absolute humidity in spring (March to May), summer (June to August),

313

autumn (September to November) and winter (December to February), during the whole period, for

314

non-rainy days and rainy days (see Fig. 5). Under the three circumstances, the stable negative relationship

315

between δ18Ov and precipitation was observed in Changsha, although it did not reach the confidence level

316

of 0.01. Significant positive linear relationships of δ18Ov with temperature and absolute humidity were

317

recorded, which were especially pronounced in fall and winter primarily controlled by westerly winds.

318

However, due to the frequent interactions between cold and warm air masses in the middle and lower

AC C

311

16

ACCEPTED MANUSCRIPT reaches of the Yangtze River in spring and in the “plum rains” season (June to July), the relationships

320

between δ18Ov and temperature and between δ18Ov and absolute humidity varied greatly. For instance,

321

significant positive linear relationships in the spring of 2016, significant negative linear relationships in

322

the spring of 2015, and no significant relationships in the summer of 2015 were recorded between δ18Ov

323

and temperature and between δ18Ov and absolute humidity.

0.6 0.4

(a)

**

**

0.8

**

**

* *

0.6

* *** *

-0.2 -0.4

-0.8

0.6

0.6

0.4

0.4

0.0

dv/P

δ18Ov/P

* *

*

0.2

-0.2 -0.4

0.0

-0.2 -0.4

*

-0.6 -0.8

-0.6 -0.8

0.6

**

* *

**

0.6

*

*

*

**

0.4

dv/Q

0.2 0.0 -0.2

0.2 0.0

-0.2

-0.4

**

***

-0.8

TE D

-0.4

-0.6

autumn spring autumn spring autumn winter summer winter summer winter 2014

324

** *

M AN U

-0.8

** * * *

**

-0.6

0.2

δ18Ov/Q

0.0 -0.2 -0.4

***

-0.6

Whole period Non-rainy days Rainy days

0.2

0.0

0.4

(b)

0.4 dv/T

δ18Ov/T

0.2

** *

SC

0.8

RI PT

319

2015

-0.6

* *

-0.8

* *** *** * * * **

autumn spring autumn spring autumn winter summer winter summer winter

2016

2014

2015

2016

Fig. 5. Comparisons of linear relationships between daily mean δ18Ov and temperature, precipitation, absolute

326

humidity near the surface in different seasons in Changsha (a). Same as (a) but for dv (b). Asterisk indicates p < 0.01,

327

the same below.

AC C

328

EP

325

As shown in Figs. 2, 4 and 5, on the three time scales, including diurnal, intra-seasonal and seasonal

329

time scales, dv was mostly inversely proportional to temperature and absolute humidity, and showed no

330

significant relationship with precipitation. In addition, precipitation always depleted atmospheric water

331

vapor in

332

various on different time scales.

333

18

O. By comparison, the relationships of δ18Ov with temperature and absolute humidity were

On the short time scales, including diurnal and intra-seasonal time scales, the positive correlations of 17

ACCEPTED MANUSCRIPT δ18Ov with temperature and absolute humidity reflected the relationship between the stable isotopes and

335

temperature of the same air mass. Increased atmospheric humidity induced by elevated temperature and

336

enhanced evaporation led to the enrichment of stable isotopic signatures in atmospheric water vapor; on

337

the contrary, decreased atmospheric humidity induced by a declining temperature and evaporation led to

338

the depletion of stable isotopic signatures in atmospheric water vapor.

RI PT

334

On seasonal and sometimes on intra-seasonal time scales, the negative correlations of δ18Ov with

340

temperature and absolute humidity reflected the interactions across different air masses. In the cold

341

season, continental air masses resulted in a low mean temperature, atmospheric humidity, and enriched

342

stable vapor isotopic signatures. In the warm season, however, marine air masses resulted in a high mean

343

temperature, atmospheric humidity, and depleted stable vapor isotopic signatures. Consequently, changes

344

in vapor isotopes showed markedly negative correlations with those of temperature and atmospheric

345

humidity. Further analysis of the relationships between vapor isotopic compositions and temperature and

346

atmospheric humidity showed that on different time scales, the effects of temperature and atmospheric

347

humidity on vapor isotopes were differential, and sometimes opposite relationships could be recorded.

348

For example, significant positive correlations were recorded between δ18Ov and temperature and between

349

δ18Ov and absolute humidity in the fall and winter of 2014 (Fig. 6). However, during the whole period

350

from October 2014 to February 2015, significant negative relationships were observed between δ18Ov and

351

temperature and between δ18Ov and absolute humidity. These results illustrate the strong influence of

352

water vapor sources on vapor isotopic compositions on an intra-seasonal time scale. But it can be noted

353

that the correlation patterns between δ18Ov and temperature and absolute humidity discussed above do not

354

always exist, suggesting there are more influencing factors besides water vapor sources, which needs

355

further investigation.

AC C

EP

TE D

M AN U

SC

339

18

ACCEPTED MANUSCRIPT 0

0

(a)

-5

(b) δ18Ov=0.74Q-9.42

-5

δ18Ov=0.19T-7.85

(n=82, r=0.47)

-10

-10

-15

-15 δ18Ov=-0.49T-4.03

-20 18

-25

(n=50, r=0.54) autumn

δ18Ov=-0.49Q-4.03

δ18Ov=0.72Q-21.90

δ Ov=0.37T-22.48

-25

-20

(n=132, r= -0.53)

(n=132, r= -0.59)

(n=50, r=0.41)

winter

-30

autumn

-30 0

5

10

15 T (°°C)

356

20

25

30

RI PT

δ18Ov (‰)

(n=82, r=0.34)

2

4

6

8

10

winter 12

Q (g kg-1)

Fig. 6. The linear relationships between daily mean δ18Ov and temperature (a), absolute humidity (b) near the surface

358

in fall and winter, 2014. Linear regression lines and correlation coefficient r are also shown.

SC

357

3.3. Comparisons between isotopic compositions in vapor and precipitation

360

3.3.1. Comparisons of temporal variations

M AN U

359

It is of great significance to understand the interactions between precipitation and water vapor during

362

the descent process of each precipitation event and thus to explain the mechanisms of the variations of

363

stable isotopic compositions in atmospheric water vapor and precipitation (Dansgaard, 1964; Zhang, Xie,

364

& Yao, 1998). Fig. 7 shows the daily variations of δ18O and d in atmospheric water vapor and

365

precipitation, as well as the δ18Oe predicted from the equilibrium theory during the whole sampling period

366

in Changsha.

