Temporal and Spatial Variations of Precipitation <i>δ</i><sup>18</sup>O and Controlling Factors on the Pearl River Basin and Adjacent Regions

Ruan, Yunfeng; Liu, Zhaofei; Yao, Zhijun; Wang, Rui

doi:https://doi.org/10.1155/2018/1419326

Advances in Meteorology

On this page

Abstract Introduction Methods Results and Discussion Conclusions Conflicts of Interest Acknowledgments References Copyright Related Articles

Research Article | Open Access

Volume 2018 | Article ID 1419326 | https://doi.org/10.1155/2018/1419326

Temporal and Spatial Variations of Precipitation δ¹⁸O and Controlling Factors on the Pearl River Basin and Adjacent Regions

Yunfeng Ruan,^1,2Zhaofei Liu,¹Zhijun Yao,¹and Rui Wang¹

Academic Editor: Paolo Madonia

Received24 Sept 2017

Revised15 Dec 2017

Accepted31 Dec 2017

Published20 Feb 2018

Abstract

Based on the precipitation δ¹⁸O values from the datasets of the Global Network of Isotopes in Precipitation (GNIP), the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) Reanalysis data, and previous researches, we explored the temporal and spatial variations of precipitation δ¹⁸O in a typical monsoon climate zone, the Pearl River basin (PRB), and adjacent regions. The results showed that the temporal variations of precipitation δ¹⁸O for stations should be correlated with water vapor sources, the distance of water vapor transport, the changes in location, and intensity of the intertropical convergence zone (ITCZ) rather than “amount effect.” Meanwhile, local meteorological and geographical factors showed close correlations with mean weighted precipitation δ¹⁸O values, suggesting that “altitude effect” and local meteorological conditions were significant for the spatial variations of precipitation δ¹⁸O. Moreover, we established linear regression models for estimating the mean weighted precipitation δ¹⁸O values, which could better estimate variations in precipitation δ¹⁸O than the Bowen and Wilkinson model in the PRB and adjacent regions.

1. Introduction

Meteoric precipitation is a significant part of the circulation of natural water, and the composition of stable isotopes is closely related to meteorological factors (e.g., temperature, relative humidity, and precipitation amount) and geographical factors (e.g., latitude, altitude, and distance from the moisture transport source), which have shown a sensitive response to environmental change [1–3]. The stable isotopic composition in precipitation can be used as a tracer to reveal the moisture sources [4], retrieve the atmospheric processes [5, 6], and reflect the characteristics of regional climate [7].

Based on the Global Network of Isotopes in Precipitation (GNIP) established by the International Atomic Energy Agency (IAEA) and the World Meteorological Organization, a number of scholars have explored the characteristics of precipitation δ¹⁸O and factors controlling precipitation δ¹⁸O variations [2, 8–10], revealed the original moisture sources of precipitation [11, 12], calculated the contribution ratio of every moisture source to precipitation [9], found significant impacts of climatic events on precipitation δ¹⁸O [13], and utilized models to predict the temporal and spatial distribution of precipitation δ¹⁸O [3, 14–16]. Previous studies have shown that precipitation δ¹⁸O variation for the mid and high-latitude regions is mainly controlled by “temperature effect” and in the low-latitude regions is mainly controlled by “amount effect” [2, 8]. However, not all studies showed “amount effect” in low-latitude regions, but the significance of variation in moisture sources on the precipitation δ¹⁸O variation [17, 18] and different models have represented different simulation accuracy on precipitation δ¹⁸O [3, 14–16]. In China, many studies have explored the temporal and spatial variations of precipitation δ¹⁸O in different geographical zones [15, 16, 19–21], river basins [22–24], and cities [18, 25, 26]. Different moisture sources, which originate from the Arabian Sea, the Bay of Bengal, the South China Sea and the Western Pacific Ocean, and the westerly winds, have remarkable influence on the stable isotopic variations in China [3, 15, 16]. However, few studies have focused on the Pearl River basin (PRB) and adjacent regions, which have the closer distance to the moisture transport sources for being located in the south of China. This may reveal the more complicated patterns of temporal and spatial distribution of precipitation δ¹⁸O and controlling factors. Therefore, this study utilizes precipitation δ¹⁸O data from the GNIP and previous studies to understand the temporal and spatial variations of precipitation δ¹⁸O in the PRB and adjacent regions; explore the controlling factors on the variation of precipitation δ¹⁸O in the PRB and adjacent regions; and establish appropriate spatial models for predicting precipitation δ¹⁸O in the PRB and adjacent regions.

2. Study Area, Data, and Methods

2.1. Study Area

The PRB is situated in the south of China, located from 97.65°E to 117.30°E and 3.68°N to 29.25°N. The Pearl River is the second largest Chinese river in terms of streamflow, with a drainage area of 4.42 10⁵ km² (Pearl River Water Resources Committee (PRWRC), 1991). The mean annual temperature and precipitation range from 14 to 22°C and from 1200 to 2200 mm. In general, the elevation mainly increases from the southeast (southeast delta area) to the northwest (Yunnan–Guizhou Plateau) (Figure 1).

2.2. Data

The precipitation δ¹⁸O data in this study mainly come from the GNIP database (http://isohis.iaea.org) and previous studies by Chinese scholars. The stations in the GNIP utilized in this study are Hong Kong (114.17°E, 22.32°N), Guangzhou (113.32°E, 23.13°N), Guilin (110.08°E, 25.07°N), and Liuzhou (109.40°E, 24.35°N) in the PRB and Haikou (110.35°E, 20.03°N), Guiyang (106.72°E, 26.58°N), Kunming (102.68°E, 25.02°N), and Fuzhou (119.28°E, 26.08°N) in the adjacent regions of the PRB. The data included the monthly temperature, precipitation amount, vapor pressure, and precipitation δ¹⁸O. Data from previous studies came from the stations of Liangfengdong (108.05°E, 25.27°N) [27], Guilin (110.08°E, 25.07°N) [28], and Huanjiang (108.33°E, 24.74°N) [21] which are located in the PRB, and the stations of Ailaoshan (101.03°E, 24.55°N) [21] and Mengzi (103.23°E, 23.23°N) [30] which are located adjacent to the PRB. The spatial distributions of the stations are shown in Figure 1, and the basic information of sampling stations is shown in Table 1.

2.3. Methods

2.3.1. Precipitation δ¹⁸O

All precipitation δ¹⁸O data are on a per mil basis, and δ notation is relative to the Vienna Standard Mean Ocean Water (VSMOW) standard. The δ¹⁸O are calculated as follows: where and represent the precipitation samples and (¹⁸O/¹⁶O) of VSMOW, respectively.

In this study, the sampling period is divided into the summer monsoon period (May to September), nonsummer monsoon period (October to April), and the annual time scale (January to December).

The mean precipitation δ¹⁸O values are the precipitation amount-weighted δ¹⁸O values, which are calculated as follows [29]:where and represent each monthly precipitation δ¹⁸O value and the corresponding precipitation amount, respectively.

2.3.2. OLR and Vertically Integrated Moisture Transport

The monthly mean variations of the outgoing longwave radiation (OLR) and vertically integrated moisture transport were adopted from the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) Reanalysis data (https://www.esrl.noaa.gov/psd/) during 1980 and 2011.

2.3.3. Linear Correlation Analysis

Monthly precipitation δ¹⁸O values for the summer monsoon, nonsummer monsoon periods, and the annual time scale are used to explore the linear correlations with precipitation (mm) and temperature (°C). The correlation is explored for each station alone.

2.3.4. Simulation of the Spatial Distribution of Precipitation δ¹⁸O

(1) Nonlinear Regression Model. The Bowen and Wilkinson (BW) model was used as the nonlinear regression model, which was established by Bowen and Wilkinson [14], and has been successfully applied to estimate the spatial distribution of precipitation δ¹⁸O [16]. The model considers that the isotopic composition of precipitation is affected by the temperature of driving rainout and the regional patterns of the origin and delivery of moisture. Therefore, the latitude and altitude of the precipitation observation stations are substituted for the temperature effect as the geographic parameters to estimate the isotopic composition of precipitation [14]. The modeled precipitation δ¹⁸O has the following form:where δ¹⁸O is an observation of the mean weighted value of the oxygen isotopic composition of precipitation and , , , and are regression parameters. LAT and ALT represent the latitude and altitude, respectively.

(2) Linear Regression Model. Variations in precipitation δ¹⁸O are cocontrolled by different meteorological and geographical factors [3, 21, 31]. Meteorological factors in this study come from the China meteorological data sharing service system and include AP, S, WS, P, RH, VP, and T, which represent atmospheric pressure (hPa), sunshine duration (h), wind speed (m/s), precipitation (mm), relative humidity (%), vapor pressure (hPa), and air temperature (°C), respectively, and altitude (m) (ALT) is used as the geographical factor. In order to compare the factors controlling precipitation δ¹⁸O values among stations under different time scales in the PRB and adjacent regions, linear correlations between the mean weighted precipitation δ¹⁸O values of all stations and their corresponding mean meteorological values and altitude were calculated on the three time periods, while stations of Liangfengdong, Huanjiang, Ailaoshan, and Mengzi were only calculated for the annual time scale.

The modeled precipitation δ¹⁸O has the following form:where δ¹⁸O is an observation of the mean weighted value of the oxygen isotopic composition of precipitation, represent different meteorological and geographical factors that are correlated with the mean weighted precipitation δ¹⁸O value at value < 0.05 or 0.01, and are regression parameters.

3. Results and Discussion

3.1. Correlations of Precipitation δ¹⁸O for Stations with Temperature and Precipitation Amount

The positive correlation between isotopic composition and temperature was called “temperature effect,” while negative correlation between isotopic composition and precipitation amount was called “amount effect” [8, 32]. The correlations between monthly precipitation δ¹⁸O of each station and precipitation amount, temperature, and the slopes δ¹⁸O_p/ and δ¹⁸O_p/ are shown in Table 2. Precipitation δ¹⁸O shows significant negative correlations with temperature in all stations under the annual time scale and in partial stations under the summer monsoon and nonsummer monsoon periods, which shows opposite result to the “temperature effect” of precipitation δ¹⁸O, suggesting the “temperature effect” is rarely found in the summer monsoon areas with low latitudes but mainly occurs in mid-high latitudes, especially near the poles [32], and in nonsummer monsoon areas [11]. Except for Haikou under the annual time scale, precipitation δ¹⁸O has stronger negative correlations with the precipitation amount, especially at the stations of Hong Kong and Kunming, with correlation coefficients of −0.61 and −0.62 and values < 0.01 and 0.01, respectively, which suggests that precipitation δ¹⁸O is significantly influenced by the “amount effect” in the monsoon regions [3, 8]. However, only three stations have correlations with the precipitation amount under the summer monsoon and nonsummer monsoon periods, suggesting the “amount effect” mainly occurs under the annual time scale. Except for Fuzhou under the summer monsoon period, the slopes of δ¹⁸O_p/ for Guiyang and Kunming are significantly higher than other stations in the PRB and adjacent regions (Table 2), and stations of at value < 0.05 or 0.01 for the annual time scale show significant correlations ( value < 0.01) with altitude (Figure 2). This is mainly because the higher altitude areas experience greater temperature drop for wet air-parcels as the decrease in absolute humidity, resulting in greater slopes of δ¹⁸O_p/ [2].