AC C

EP

TE D

361

19

ACCEPTED MANUSCRIPT 10 5

(a)

δ18Ov

(b)

dv

δ18Op

δ18Oe

18

δ O (‰)

0 -5 -10 -15 -20 -30 60 50

dp

de

40 d (‰)

30 20 10 -10 -20

367

11 2014

1

3

5 7 2015

9

11

1

3

5 7 2016

M AN U

9

SC

0

RI PT

-25

9

11

1 3 2017

368

Fig. 7. Comparisons of the temporal variations of daily mean δ18O (a) and d (b) in atmospheric water vapor and in

369

precipitation near the surface in Changsha. The subscript e identifies the daily mean stable isotopic compositions in

370

atmospheric water vapor in equilibrium with those in precipitation.

During the whole survey period, the δ18Ov values varied from -28.15 ‰ to -1.18 ‰, with the mean

372

value of -13.94 ‰, and the dv values varied from -17.47 ‰ to 52.49 ‰, with the mean value of 18.18 ‰.

373

By contrast, the δ18Op values varied from -17.52 ‰ to 6.73 ‰, with the amount-weighted mean value of

374

-4.90 ‰, and the dp values varied from -8.82 ‰ to 33.32 ‰, with the amount-weighted mean value of

375

14.25 ‰. More negative δ18O values were observed in atmospheric water vapor compared to those in

376

precipitation. The difference between δ18O in precipitation and in atmospheric water vapor (∆δ) varied

377

from 1.24 ‰ to 20.13 ‰ with the mean value of 9.18 ‰ on rainy days. The low ∆δ values mainly

378

occurred during the winter of 2014 (Fig. 7a). The more depleted stable isotopic signatures in atmospheric

379

water vapor were closely associated with the evaporative processes of new rainfall reaching the ground

380

and of raindrops during their descent. In general, raindrops that fell in the unsaturated atmospheric

381

environment with low humidity would experience strong evaporative processes. These processes caused

AC C

EP

TE D

371

20

ACCEPTED MANUSCRIPT the heavy isotopic water molecules (e.g., 1H218O and 1H2H16O) in raindrops to become enriched and their

383

counterparts in atmospheric water vapor to gradually become depleted. Saxena and Eriksson (1985) have

384

identified the evaporated strength of falling raindrops below the cloud according to the magnitude of

385

difference between predicted-δ18Oe and δ18Ov, showing that smaller difference indicated the greater

386

evaporative enrichment of stable isotopic signatures in raindrops below the cloud and vice versa.

387

Following Saxena’s inference, the weaker evaporative enrichment of stable isotopic signatures in

388

raindrops below the cloud appeared during the winter of 2014, whereas there was stronger evaporative

389

enrichment during other periods. The d was higher in atmospheric water vapor than that in precipitation

390

because the light isotopic water molecules were preferentially lost from raindrops during the evaporative

391

processes. The average ∆d value of 3.43 ‰ was found during rainy days. The trends of seasonal

392

variations of both δ18O and d in atmospheric water vapor and precipitation were generally consistent,

393

suggesting that these variations were influenced by large-scale climate systems and the interactions

394

between local precipitation and water vapor.

395

3.3.2. Comparisons of the relationships between δ2H and δ18O in vapor and precipitation

TE D

M AN U

SC

RI PT

382

The linear correlation between δ2H and δ18O in atmospheric water vapor here was described as the

397

Meteoric Vapor Line (MVL), similar to the definition of the Meteoric Water Line (MWL; Craig, 1961).

398

Both MVL and MWL are powerful tools and are widely applied in isotope hydrology. The MVL and

399

MWL in Changsha calculated from daily stable isotope data for the whole sampling period were

400

expressed as Equations (3) and (4), respectively.

AC C

EP

396

401

δ2Hv = 7.59δ18Ov + 12.43 (n=868, r=0.98)

(3)

402

δ2Hp = 8.28δ18Op + 15.98 (n=352, r=0.98),

(4)

403

The slope of the MVL was lower than 8, while that of the MWL was greater than 8, although the 21

ACCEPTED MANUSCRIPT 404

intercepts of both the MVL and MWL were greater than 10 therein. Numerous previous studies have demonstrated that the slope of MWL and the corresponding linear

406

correlation coefficient between δ2H and δ18O in precipitation may be considered as an important indicator

407

to identify the degree of isotopic enrichment of raindrops below the cloud (Crawford, Hollins, Meredith,

408

& Hughes, 2017; Dansgaard, 1964; Peng, Mayer, Harris, & Krouse, 2007; Salamalikis, Argiriou, &

409

Dotsika, 2016). In general, greater slopes of the MWLs and more significant linear correlations indicate

410

the weaker secondary evaporative effect on the stable isotopic compositions of raindrops below the cloud

411

and vice versa. The slopes of the MWLs and δ2Hp/δ18Op correlations in different seasons were analysed in

412

Changsha (Table 2). As shown in Table 2, the lowest slope (7.64) and a relatively non-significant linear

413

correlation (r = 0.95) occurred in the cold season of 2014, suggesting that the stable isotopic signatures in

414

raindrops experienced great evaporative enrichment below the cloud; by contrast, the higher slopes and

415

more significant linear correlations were recorded during other periods, indicating that the stable isotopic

416

signatures in raindrops experienced relatively weak evaporative enrichment below the cloud. This result

417

contrasts with the results derived from the difference between δ18Oe and δ18Ov.

TE D

M AN U

SC

RI PT

405

From the seasonal variations of δ18Ov, the stable isotopic signatures in the atmospheric water vapor

419

tend to be depleted with the occurrence of rainfall, probably resulting from the evaporation of new rainfall

420

and raindrops below the cloud. By comparing the values between the MVLs on non-rainy and rainy days

421

in the same period (Table 2), it can be found that the contribution of rainfall events caused the increases

422

of the slope and intercept of the MVL, supplemented by evaporated vapor. In classic isotope fractionation

423

theory (Clark & Fritz, 1997; Dansgaard, 1964), the evaporation of raindrops will result in the decreases of

424

the slope and intercept of the δ2H-δ18O linear relationship in falling raindrops but the increases of the

425

slope and intercept of the δ2H-δ18O linear relationship in the evaporated vapor. Affected by the evaporated

AC C

EP

418

22

ACCEPTED MANUSCRIPT vapor, which has a higher slope and intercept, the slope and intercept of the MVL will increase. Because

427

the secondary evaporation of raindrops is a continuous process from cloud to ground, its effect on

428

atmospheric water vapor near the surface is lighter than the evaporation from the underlying surface on

429

rainy days.

RI PT

426

430

Table 2 Comparisons of the relationships between daily mean δ2H and δ18O in atmospheric water vapor and

431

precipitation in different seasons in Changsha.