3.2. Precipitation δ¹⁸O Variations at Stations in the PRB and Adjacent Regions

The mean monthly precipitation amount, air temperature, and mean weighted precipitation δ¹⁸O values in the sampling period for each precipitation station of the GNIP are shown in Figure 3. The precipitation δ¹⁸O values present “V-shaped” patterns in all stations, which show a decreasing trend as the temperature and the precipitation amount increase and an increasing trend as the temperature and the precipitation amount decrease (Figure 3). Precipitation δ¹⁸O values at the stations show significant seasonal variation, with more depleted mean precipitation δ¹⁸O values in the summer monsoon period and more enriched δ¹⁸O values in the nonmonsoon period (Figure 3). Our results reveal significantly negative correlation between precipitation δ¹⁸O values for each station and precipitation amount in the annual time scale (Table 2), which shows “amount effect” for each station. However, the most depleted precipitation δ¹⁸O value for each station does not occur in the month with the most precipitation amount. This phenomenon has also been observed in other monsoon regions [18, 21, 33, 34]. Mean weighted precipitation δ¹⁸O values (the annual time scale, the summer monsoon, and the nonsummer monsoon periods) are more enriched at the stations of Guangzhou (−5.83, −6.27, and −4.09%), Haikou (−6.09, −7.80, and −4.12), Hong Kong (−6.59, −7.16, and −4.09), Liuzhou (−6.37, −8.09 and, −3.74), Guilin (−6.12, −7.76, and −3.56), Fuzhou (−6.6, −7.73, and −4.44), and Huanjiang (−6.16) with the lower altitude comparison with the more depleted values at the stations of Guiyang (−8.32, −9.42, and −5.55), Kunming (−10.11, −10.67, and −8.00), Liangfengdong (−7.3), Mengzi (−7.26), and Ailaoshan (−9.22) with higher altitude (Huanjiang, Liangfengdong, Mengzi, and Ailaoshan do not have the mean weighted values for the summer monsoon and nonsummer monsoon periods), suggesting differences in spatial distribution characteristics of precipitation δ¹⁸O. By comparing these values with the mean annual precipitation amount Figure 3, the result shows that the most depleted mean annual precipitation δ¹⁸O values do not occur in the station with the most mean annual precipitation amount. Both of the seasonal and spatial characteristics of variations show that there exist other factors controlling the precipitation δ¹⁸O variations rather than “amount effect.”

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

3.3. Controlling Factors on the Precipitation δ¹⁸O Variation

3.3.1. Effects of Changes in Water Vapor Sources and Vapor Transport on Seasonal Precipitation δ¹⁸O Variation

Water vapor sources and vapor transport influence significantly the precipitation δ¹⁸O variations [3, 15, 16, 18, 31]. The intertropical convergence zone (ITCZ) is a major convergence zone of the troposphere wind field that always closely tracks with monsoon activity and moisture source for precipitation [18]. Outgoing long wave radiation (OLR) is closely related to convection activities, with low OLR values reflecting strong convection activities, conversely, reflecting weak convection activities. The distribution area of low OLR values can accurately reflect the location and intensity changes of the ITCZ to reveal the moisture sources for the summer monsoon precipitation [35]. Therefore, the average OLR values for each month from 1980 to 2011 were calculated (Figure 4). In addition, the average vertically integrated moisture transport for each month from 1980 to 2011 which was calculated by multiplying the zonal and meridional winds by specific humidity from the surface to the 300 hPa level (Figure 5) was used to explore the effects of changes in moisture sources and vapor transport on precipitation δ¹⁸O.

In spring, the ITCZ is mainly distributed near the equator regions in March and April, with a weak intensity in the ITCZ and a long distance from the water vapor sources to the PRB and adjacent regions (Figure 4). The PRB and adjacent regions are dominated by a westerly system, with a small water vapor transport (Figure 5), resulting in enriched precipitation δ¹⁸O (Figure 3) contributed by local evaporation and inland evaporation taken by the westerly system. In May, the ITCZ gradually moves to the Bay of Bengal, with a strengthening in the intensity of the ITCZ that suggests the onset of the Asian summer monsoon (Figure 4). The prevailing wind direction gradually shifts from a westerly system to a south wind (Figure 5). Large amounts of moisture are brought from the Bay of Bengal and strong convection occurs in the water vapor sources. The vapor transportation depletes the δ¹⁸O in water vapor, which causes the shape decrease in the precipitation δ¹⁸O for May in comparison to April (Figure 3).

In summer, the convection center of the ITCZ moves further to the north with the rapidly strengthening intensity in the Bay of Bengal, the South China Sea, and the Arabian Sea in June (Figure 4), and the prevailing wind direction is dominated by a southwest wind (Figure 5). As large amounts of moisture are brought to the PRB and adjacent regions by the southwest wind, a stronger convection occurs in the water vapor sources and vapor transportation, resulting in a continuous decrease in the precipitation δ¹⁸O in the PRB and adjacent regions. In July, the prevailing wind direction and water vapor transport do not change too much compared with that in June, but the strong intensity convention for the ITCZ is enlarged further in the Bay of Bengal and the Indo-China Peninsula, depleting the δ¹⁸O in the water vapor and continuously decreasing the precipitation δ¹⁸O in the PRB and adjacent regions. In August, by comparison with July, the prevailing wind direction in the PRB and adjacent regions becomes jointly controlled by southwest and southeast winds (Figure 5). The scope of the convection in the ITCZ is narrowed in the Bay of Bengal, but the intensity of the convection increases in the western Pacific Ocean (Figure 4), which increases the distance for water vapor transport to the PRB and adjacent regions. These changes contribute to the depleted precipitation δ¹⁸O values in the PRB and adjacent regions (Figure 3).

In autumn, the prevailing wind direction does not change much in September as compared with August (Figure 5). However, the convective centers of the ITCZ in the Bay of Bengal and the South China Sea move southward (Figure 4), which increases the distance for water vapor transport to the PRB and adjacent regions. As a result, the stations experience decreasing precipitation δ¹⁸O values, with Guilin and Liuzhou obviously reaching their most depleted precipitation δ¹⁸O values during this month. However, Haikou and Guangzhou do not reach their most depleted precipitation δ¹⁸O values in September but have significantly higher values than the most depleted values obtained in August (Figure 3). As shown in Figure 5, the water vapor sources originate from the Bay of Bengal and South China Sea and also originate from the adjacent seas in September. Moreover, Haikou and Guangzhou are all coastal stations in the south of PRB, which results in higher precipitation δ¹⁸O values than in August due to the relatively shorter distance for water vapor transport. From October to November, the prevailing wind direction changes from southwest and southeast to west, and the ITCZ moves further southward to the vicinity of the equator, suggesting that the summer monsoon has disappeared. The water vapor sources mainly originate from the inland [3], resulting in enriched precipitation δ¹⁸O values in the PRB and adjacent regions (Figure 3).

In winter, water vapor mainly originates from local evaporation and inland evaporation due to the westerly system (Figure 5), resulting in enriched precipitation δ¹⁸O values in the PRB and adjacent regions (Figure 3).

3.3.2. Local Meteorological and Geographical Factors Controlling Spatial Precipitation δ¹⁸O Variation

In order to explore the factors controlling spatial distribution of precipitation δ¹⁸O in the PRB and adjacent regions, we compared the correlations between mean weighted precipitation δ¹⁸O values and meteorological and geographical factors among stations, showing that mean values for atmospheric pressure, temperature, water vapor pressure, precipitation amount, and altitude show significant correlations with mean weighted precipitation δ¹⁸O values in the annual time scale of the PRB and adjacent regions, with correlation coefficients of 0.929, 0.787, 0.786, 0.589, and −0.908, respectively, with values < 0.01 for atmospheric pressure, temperature, water vapor pressure, altitude, and 0.05 for precipitation amount (Figures 6(a), 6(b), 6(c), 6(e), and 6(d)). This suggests that they are the main factors controlling spatial precipitation variation. The slope of δ¹⁸O_p/ and δ¹⁸O_p/ is 0.34/°C (Figure 6(b)) and 1.7/100 mm (Figure 6(d)), respectively. The slope of δ¹⁸O_p/ is higher than the northeast regions of China (0.27/°C) and slightly lower than the northwest regions of China (0.37/°C) [21]. The slope of δ¹⁸O_p/ALT is −0.15/100 m (Figure 6(e)), which indicates the “altitude effect” on the weighted mean annual precipitation δ¹⁸O values, which is lower than that found for the Tibet Plateau (−0.3/100 m) by Liu et al. [21] and for the globe (−0.28/100 m) by Poage and Chamberlain [36]. However, this is similar to the whole country scale found for China (−0.15/100 m and −0.13/100 m) by Liu et al. [16] and Liu et al. [21], respectively.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

In the summer monsoon period, the controlling factors for spatial precipitation δ¹⁸O variation is atmospheric pressure, temperature, vapor pressure, and altitude, exhibiting correlation coefficients of 0.930, 0.889, 0.913, and −0.921, respectively, with all the values < 0.01 (Figures 6(f), 6(g), 6(h), and 6(i)). The slope of δ¹⁸O_p/ and δ¹⁸O_p/ALT is 0.35/°C (Figure 6(g)) and −0.17/100 m (Figure 6(i)), respectively, which is similar to that for the annual time scale.

In the nonsummer monsoon period, the controlling factors for the precipitation δ¹⁸O values are sunshine duration, atmospheric pressure, and altitude, with correlation coefficients of −0.776, 0.959, and −0.957, respectively, explaining 60.2%, 91.9%, and 91.6% of the variance, respectively, with values < 0.05, 0.01, and 0.01 (Figures 6(j), 6(k), and 6(l)). The δ¹⁸O_p/ALT is −0.20/100 m (Figure 6(l)), suggesting that the “altitude effect” for the nonsummer monsoon period is the most significant among the three time periods.