Cold season in 2014

Warm season in 2015

Whole period

7.83

12.92

0.98*

Non-rainy days

7.73

11.36

0.98*

Rainy days

8.01

15.60

Whole period

7.54

13.48

Non-rainy days

7.47

12.91

Rainy days

7.66

14.26

Whole period

7.21

11.78

Non-rainy days

7.22

11.93

Rainy days

7.33

12.99

Whole period

8.00

14.57

Non-rainy days

7.99

13.69

Rainy days Whole period Cold season in 2015

Non-rainy days Rainy days Whole period Non-rainy days Rainy days

0.98*

13.96

0.99*

2015/04-2015/09 2016/04-2016/09

7.96

16.13

0.97*

2014/10-2015/03 2015/10-2016/03 2016/10-2017/03

7.64

14.01

0.95*

2014/10-2015/03

8.53

14.19

0.99*

2015/04-2015/09

8.32

19.34

0.98*

2015/10-2016/03

8.40

13.72

0.99*

2016/04-2016/09

7.76

16.15

0.97*

2016/10-2017/03

0.99*

0.97*

0.98*

0.98*

0.98*

17.20

0.99*

8.24

0.97*

6.98

5.08

0.97*

7.48

12.46

0.96*

7.61

10.19

0.98*

7.49

8.99

0.98*

7.80

12.11

0.98*

Whole period

7.67

15.31

0.98*

7.62

15.37

0.97*

16.31

0.98*

7.82

δ2Hp/δ18Op

0.98*

Non-rainy days Rainy days

432

0.98*

8.10

AC C

Cold season in 2016

8.46

Period

Intercept (‰)

0.99*

7.19

EP

Warm season in 2016

Slope

M AN U

Cold season

Intercept (‰)

TE D

Warm season

MWL δ2Hp/δ18Op

Slope

SC

MVL Timescale

According to the aforementioned analyses, on average, the strong evaporative enrichment in falling

433

raindrops, but weak evaporation from the underlying surface, as well as a great difference between

434

predicted-δ18Oe and δ18Ov near the surface, appeared in the cold season of 2014, suggesting that the

435

isotopic compositions in atmospheric water vapor and precipitation were in a non-equilibrium state.

436

Under the action of weak surface evaporation, the slope and intercept of the MVL increased by only 0.11

437

and 1.06 ‰ on rainy days, respectively, compared with those on non-rainy days (the increment is 23

ACCEPTED MANUSCRIPT calculated as arithmetic difference between the both). By contrast, the weak evaporative enrichment in

439

falling raindrops but the strong evaporation from the underlying surface, as well as the small difference

440

between the predicted-δ18Oe and δ18Ov near the surface, happened in other periods, indicating that the

441

isotopic compositions of atmospheric water vapor and precipitation were in or near an equilibrium state.

442

Under the action of strong surface evaporation, compared with those on non-rainy days, the slope and

443

intercept of the MVL increased by 0.11 and 3.51 ‰ in the warm season of 2015, respectively; by 0.5 and

444

7.38 ‰ in the cold season of 2015, respectively; by 0.31 and 3.12 ‰ in the warm season of 2016,

445

respectively; and by 0.20 and 0.54 ‰ in the cold season of 2016, respectively. For the whole survey

446

period, the evaporation of new rainfall and falling raindrops induced the increases of the slope and

447

intercept of the MVL by 0.2 and 1.98 ‰ on rainy days, respectively, compared with those on non-rainy

448

days.

449

3.4. Variations of stable water isotopes during typical weather patterns

450

3.4.1. Variations of vapor isotopes during continuous non-rainy days

SC

M AN U

TE D

451

RI PT

438

Under the influence of the seasonal movement of the subtropical high over the western Pacific, the climate exhibits a dry and hot weather pattern from July to September of every year in Changsha. For

453

example, there was not any precipitation from September 12 to September 27 of 2016 in Changsha.

454

During this period, the temperature varied from 19.24 to 34.90 °C, with a mean value of 26.15 °C; the

455

absolute humidity varied from 7.51 to 18.58 g kg-1, with a mean value of 13.75 g kg-1; and the δ18Ov

456

values varied from -24.48 ‰ to -12.77 ‰, with a mean value of -17.04 ‰. Interestingly, all these

457

statistics were lower than the mean values of corresponding variables in September except for the air

458

temperature. The heavy isotopes in different water bodies of the underlying surface and thus those in

459

atmospheric water vapor near the surface were successively enriched because no rainfall event happened

AC C

EP

452

24

ACCEPTED MANUSCRIPT throughout this process. From Fig. 8, the isotopes in atmospheric water vapor exhibited a positive trend

461

from September 12 to September 27 but showed a negative trend from September 15 to September 18. It

462

can be found that, after analysing simultaneous synoptic charts, super typhoon Morandi landed in Xiamen,

463

China on September 15. The daily minimum temperature and atmospheric precipitable water dropped to

464

22.28 °C and 32.84 mm on September 18 from 24.73 °C and 54.92 mm on September 15, respectively,

465

influenced by the peripheral cloud system of the super typhoon. Simultaneously, dv also showed obvious

466

fluctuations. The changes in δ18Ov and dv during the passage of typhoon Morandi are similar to the results

467

reported by Fudeyasu et al. (2008), which showed the water vapor was isotopically depleted due to the

468

rainout effect. This synoptic process indicates that the effects of a weather system can still be detected by

469

tracers of the vapor isotopes, even though no rainfall occurs. 14

16

17

Date 19 20

18

21

22

23

24

25

26

27

(a)

20 18 16 14 12 10 8 6

EP

(b)

(c)

T

Q

Q (g kg-1)

TE D

typhoon "Morandi"

00:00

00:00

00:00

00:00

00:00

00:00

00:00

00:00

00:00

00:00

00:00

00:00

00:00

00:00

00:00

(d)

00:00

PW (mm)

470

15

00:00

-12 -14 -16 -18 -20 -22 -24 -26 30 25 20 15 10 5 0 -5 36 34 32 30 28 26 24 22 20 18 55 50 45 40 35 30 25

13

AC C

T (°°C)

dv (‰)

δ18Ov (‰)

12

M AN U

SC

RI PT

460

Time (CST)

471

Fig. 8. Variations of the hourly mean δ18Ov (a), dv (b), temperature, and absolute humidity (c) near the surface, as

472

well as the hourly mean precipitable water (d) from September 12 to September 27 of 2016 in Changsha.