However, mean temperature shows significantly positive correlation with mean weighted precipitation δ¹⁸O in the annual time scale and the summer monsoon period (Figures 6(b) and 6(g)). Meanwhile, mean precipitation also shows significantly positive correlation with mean weighted precipitation δ¹⁸O values (Figure 6(d)). Moreover, both temperature and precipitation show significantly negative correlations with altitude (Figure 7). This indicates that the correlations of precipitation and temperature may reflect the result of the “altitude effect,” suggesting that altitude should be the more significant factor for spatial precipitation δ¹⁸O variation in the PRB and adjacent regions.

(a)

(b)

(c)

3.4. Simulation of Precipitation δ¹⁸O in the PRB and Adjacent Regions

3.4.1. Nonlinear Regression Model and Linear Regression Model

The Bowen and Wilkinson model has been applied widely for simulating the precipitation δ¹⁸O in monsoon climatic zones [3, 10, 16]. No significant interaction effect of latitude and altitude has been found since the stations are mostly situated below 200 m [14]. The interaction effect of latitude and altitude occurs when stations are both below 200 m and above 200 m [16], and the elevation of our research area ranges from 0 to 2852 m. Therefore, we adopt a nonlinear fit to consider the correlation between latitude (LAT) and altitude (ALT) to establish the mean weighted precipitation δ¹⁸O patterns under the three time scales: the annual time scale, summer monsoon, and nonsummer monsoon periods. The model equations are as follows:where (5), (6), and (7) represent the mean weighted precipitation δ¹⁸O patterns for the annual time scale, summer monsoon, and nonsummer monsoon periods, respectively.

According to the above results (Figure 6), we establish linear regression models using those meteorological and geographical variables for modeling in the three time periods. The resulting models are as follows:where (8), (9), and (10) represent the mean weighted precipitation δ¹⁸O patterns for the annual time scale, summer monsoon, and nonsummer monsoon periods, respectively. The parameters, S, VP, T, AP, P, and ALT represent sunshine duration (h), vapor pressure (hPa), air temperature (°C), atmospheric pressure (hPa), precipitation amount (mm), and altitude (m), respectively.

By comparing the results of the Bowen and Wilkinson model with other researches, the regression coefficients of the second-order term (−0.062) (see (5)) are significantly more negative than that modeled on a global scale (−0.0051) [14] or modeled for the United States (−0.0057) [37] and China (−0.0073) [16]. This may be caused by the smallest area of our study. And the PRB and adjacent regions are significantly influenced by both the southeast and southwest monsoons, and precipitation δ¹⁸O depleted as moisture is transferred from the oceans to the PRB and adjacent regions. Moreover, the Bowen and Wilkinson model estimates the best precipitation δ¹⁸O patterns for nonsummer monsoon period (Figure 8). Zhao et al. [3] found that the Bowen and Wilkinson model could estimate precipitation δ¹⁸O well only in winter for China. This suggests that there are differences in modeling the precipitation δ¹⁸O by the Bowen and Wilkinson model for different regional scales and time scales in the monsoon climatic regions.

(a)

(b)

(c)

(d)

(e)

(f)

By comparing the results of the Bowen and Wilkinson model with the linear regression model, the results show that the linear regression model also estimates the best precipitation δ¹⁸O patterns for the nonsummer monsoon period (Figure 8). The amount of variance explained by the linear regression model is higher (93.5%, 92.8%, and 99.0%, resp.) than those (85.8%, 90%, and 92.4%, resp.) (Figure 8) by the Bowen and Wilkinson model for the annual time scale, summer monsoon, and nonsummer monsoon periods, respectively. Moreover, the mean errors explained by the linear regression model are lower (0.27 and 0.24, resp.) than those (0.48 and 0.35, resp.) by the Bowen and Wilkinson model for the annual time scale and the summer monsoon period. There is a higher mean error (0.41) in the linear regression model than that in the Bowen and Wilkinson model (0.32) for the nonsummer monsoon period, but the linear regression model explains more variance with a lower standard deviation of mean error (99.0 and 0.19, resp.) than does the Bowen and Wilkinson model (92.4 and 0.22, resp.). The results suggest that precipitation δ¹⁸O can be estimated more accurately by the linear regression model than by the Bowen and Wilkinson model in the PRB and adjacent regions.

3.5. Spatial Distribution of Precipitation δ¹⁸O in the PRB

Spatial distribution of precipitation δ¹⁸O values from 1980 to 2011 for the annual time scale, summer monsoon, and nonsummer monsoon periods of the PRB are presented based on (8), (9), and (10), respectively. The data utilized come from the meteorological data network of China (Figure 9(a)). Precipitation δ¹⁸O values are gradually more depleted (Figure 9) from the south and east to the north and northwest, especially in the annual time scale and the summer monsoon period, which are mainly caused by “altitude effect.” Precipitation δ¹⁸O values are more depleted in the summer monsoon period and more enriched in the nonsummer monsoon period (Figures 9(c) and 9(d)). However, spatial distribution of precipitation δ¹⁸O reflect only the results of the linear regression model based on limited quantities and years of observation stations. As a result, more of the long-time observation stations should be established for more accurate estimation of precipitation δ¹⁸O variations in the monsoon climatic regions.