473 25

ACCEPTED MANUSCRIPT 474

3.4.2. Variations of stable water isotopes during continuous rainy days In the Yangtze River Basin and Huai River Basin of China, the period from middle June to early July

476

of every year is called the “plum rains” season (Tao, Zhao, & Chen, 1958; Zhang, Kuang, Guo, & Zhou,

477

2006). The precipitation has the characteristics of continuity and high intensity during this season. As an

478

example, Fig. 9 shows the hourly variations in δ18O and d in atmospheric water vapor and in precipitation,

479

precipitation, air temperature, and absolute humidity near the surface, as well as the precipitable water

480

during a continuous rainfall process in the plum rains season of 2016 in Changsha. From June 30 to July 6,

481

the accumulated precipitation reached 226.59 mm, accounting for 20.35 % of the warm season of 2016;

482

the mean temperature was 26.0 °C and the mean absolute humidity was 18.22 g kg-1 (Table 3). Based on

483

the temporal patterns of precipitation variations in this period, the variation of δ18Ov is divided into three

484

stages. In the first stage, from 00:00 on June 30 to 19:00 on July 2, δ18Ov was generally maintained at a

485

relatively high level with less precipitation, high mean temperature, and absolute humidity; in the second

486

stage, from 20:00 on July 2 to 16:00 on July 4, δ18Ov fluctuated significantly and

487

averagely relative to that in the first stage, with the accumulated precipitation of 197.13 mm, accounting

488

for 87.0 % of total precipitation in this period, as well as reduced mean temperature and absolute

489

humidity; in the third stage, from 17:00 on July 4 to 00:00 on July 7, 18Ov was gradually depleted, with

490

less precipitation, low mean temperature, and absolute humidity.

AC C

EP

TE D

M AN U

SC

RI PT

475

26

18

Ov was depleted

1 First stage

5 6 Third stage

T

P

36 34 32 30 28 26 24 22

Q

30 20

00:00

16:00

08:00

00:00

16:00

08:00

00:00

M AN U

16:00

08:00

00:00

16:00

08:00

00:00

16:00

08:00

00:00

(d)

SC

(c)

16:00

0 75 70 65 60 55 50

08:00

10

22 21 20 19 18 17 16 15

-1

dp

Q (g kg )

dv

T (°°C)

(b)

00:00

δ Op

RI PT

18

δ Ov

16:00

18

Date 3 4 Second stage

2

(a)

00:00

PW (mm)

30

08:00

-4 -8 -12 -16 -20 -24 30 25 20 15 10 5 0 40

P (mm)

d (‰)

18

δ O (‰)

ACCEPTED MANUSCRIPT

Time (CST)

491

Fig. 9. Variations of the hourly mean δ18O (a), d (b) in atmospheric water vapor and in precipitation, precipitation,

493

temperature, and absolute humidity (c) near the surface, as well as the hourly mean precipitable water (d) from June

494

30 to July 6 of 2016 in Changsha.

495

Table 3 Comparisons of the mean δ18Ov, precipitation, mean temperature and absolute humidity near the surface at

496

different stages of continuous rainfall in Changsha.

Whole period First stage

Third stage

497 498

P (mm)

Mean T (°C)

Mean Q (g kg-1)

Relationships 18

δ Ov/T

δ18Ov/P

δ18Ov/Q

-18.89(2.55)

226.59

26.00

18.22

0.64*

-0.21

0.78*

-16.67(0.92)

11.43

27.46

18.89

-0.13

-0.17

0.39*

-19.35(2.13)

197.13

24.89

18.31

0.68*

-0.09

0.72*

17.32

*

-0.30

0.84*

AC C

Second stage

δ18Ov (‰)

EP

Period

TE D

492

-21.27(1.80)

18.04

25.11

0.79

Note. The number in the parentheses represents the standard deviation.

From the statistics in Table 3, the δ18Ov values of every stage displayed significant positive

499

correlations with absolute humidity, which were consistent with their relationships on a diurnal time scale.

500

However, the correlations of δ18Ov with temperature varied in different stages. Specifically, δ18Ov showed

501

an insignificant negative correlation with temperature in the first stage, but it did show significant positive

27

ACCEPTED MANUSCRIPT correlations with temperature in the last two stages, with an increase in temperature and more positive

503

δ18Ov (Fig. 9). Although rainfall always depleted vapor isotopes, the negative correlations between them

504

were not significant, especially in the second stage. It can be seen that 18Ov was not significantly depleted

505

in the second stage, which had the most precipitation, compared to that in the first stage, which had the

506

least precipitation. In addition, by analysing the variation of the precipitable water in the second stage, a

507

steady input of ambient water vapor might compensate for the loss of water vapor due to condensation

508

and offset the depletion effect of precipitation on vapor isotopes; thus, the decrease of δ18Ov was not

509

obvious in the second stage. The small fluctuation in dv also supports this hypothesis to some extent. With

510

the decrease of water vapor input and change of water vapor transport type (Li et al., 2015; Munksgaard,

511

Wurster, Bass, & Bird, 2012), as well as the contribution from surface evapotranspiration, the precipitable

512

water and precipitation tended to decrease, but δ18Ov began to rise steadily in the third stage. Although the

513

variations of precipitation isotopes can partly reflect the characteristics of weather conditions and water

514

vapor transports in different stages, the vapor isotopes provided details about the continuous variations of

515

relevant elements.

516

3.4.3. Variations of stable water isotopes during a strong cold wave

EP

TE D

M AN U

SC

RI PT

502

The cold wave and cold air moving down southward are the main weather systems in winter in East

518

Asia. In the period from January 20 to January 25, 2016, the whole of East Asia had experienced a wide

519

range of cooling, heavy rain and snow, and gale weather from north to south. Under the influence of

520

strong cold air from Central Siberian and East Siberian, the air temperatures of 179 weather stations over

521

North China, Huang Huai Basin, south of the Yangtze River, Southern China, and the Southwest fell

522

below the historical minimum in January since the establishments of these stations and those of 82 cities

523

and counties in Shandong, Jiangsu, Fujian, Sichuan, and other 19 provinces in China fell below the

AC C

517

28

ACCEPTED MANUSCRIPT 524

historical extreme values of minimum temperature (Jiang, Ma, & Wang, 2016). Fig. 10 exhibits the

525

temporal variations of vapor isotopes, precipitation isotopes, and related meteorological factors before

526

and after this strong cold wave passed through Changsha.

23

24

25

26

27

00:00

12:00

12:00

12:00 00:00

00:00

12:00

00:00

12:00

00:00

12:00

(d)

12:00 00:00

(c)

00:00

1

Q

15 10 5 0 -5

5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0

-1

P

M AN U

T

Q (g kg )

dp

T (°° C)

dv

00:00

(b)

12:00

δ18Op

18

12:00 00:00

δ Ov

12:00 00:00

(a)

28

SC

cold wave

2

0 25 20 15 10 5

22

RI PT

21

00:00

20

00:00 12:00

19

TE D

527

18

12:00

17

00:00

PW (mm)

P (mm)

d (‰)

δ18O (‰)

Date 4 0 -4 -8 -12 -16 -20 -24 50 40 30 20 10 0 -10 3

Time (CST)

Fig. 10. Variations of the hourly mean δ18O (a), d (b) in atmospheric water vapor and in precipitation, precipitation,

529

temperature, and absolute humidity (c) near the surface, as well as the hourly mean precipitable water (d) from

531

January 17 to January 28 of 2016 in Changsha.