(a)

(b)

(c)

(d)

Figure 9

Spatial distribution of mean weighted precipitation δ¹⁸O values over Pearl River basin of China based on the Kriging method from 1980 to 2011. Note: (a) represents meteorological stations in the Pearl River basin; (b) represents spatial distribution of the mean weighted annual precipitation δ¹⁸O values by (8); (c) represents spatial distribution of mean weighted precipitation δ¹⁸O values for the summer monsoon period by (9); (d) represents spatial distribution of mean weighted precipitation δ¹⁸O values for the nonsummer monsoon period by (10).

4. Conclusions

Based on the precipitation δ¹⁸O values from the datasets of the Global Network of Isotopes in Precipitation (GNIP), the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) Reanalysis data, and previous researches, we explored the temporal and spatial variations of precipitation δ¹⁸O in the PRB and adjacent regions. The principal conclusions are as follows:(1)There was no “temperature effect” for the precipitation δ¹⁸O for each station in the PRB and adjacent regions. “Amount effect” has been found at all stations (except for Haikou) for the annual time scale, while only three stations show the “amount effect” in the summer monsoon and nonsummer monsoon periods.(2)Temporal variations show that the most depleted mean monthly precipitation δ¹⁸O value does not occur in the month with the most mean monthly precipitation amount, which should be correlated with water vapor sources, the distance of water vapor transport, the changes in location, and intensity of the intertropical convergence zone (ITCZ). Spatial variations show that the most depleted mean precipitation δ¹⁸O values does not occur in the station with the most mean precipitation amount for the annual time scale. Meanwhile, “altitude effect” has been significantly found among stations and precipitation values among stations show close correlation with local meteorological conditions, which can account for the characteristics in the spatial variations. This indicates that water vapor sources and changes in location and intensity of the ITCZ, “altitude effect” as well as local meteorological conditions should be taken significantly into consideration when interpreting the temporal and spatial variations of precipitation δ¹⁸O rather than “amount effect” in the PRB and adjacent regions.(3)We established linear regression models combining multiple meteorological variables for estimating the mean weighted precipitation δ¹⁸O in the PRB for the annual time scale, summer monsoon, and nonsummer monsoon periods. The amounts of variances explained by the linear regression model are all more than 92%, which are higher than those explained by the Bowen and Wilkinson model, suggesting that the linear regression models can estimate more accurately than the Bowen and Wilkinson model in the PRB and adjacent regions.(4)This research can provide useful information for rebuilding temporal and spatial precipitation δ¹⁸O in the monsoon climatic regions. Meanwhile, more of the long-time observation stations should be established for more accurate estimation of precipitation δ¹⁸O variations in the monsoon climatic regions.

Conflicts of Interest

The authors declare that there are no conflicts of interest.

Acknowledgments

This research was supported by the National Natural Science Foundation of China (41371058, 4161101254, and 41561144012).

References

K. Rozanski, L. Araguas-Araguas, and R. Gonfiantini, “Isotopic patterns in modern global precipitation,” in Climate Change in Continental Isotopic Records, P. K. Swart, K. C. Lohmann, J. McKenzie, and S. Savin, Eds., American Geophysical Union, pp. 1–36, Washington, DC, USA, 1993.
View at: Google Scholar
X. Zhang, J. Liu, L. Tian, Y. He, and T. Yao, “Variations of δ¹⁸O in precipitation along vapor transport paths,” Advances in Atmospheric Sciences, vol. 21, no. 4, pp. 562–572, 2004.
View at: Publisher Site | Google Scholar
L. Zhao, H. Xiao, M. Zhou et al., “Factors controlling spatial and seasonal distributions of precipitation δ¹⁸O in China,” Hydrological Processes, vol. 26, no. 1, pp. 143–152, 2012.
View at: Publisher Site | Google Scholar
N. Kurita, N. Yoshida, G. Inoue, and E. A. Chayanova, “Modern isotope climatology of Russia: A first assessment,” Journal of Geophysical Research: Atmospheres, vol. 109, article D03102, 2004.
View at: Google Scholar
D. Brown, J. Worden, and D. Noone, “Comparison of atmospheric hydrology over convective continental regions using water vapor isotope measurements from space,” Journal of Geophysical Research: Atmospheres, vol. 113, no. 15, Article ID D15124, 2008.
View at: Publisher Site | Google Scholar
V. I. Zakharov, R. Imasu, K. G. Gribanov, G. Hoffmann, and J. Jouzel, “Latitudinal distribution of the deuterium to hydrogen ratio in the atmospheric water vapor retrieved from IMG/ADEOS data,” Geophysical Research Letters, vol. 31, no. 12, Article ID L12104, 2004.
View at: Publisher Site | Google Scholar
I. Clark and P. Fritz, Environmental Isotopes in Hydrogeology, Lewis Publishers, New York, NY, USA, 1997.
L. Araguás-Araguás, K. Froehlich, and K. Rozanski, “Stable isotope composition of precipitation over southeast Asia,” Journal of Geophysical Research: Atmospheres, vol. 103, no. D22, Article ID 98JD02582, pp. 28721–28742, 1998.
View at: Publisher Site | Google Scholar
S. Sengupta and A. Sarkar, “Stable isotope evidence of dual (Arabian Sea and Bay of Bengal) vapour sources in monsoonal precipitation over north India,” Earth and Planetary Science Letters, vol. 250, no. 3-4, pp. 511–521, 2006.
View at: Publisher Site | Google Scholar
X. Liu, Z. Rao, X. Zhang, W. Huang, J. Chen, and F. Chen, “Variations in the oxygen isotopic composition of precipitation in the Tianshan Mountains region and their significance for the Westerly circulation,” Journal of Geographical Sciences, vol. 25, no. 7, pp. 801–816, 2015.
View at: Publisher Site | Google Scholar
W. Yu, T. Yao, S. Lewis et al., “Stable oxygen isotope differences between the areas to the north and south of Qinling Mountains in China reveal different moisture sources,” International Journal of Climatology, vol. 34, no. 6, pp. 1760–1772, 2014.
View at: Publisher Site | Google Scholar
B.-L. Cui and X.-Y. Li, “Stable isotopes reveal sources of precipitation in the Qinghai Lake Basin of the northeastern Tibetan Plateau,” Science of the Total Environment, vol. 527-528, pp. 26–37, 2015.
View at: Publisher Site | Google Scholar
M. Tan, “Circulation effect: response of precipitation to the ENSO cycle in monsoon regions of China,” Climate Dynamics, vol. 42, no. 3-4, pp. 1067–1077, 2014.
View at: Publisher Site | Google Scholar
G. J. Bowen and B. Wilkinson, “Spatial distribution of δ¹⁸O in meteoric precipitation,” Geology, vol. 30, no. 4, pp. 315–318, 2002.
View at: Publisher Site | Google Scholar
J. Liu, X. Song, G. Yuan et al., “Stable isotopes of summer monsoonal precipitation in southern China and the moisture sources evidence from δ¹⁸O signature,” Journal of Geographical Sciences, vol. 18, no. 2, pp. 155–165, 2008.
View at: Publisher Site | Google Scholar
Z. Liu, L. Tian, X. Chai, and T. Yao, “A model-based determination of spatial variation of precipitation δ¹⁸O over China,” Chemical Geology, vol. 249, no. 1-2, pp. 203–212, 2008.
View at: Publisher Site | Google Scholar
T.-R. Peng, C.-H. Wang, C.-C. Huang, L.-Y. Fei, C.-T. A. Chen, and J.-L. Hwong, “Stable isotopic characteristic of Taiwan's precipitation: A case study of western Pacific monsoon region,” Earth and Planetary Science Letters, vol. 289, no. 3-4, pp. 357–366, 2010.
View at: Publisher Site | Google Scholar
Y. Tang, H. Pang, W. Zhang, Y. Li, S. Wu, and S. Hou, “Effects of changes in moisture source and the upstream rainout on stable isotopes in precipitation -a case study in Nanjing, eastern China,” Hydrology and Earth System Sciences, vol. 19, no. 10, pp. 4293–4306, 2015.
View at: Publisher Site | Google Scholar
L. Tian, T. Yao, K. MacClune et al., “Stable isotopic variations in west China: A consideration of moisture sources,” Journal of Geophysical Research: Atmospheres, vol. 112, no. 10, Article ID D10112, 2007.
View at: Publisher Site | Google Scholar
X. P. Zhang, J. M. Liu, W. Z. Sun, Y. M. Huang, and J. M. Zhang, “Relations between oxygen stable isotopic ratios in precipitation and relevant meteorological factors in southwest China,” Science China Earth Sciences, vol. 50, no. 4, pp. 571–581, 2007.
View at: Publisher Site | Google Scholar
J. Liu, X. Song, G. Yuan, X. Sun, and L. Yang, “Stable isotopic compositions of precipitation in China,” Tellus B: Chemical and Physical Meteorology, vol. 66, no. 1, Article ID 22567, 2014.
View at: Publisher Site | Google Scholar
J. Wu, Y. Ding, B. Ye, Q. Yang, X. Zhang, and J. Wang, “Spatio-temporal variation of stable isotopes in precipitation in the Heihe River Basin, Northwestern China,” Environmental Earth Sciences, vol. 61, no. 6, pp. 1123–1134, 2010.
View at: Publisher Site | Google Scholar
X. F. Li, M. J. Zhang, S. J. Wang, X. N. Ma, and F. Li, “Spatial and temporal variations of hydrogen and oxygen isotopes in precipitation in the yellow river basin and its environmental significance,” Acta Geologica Sinica, vol. 87, no. 2, pp. 269–277, 2013 (Chinese).
View at: Google Scholar
S. G. Pang, S. K. Zhao, R. Wen, and Z. F. Liu, “Spatial and temporal variation of stable isotopes in precipitation in the Haihe River basin,” Chinese Science Bulletin, vol. 60, no. 13, pp. 1218–1226, 2015 (Chinese).
View at: Google Scholar
J. B. Xue, W. Zhong, and Y. J. Zhao, “Stable oxygen isotope in precipitation in Guangzhou in relation to the meteorological factors and the monsoon activity,” Journal of Glaciology and Geocryology, vol. 30, no. 5, pp. 761–768, 2008 (Chinese).
View at: Google Scholar
Y. M. Huang, X. P. Zhang, X. Sun, Y. B. Huang, and N. Q. Wei, “Seasonal variations of stable isotope in precipitation and atmospheric water vapor and their relationship with moisture transportation in Changsha city,” Scientia Geographica Sinica, vol. 35, no. 4, pp. 498–506, 2015 (Chinese).
View at: Google Scholar
W. J. Luo and S. J. Wang, “Signal transmission of in precipitation, soil water and dropping water of Liang Feng cave and its implications,” Chinese Science Bulletin, vol. 53, no. 17, pp. 2071–2076, 2008 (Chinese).
View at: Google Scholar
X. Wu, X. Y. Zhu, M. L. Zhang, X. Bai, and B. Y. Zhang, “High-resolution stable isotope record of atmospheric precipitation,” Resources and Environment in the Yangtze Basin, vol. 22, no. 2, pp. 182–188, 2013 (Chinese).
View at: Google Scholar
T. B. Coplen, “Reporting of stable carbon, hydrogen and oxygen abundances,” IAEA Tecdoc 825, International Atomic Energy Agency, Vienna, Austria, 1995.
View at: Google Scholar
G. Li, X. P. Zhang, H. W. Wu, J. M. Zhang, N. Q. Wei, and H. Huang, “Stable oxygen isotope in precipitation in relation to the meteorological factors and the moisture sources in Yunnan,” Journal of Natural Resources, vol. 29, no. 6, pp. 1043–1052, 2014 (Chinese).
View at: Google Scholar
M. Vuille, M. Werner, R. S. Bradley, and F. Keimig, “Stable isotopes in precipitation in the Asian monsoon region,” Journal of Geophysical Research: Atmospheres, vol. 110, no. 23, Article ID D23108, pp. 1–15, 2005.
View at: Publisher Site | Google Scholar
W. Dansgaard, “Stable isotopes in precipitation,” Tellus, vol. 16, no. 4, pp. 436–468, 1964.
View at: Publisher Site | Google Scholar
S. K. Bhattacharya, K. Froehlich, P. K. Aggarwal, and K. M. Kulkarni, “Isotopic variation in indian monsoon precipitation: records from bombay and new delhi,” Geophysical Research Letters, vol. 30, no. 24, pp. 2285–2289, 2003.
View at: Publisher Site | Google Scholar
S. F. M. Breitenbach, J. F. Adkins, H. Meyer, N. Marwan, K. K. Kumar, and G. H. Haug, “Strong influence of water vapor source dynamics on stable isotopes in precipitation observed in Southern Meghalaya, NE India,” Earth and Planetary Science Letters, vol. 292, no. 1-2, pp. 212–220, 2010.
View at: Publisher Site | Google Scholar
K. D. Prasad and S. D. Bansod, “Interannual variations of outgoing longwave radiation and Indian summer monsoon rainfall,” International Journal of Climatology, vol. 20, no. 15, pp. 1955–1964, 2000.
View at: Publisher Site | Google Scholar
M. A. Poage and C. P. Chamberlain, “Empirical relationships between elevation and the stable isotope composition of precipitation and surface waters: Considerations for studies of paleoelevation change,” American Journal of Science, vol. 301, no. 1, pp. 1–15, 2001.
View at: Publisher Site | Google Scholar
A. Dutton, B. H. Wilkinson, J. M. Welker, G. J. Bowen, and K. C. Lohmann, “Spatial distribution and seasonal variation in ¹⁸O/¹⁶O of modern precipitation and river water across the conterminous USA,” Hydrological Processes, vol. 19, no. 20, pp. 4121–4146, 2005.
View at: Publisher Site | Google Scholar