From Fig. 10, the temperature and absolute humidity with mean values of 7.20 °C and 4.12 g kg-1,

AC C

530

EP

528

532

respectively, showed great fluctuations, but δ18Ov with a mean value of -15.94 ‰ varied smoothly from

533

00:00 on January 17 to 04:00 on January 20. Since the cold front arrived on January 20, the Changsha

534

area had begun an obvious precipitation and cooling process. The main rainfall processes occurred from

535

14:00 on January 20 to 07:00 on January 21 with 20.06 mm of precipitation and from 23:00 on January

536

21 to 09:00 on January 22 with 6.35 mm of precipitation. The daily mean temperature decreased from

537

6.62 °C on January 19 to 1.80 °C on January 22 with the lowest temperature of 0.88 °C; the absolute 29

ACCEPTED MANUSCRIPT humidity, with a mean value of 3.60 g kg-1, decreased slightly; δ18Ov showed a small fluctuation and was

539

almost equivalent with those in the period from 00:00 on January 17 to 04:00 on January 20, which were

540

unaffected by the cold front. After the cold front passed through Changsha, the δ18Ov and absolute

541

humidity decreased sharply, from -17.30 ‰ to -24.45 ‰ and from 2.88 to 1.60 g kg-1, respectively, from

542

00:00 to 13:00 on January 23. They were all maintained at low levels until 00:00 on January 25. With the

543

end of the cold wave process, the temperature, absolute humidity, and δ18Ov all began to rise and basically

544

returned to their former levels on January 26. During the strong cold wave, only 7 precipitation samples

545

were collected. All water samples, except one collected at 08:00 on January 27, showed small differences

546

in δ18Op and dp.

M AN U

SC

RI PT

538

During the cold wave, there was a close relationship between δ18Ov and absolute humidity, with a

548

correlation coefficient of 0.85 (p < 0.01). Because absolute humidity was primarily controlled by the

549

precipitable water, and the latter was dependent on the transportation of total water vapor during the

550

invasion of the cold wave, it can be inferred that the variation of δ18Ov was mainly caused by that of the

551

cold wave. Indeed, the variation of δ18Ov well indicates the activities of the cold wave.

TE D

547

The analyses above indicate that the short-term weather processes can be reflected in the temporal

553

evolutions of vapor isotopes. In addition, atmospheric humidity is a good predictor for vapor isotope on

554

the time scales from diurnal to a few days, which is similar to the finding by Lee et al. (2006).

555

4. Conclusions

AC C

556

EP

552

Long-term and continuous measurements of stable isotopic compositions in atmospheric water vapor

557

near the surface show the obvious multi time scale variation characteristics of vapor isotopes. On a

558

diurnal time scale, the δ18Ov values were more negative in the daytime and more positive at night, and the

559

variation of dv was the opposite on non-rainy days; however, the δ18Ov values were more positive in the 30

ACCEPTED MANUSCRIPT 560

daytime and more negative at night, and the variation of dv was not significant on rainy days; on a

561

seasonal time scale, affected by air masses with distinct thermodynamic properties, the δ18O and d values

562

in atmospheric water vapor were more negative in the warm season than in the cold season. On different time scales, including diurnal, a few days, intra-seasonal and seasonal time scales, dv

564

was generally inversely proportional to temperature and absolute humidity, and had no significant

565

relationship with precipitation; moreover, δ18Ov had a stable negative linear relationship with precipitation,

566

although sometimes the relationship was statistically insignificant. However, the relationships of δ18Ov

567

with temperature and absolute humidity were obviously different on different time scales. Overall, δ18Ov

568

had very significantly positive linear relations with temperature and absolute humidity on short time

569

scales, including diurnal, a few days, and intra-seasonal time scales, which indicated the relationship

570

between the stable isotopes and the temperature of the same air mass. The negative linear relationships

571

between δ18Ov and temperature and absolute humidity reflected the interactions between different air

572

masses on seasonal and sometimes intra-seasonal timescales.

SC

M AN U

TE D

573

RI PT

563

The change patterns of δ18O and d in atmospheric water vapor and precipitation were highly 18

consistent, but affected by below-cloud secondary evaporation, there was an average

575

9.18 ‰ relative to

576

whole sampling period, the local Meteoric Water Line in Changsha was δ2Hp = 8.28δ18Op + 15.98, the

577

Meteoric Vapor Line was δ2Hv = 7.59δ18Ov + 12.43. Under the influence of the evaporation of falling

578

raindrops and the underlying surface, the slopes and intercepts of the MVLs on rainy days all increased,

579

with average increases of 0.2 and 1.98 ‰, respectively, relative to those on non-rainy days. Generally, the

580

isotopic compositions of atmospheric water vapor were in equilibrium with those of precipitation during

581

each rainfall event, except for those in the cold season in 2014. And further analysis suggested that

EP

574

Op and an average dv enrichment of 3.43 ‰ relative to dp on rainy days. For the

AC C

18

Ov depletion of

31

ACCEPTED MANUSCRIPT compared with precipitation isotopes, the vapor isotopes can provide details about weather processes due

583

to the time-space continuity of water vapor.

584

Acknowledgements

585

RI PT

582

This work was supported by the National Natural Science Foundation of China (Grant No. 41571021) and the first-class disciplines (Geography) in Hunan Normal University (Grant No. 810006). The authors

587

would like to thank Huawu Wu, Guang Li, Yongqiang Zhou and Zidong Luo for their help.

588

Competing Interests: The authors declare that they have no competing interests.

589

References

590

Araguás, L. A., Froehlich, K., & Rozanski, K. (1998). Stable isotope composition of precipitation over

591

southeast Asia. Journal of Geophysical Research Atmospheres, 103(D22), 28721-28742.

592

https://doi.org/10.1029/98jd02582.

595 596 597 598 599 600 601

M AN U

TE D

30(4), 315-318. https://doi.org/10.1130/0091-7613(2002)030<0315:SDOOIM>2.0.CO;2.

EP

594

Bowen, G. J., & Wilkinson, B. (2002). Spatial distribution of δ18O in meteoric precipitation. Geology,

Cheng, H., Edwards, R. L., Sinha, A., Spötl, C., Yi, L., Chen, S. T., Kelly, M., Kathayat, G., Wang, X. F.,

AC C

593

SC

586

Li, X. L., Kong, X. G., Wang, Y. J., Ning, Y. F., & Zhang, H. W. (2016). The Asian monsoon over the

past

640,000

years

and

ice

age

terminations.