Copyright

Copyright © 2018 Yunfeng Ruan et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies

Views

1254

Downloads

1171

Citations

Advances in Meteorology

Temporal and Spatial Variations of Precipitation δ18O and Controlling Factors on the Pearl River Basin and Adjacent Regions

Abstract

1. Introduction

2. Study Area, Data, and Methods

2.1. Study Area

2.2. Data

2.3. Methods

2.3.1. Precipitation δ18O

2.3.2. OLR and Vertically Integrated Moisture Transport

2.3.3. Linear Correlation Analysis

2.3.4. Simulation of the Spatial Distribution of Precipitation δ18O

3. Results and Discussion

3.1. Correlations of Precipitation δ18O for Stations with Temperature and Precipitation Amount

3.2. Precipitation δ18O Variations at Stations in the PRB and Adjacent Regions

3.3. Controlling Factors on the Precipitation δ18O Variation

3.3.1. Effects of Changes in Water Vapor Sources and Vapor Transport on Seasonal Precipitation δ18O Variation

3.3.2. Local Meteorological and Geographical Factors Controlling Spatial Precipitation δ18O Variation

3.4. Simulation of Precipitation δ18O in the PRB and Adjacent Regions

3.4.1. Nonlinear Regression Model and Linear Regression Model

3.5. Spatial Distribution of Precipitation δ18O in the PRB

4. Conclusions

Conflicts of Interest

Acknowledgments

References

Copyright

Temporal and Spatial Variations of Precipitation δ¹⁸O and Controlling Factors on the Pearl River Basin and Adjacent Regions

2.3.1. Precipitation δ¹⁸O

2.3.4. Simulation of the Spatial Distribution of Precipitation δ¹⁸O

3.1. Correlations of Precipitation δ¹⁸O for Stations with Temperature and Precipitation Amount

3.2. Precipitation δ¹⁸O Variations at Stations in the PRB and Adjacent Regions

3.3. Controlling Factors on the Precipitation δ¹⁸O Variation

3.3.1. Effects of Changes in Water Vapor Sources and Vapor Transport on Seasonal Precipitation δ¹⁸O Variation

3.3.2. Local Meteorological and Geographical Factors Controlling Spatial Precipitation δ¹⁸O Variation

3.4. Simulation of Precipitation δ¹⁸O in the PRB and Adjacent Regions

3.5. Spatial Distribution of Precipitation δ¹⁸O in the PRB