Nature,

534(7609),

640-646.

https://doi.org/10.1038/nature18591.

Clark, I. D., & Fritz, P. (1997). Tracing the hydrological cycle. In Environmental isotopes in hydrogeology. Florida: CRC press. Craig, H. (1961). Isotopic variations in meteoric waters. Science, 133(3465), 1702-1703.

32

ACCEPTED MANUSCRIPT 602

https://doi.org/10.1126/science.133.3465, 1702. Crawford, J., Hollins, S. E., Meredith, K. T., & Hughes, C. E. (2017). Precipitation stable isotope

604

variability and subcloud evaporation processes in a semi-arid region. Hydrological Processes, 31(1),

605

20-34. https://doi.org/10.1002/hyp.10885.

RI PT

603

Cui, J. P., Tian, L. D., Liu, Q., & Wen, R. (2014). Signal of Typhoon Phailin from Indian Ocean captured

607

by atmospheric water vapor isotope, central Tibetan Plateau (in Chinese with English abstract).

608

Chinese Science Bulletin, 59(35), 3526-3532. https://doi.org/10.1360/N972014-00601.

611 612 613 614

M AN U

610

Dansgaard, W. (1953). The abundance of O18 in atmospheric water and water vapour. Tellus, 5(4), 461-469. https://doi.org/10.3402/tellusa.v5i4.8697. Dansgaard,

W.

(1964).

Stable

isotopes

in

precipitation.

Tellus

A,

16(4),

436-468.

https://doi.org/10.1111/j.2153-3490.1964.tb00181.x.

Ehhalt, D. H. (1974). Vertical profiles of HTO, HDO, and H2O in the troposphere. National Center for

TE D

609

SC

606

Atmospheric Research Technical Note, NCAR-TN/STR-100: Boulder, Co. Frankenberg, C., Yoshimura, K., Warneke, T., Aben, I., Butz, A., Deutscher, N., Griffith, D., Hase, F.,

616

Notholt, J., Schneider, M., Schrijver, H., & Rockmann, T. (2009). Dynamic processes governing

617

lower-tropospheric HDO/H2O ratios as observed from space and ground. Science, 325(5946),

AC C

618

EP

615

1374-1377. https://doi.org/10.1126/science.1173791.

619

Fudeyasu, H., Ichiyanagi, K., Sugimoto, A., Yoshimura, K., Ueta, A., Yamanaka, M. D., & Ozawa, K.

620

(2008). Isotope ratios of precipitation and water vapor observed in Typhoon Shanshan. Journal of

621

Geophysical Research, 113(D12). https://doi.org/10.1029/2007JD009313.

622

Griffis, T. J., Wood, J. D., Baker, J. M., Lee, X., Xiao, K., Chen, Z., Welp, L. R., Schultz, N. M., Gorski,

623

G., Chen, M., & Nieber, J., (2016). Investigating the source, transport, and isotope composition of 33

ACCEPTED MANUSCRIPT 624

water vapor in the planetary boundary layer. Atmospheric Chemistry and Physics, 16(8), 5139-5157.

625

https://doi.org/10.5194/acp-16-5139-2016. Hu, Z. M., Yu, G. R., Zhou, Y. L., Sun, X. M., Li, Y. N., Shi, P. L., Wang, Y. F., Song, X., Zheng, Z. M.,

627

Zhang, L., & Li, S. G. (2009). Partitioning of evapotranspiration and its controls in four grassland

628

ecosystems: application of a two-source model. Agricultural and Forest Meteorology, 149(9),

629

1410-1420. https://doi.org/10.1016/j.agrformet.2009.03.014.

Jacob, H., & Sonntag, C. (1991). An 8-year record of the seasonal variation of 2H and 18O in atmospheric water

vapour

and

precipitation

at

Heidelberg,

632

https://doi.org/10.1034/j.1600-0889.1991.t01-2-00003.x.

Germany.

Tellus

B,

43(3),

291-300.

M AN U

631

SC

630

RI PT

626

Jiang, Q., Ma, X. K., & Wang, F. (2016). Analysis of the January 2016 Atmospheric Circulation and

634

Weather (in Chinese with English abstract). Meteorological Monthly, 42(4), 514-520.

635

https://doi.org/10.7519/j.issn.1000-0526.2016.04.016.

TE D

633

Johnsen, S. J., Dahl-Jensen, D., Gundestrup, N., Steffensen, J. P., Clausen, H. B., Miller, H.,

637

Masson-Delmotte, V., Sveinbjörnsdottir, A. E., & White, J. (2001). Oxygen isotope and

638

palaeotemperature records from six Greenland ice-core stations: Camp Century, Dye-3, GRIP,

639

GISP2,

Renland

and

NorthGRIP.

Journal

of

Quaternary

Science,

16(4),

299-307.

AC C

640

EP

636

https://doi.org/10.1002/jqs.622.

641

Jouzel, J. (1986). Isotopes in cloud physics: multiphase and multistage condensation processes. In

642

Handbook of Environmental Isotope Geochemistry, volume 2. The Terrestrial Environment B, Fritz,

643

P., Founter, J.C. (eds). Elsevier: Amsterdam; 6-105.

644

Lai, C. T., Ehleringer, J. R., Bond, B. J., & Paw U, K. T. (2006). Contributions of evaporation, isotopic

645

non-steady state transpiration and atmospheric mixing on the δ18O of water vapour in Pacific 34

ACCEPTED MANUSCRIPT 646

Northwest

coniferous

forests.

Plant,

647

https://doi.org/10.1111/j.1365-3040.2005.01402.x.

Cell

&

Environment,

29(1),

77-94.

Laskar, A. H., Huang, J. C., Hsu, S. C., Bhattacharya, S. K., Wang, C. H., & Liang, M. C. (2014). Stable

649

isotopic composition of near surface atmospheric water vapor and rain-vapor interaction in Taipei,

650

Taiwan. Journal of Hydrology, 519, 2091-2100. https://doi.org/10.1016/j.jhydrol.2014.10.017.

652

Lee, X. H., Smith, R., & Williams, J. (2006). Water vapour

18

O/16O isotope ratio in surface air in New

England, USA. Tellus B, 58(4), 293-304. https://doi.org/10.1111/j.1600-0889.2006.00191.x.

SC

651

RI PT

648

Li, J., Tao, T., Pang, Z. H., Tan, M., Kong, Y. L., Duan, W. H., & Zhang, Y. W. (2015). Identification of

654

different moisture sources through isotopic monitoring during a storm event. Journal of

655

Hydrometeorology, 16(4), 1918-1927. https://doi.org/10.1175/jhm-d-15-0005.1.

659 660 661

TE D

658

Chimie Physique, 68, 1423-1436. https://doi.org/10.1051/jcp/1971681423. McBean, G. A. (1971). The variations of the statistics of wind, temperature and humidity fluctuations with stability. Boundary-Layer Meteorology, 1(4), 438-457. Monin, A. S., & Obukhov, A. (1954). Basic laws of turbulent mixing in the surface layer of the

EP

657

Majoube, M. (1971). Fractionnement en oxygene 18 et en deuterium entre l’eau et sa vapeur. Journal de

atmosphere. Tr Akad Nauk SSSR Geophiz Inst, 151, 163-187.

AC C

656

M AN U

653

662

Munksgaard, N. C., Wurster, C. M., Bass, A., & Bird, M. I. (2012). Extreme short-term stable isotope

663

variability revealed by continuous rainwater analysis. Hydrological Processes, 26(23), 3630-3634.

664

https://doi.org/10.1002/hyp.9505.

665

Peng, H., Mayer, B., Harris, S., & Krouse, H. R. (2007). The influence of below-cloud secondary effects

666

on the stable isotope composition of hydrogen and oxygen in precipitation at Calgary, Alberta,

667

Canada.

Tellus

B:

Chemical

and 35

Physical

Meteorology,

59(4),

698-704.

ACCEPTED MANUSCRIPT

669

https://doi.org/10.3402/tellusb.v59i4.17049. Salamalikis, V., Argiriou, A. A., & Dotsika, E. (2016). Isotopic modeling of the sub-cloud evaporation

670

effect

671

https://doi.org/10.1016/j.scitotenv.2015.11.072.

precipitation.

Science

of

Total

Environment,

544,

1059-1072.

Saxena, R. K., & Eriksson, E. (1985). Hydrometeorological interpretation of isotopic data on atmospheric

673

moisture

674

https://doi.org/10.1017/s0260305500006133.

and

precipitation.

Annals

of

Glaciology,

7,

181-184.

Schoch-Fischer, H., Rozanski, K., Jacob, H., Sonntag, C., Jouzel, I., Ostlund, G., & Geyh, M. A. (1984).

M AN U

675

the

SC

672

in

RI PT

668

18

O and 3H in atmospheric water

676

Hydrometeorological factors controlling the time variation of D,

677

vapour and precipitation in the northern westwind belt. Isotope Hydrology 1983. International

678

Atomic Energy Agency: Vienna; 3-31.

Sidorova, O. V., Siegwolf, R. T. W., Myglan, V. S., Ovchinnikov, D. V., Shishov, V. V., Helle, G.,

680

Loader, N. J., & Saurer, M. (2013). The application of tree-rings and stable isotopes for

681

reconstructions of climate conditions in the Russian Altai. Climatic Change, 120(1-2), 153–167.

682

https://doi.org/10.1007/s10584-013-0805-5.

684 685

EP

Srivastava, R., Ramesh, R., Gandhi, N., Jani, R. A., & Singh, A. K. (2015). Monsoon onset signal in the

AC C

683

TE D

679

stable oxygen and hydrogen isotope ratios of monsoon vapor. Atmospheric Environment, 108, 117-124. https://doi.org/10.1016/j.atmosenv.2015.02.062.

686

Stewart, M. K. (1975). Stable isotope fractionation due to evaporation and isotopic exchange of falling

687

waterdrops: applications to atmospheric processes and evaporation of lakes. Journal of Geophysical

688

Research, 80(9), 1133-1146. https://doi.org/10.1029/jc080i009p01133.

689

Strong, M., Sharp, Z. D., & Gutzler, D. S. (2007). Diagnosing moisture transport using D/H ratios of 36

ACCEPTED MANUSCRIPT 690

water vapor. Geophysical Research Letters, 34(3), L03404. https://doi.org/10.1029/2006GL028307. Tao, S. Y., Zhao, Y. J., & Chen, X. M. (1958). The relationship between Meiyu in far east and the

692

behavior of circulation over Asia (in Chinese with English abstract). Acta Meteorologica Sinica,

693

29(2), 119-134. https://doi.org/10.11676/qxxb1958.014.

RI PT

691

Taylor, C. B. (1968). Die isotopenzusammensetzung des atmosphärischen Wasserdampfs in

695

Höhenbereich 500 bis 5000 m über Festland. Doctoral Thesis, 2 Physikalisches Institute der

696

University. Heildelberg, 54p.

SC

694

Thompson, L. G., Yao, T., Mosley-Thompson, E., Davis, M. E., Henderson, K. A., & Lin, P. N. (2000).

698

A high-resolution millennial record of the south Asian monsoon from Himalayan ice cores. Science,

699

289(5486), 1916-1919. https://doi.org/10.1126/science.289.5486.1916.

M AN U

697

Tremoy, G., Vimeux, F., Mayaki, S., Souley, I., Cattani, O., Risi, C., & Favreau, G., Oi, M. (2012). A

701

1-year long δ18O record of water vapor in Niamey (Niger) reveals insightful atmospheric processes

702

at

703

https://doi.org/10.1029/2012GL051298.

TE D

700

different

timescales.

Geophysical

Research

Letters,

39(8).

Wei, Z., Yoshimura, K., Okazaki, A., Ono, K., Kim, W., Yokoi, M., & Lai, C. T. (2016). Understanding

705

the variability of water isotopologues in near-surface atmospheric moisture over a humid subtropical

707

AC C

706

EP

704

rice

paddy

in

Tsukuba,

Japan.

Journal

of

Hydrology,

533,

91-102.

https://doi.org/10.1016/j.jhydrol.2015.11.044.

708

Welp, L. R., Lee, X., Kim, K., Griffis, T. J., Billmark, K. A., & Baker, J. M. (2008). δ18O of water vapour,

709

evapotranspiration and the sites of leaf water evaporation in a soybean canopy. Plant, Cell &

710

Environment, 31(9), 1214-1228. https://doi.org/10.1111/j.1365-3040.2008.01826.x.

711

Welp, L. R., Lee, X. H., Griffis, T. J., Wen, X. F., Xiao, W., Li, S. G., Sun, X. M., Hu, Z. M., Val Martin, 37

ACCEPTED MANUSCRIPT 712

M., & Huang, J. P. (2012). A meta-analysis of water vapor deuterium-excess in the midlatitude

713

atmospheric

714

https://doi.org/10.1029/2011GB004246.

surface

layer.

Global

Biogeochemical

Cycles,

26(3).

Wen, X. F., Sun, X. M., Zhang, S. C., Yu, G. R., Sargent, S. D., & Lee, X. H. (2008). Continuous

716

measurement of water vapor D/H and 18O/16O isotope ratios in the atmosphere. Journal of Hydrology,

717

349(3), 489-500. https://doi.org/10.1016/j.jhydrol.2007.11.021.

719

isotope

ratios

in

Beijing,

China.

720

https://doi.org/10.1029/2009JD012408.

Journal

of

SC

Wen, X. F., Zhang, S. C., Sun, X. M., Yu, G. R., & Lee, X. H. (2010). Water vapor and precipitation Geophysical

Research,

115(D1).

M AN U

718

RI PT

715

Wu, H. W., Zhang, X. P., Li, X. Y., Li, G., & Huang, Y. M. (2015). Seasonal variations of deuterium and

722

oxygen-18 isotopes and their response to moisture source for precipitation events in the subtropical

723

monsoon region. Hydrological Processes, 29(1), 90-102. https://doi.org/10.1002/hyp.10132.

TE D

721

Yao, T. D., Masson-Delmotte, V., Gao, J., Yu, W. S., Yang, X. X., Risi, C., Sturm, C., Werner, M., Zhao,

725

H. B., He, Y., Ren, W., Tian, L. D., Shi, C. M., & Hou, S. G. (2013). A review of climatic controls

726

on δ18O in precipitation over the Tibetan Plateau: observations and simulations. Reviews of

727

Geophysics, 51(4), 525-548. https://doi.org/10.1002/rog.20023.

AC C

EP

724

728

Yao, T. D., Thompson, L. G., Qin, D. H., Tian, L. D., Jiao, K. Q., Yang, Z. H., & Xie, C. (1996).

729

Variations in temperature and precipitation in the past 2000 a on the Xizang (Tibet)

730

Plateau——Guliya ice core record. Science in China: Series D, 39(4), 425-433.

731

Yin, C. L., Yao, T. D., Tian, L. D., Liu, D. N., Yu, W. S., & Qu, D. M. (2008). Temporal variations of

732

δ18O of atmospheric water vapor at Delingha. Science in China Series D: Earth Sciences, 51(7),

733

966-975. https://doi.org/10.1007/s11430-008-0076-6. 38

ACCEPTED MANUSCRIPT 734

Yoshimura, K. (2015). Stable water isotopes in climatology, meteorology, and hydrology: A review.

735

Journal

of

the

Meteorological

736

https://doi.org/10.2151/jmsj.2015-036.

Society

of

Japan,

93(5),

513-533.

Yu, W., Tian, L., Ma, Y., Xu, B., & Qu, D. (2015). Simultaneous monitoring of stable oxygen isotope

738

composition in water vapour and precipitation over the central Tibetan Plateau. Atmospheric

739

Chemistry and Physics, 15(18), 10251-10262. https://doi.org/10.5194/acp-15-10251-2015.

RI PT

737

Yu, W. S., Yao, T. D., Tian, L. D., Wang, Y., & Yin, C. L. (2005). Isotopic composition of atmospheric

741

water vapor before and after the monsoon’s end in the Nagqu River Basin. Chinese Science Bulletin,

742

50(23), 2755-2760. https://doi.org/10.1360/982005-802.

M AN U

SC

740

743

Yuan, G. F., Zhang, N., Sun, X. M., Wen, X. F., & Zhang, S. C. (2010). Partitioning wheat field

744

evapotranspiration using Keeling Plot method and continuous atmospheric vapor δ18O data (in

745

Chinese

746

https://doi.org/10.3773/j.issn.1005-264x.2010.02.008.

English

abstract).

Journal

of

Plant

Ecology,

34(2),

170-178.

TE D

with

Zhang, S. C., Sun, X. M., Wang, J. L., Yu, G. R., & Wen, X. F. (2011). Short-term variations of vapor

748

isotope ratios reveal the influence of atmospheric processes. Journal of Geographical Sciences,

749

21(3), 401-416. https://doi.org/10.1007/s11442-011-0853-6.

AC C

EP

747

750

Zhang, X. P., Guan, H. D., Zhang, X. Z., Zhang, W. J., & Yao, T. C. (2016). Numerical experiments on

751

the impacts of surface evaporation and fractionation factors on stable isotopes in precipitation.

752 753

Asia-Pacific

Journal

of

Atmospheric

Sciences,

52(3),

327-339.

https://doi.org/10.1007/s13143-016-0008-x.

754

Zhang, X. P., Liu, J. M., Nakawo, M., & Xie, Z. C. (2009). Vapor origins revealed by deuterium excess in

755

precipitation in southwest China (in Chinese with English abstract). Journal of Glaciology and 39

ACCEPTED MANUSCRIPT 756 757 758

Geocryology, 31(4), 613-619. Zhang, X. P., Xie, Z. C., & Yao, T. D. (1998). Mathematical modeling of variations on stable isotopic ratios in falling raindrops. Journal of Meteorological Research, 12(2), 213-220. Zhang, X. P., Yang, Z. L., Niu, G. Y., & Wang, X. Y. (2009). Stable water isotope simulation in different

760

reservoirs of Manaus, Brazil, by Community Land Model incorporating stable isotopic effect.

761

International Journal of Climatology, 29(5), 619-628. https://doi.org/10.1002/joc.1740.

RI PT

759

Zhang, X. P., Zhang, X. Z., Guan, H. D., Huang, Y. M., & Wu, H. W. (2012). Spatiotemporal

763

distributions of δD in atmospheric water vapor based on TES data during 2004-2009. Journal of

764

Meteorological Research, 26(6), 683-699. https://doi.org/10.1007/s13351-012-0602-5.

M AN U

SC

762

765

Zhang, Y. C., Cai, Y. Z., Stephen, P., Matthew, F. M., Yang, F., Wang, Q., & Shen, Y. J. (2011).

766

Preliminary research on isotopic composition of water vapor in irrigated fields (in Chinese with

767

English

768

https://doi.org/10.3724/SP.J.1011.2011.01060.

Chinese

Journal

of

Eco-Agriculture,

19(5),

1060-1066.

TE D

769

abstract).

Zhang, Y., Kuang, X., Guo, W., & Zhou, T. (2006). Seasonal evolution of the upper-tropospheric westerly

jet

core

over

771

https://doi.org/10.1029/2006GL026377.

East

Asia.

AC C

EP

770

40

Geophysical

Research

Letters,

33(11).

ACCEPTED MANUSCRIPT

1

Highlights: Rare high resolution vapor isotopic data in the subtropical monsoon region.

3

Environmental controls on vapor isotopes on different time scales are discussed.

4

Isotopic compositions of atmospheric water vapor and precipitation are compared.

5

Detection of short-duration weather processes using vapor isotopes appears possible.

AC C

EP

TE D

M AN U

SC

RI PT

2