1. Introduction
Large-scale spatial distribution maps and models of stable precipitation isotopes, δ18O and δ2H, known as precipitation isoscapes, provide a geographic perspective to aid the investigation of atmospheric and hydrological processes (West et al. 2010; Lachniet et al. 2016). The spatial information provided by precipitation isoscapes is not only of great use in understanding the transport, mixing, and phase transition of moisture in the modern atmosphere (Vachon et al. 2010; Bowen and Good 2015; Allen et al. 2018; Falster et al. 2021), but also provides a spatial basis for interpreting isotope signatures in paleoclimate records (Johnson and Ingram 2004; Dayem et al. 2010; Maher and Thompson 2012; Hatvani et al. 2017). Precipitation isoscapes also provide a key input to hydrological, ecological, and forensics applications at both local to global scales (e.g., Bowen et al. 2005; Sánchez-Murillo and Birkel 2016; Hollins et al. 2018; Hobson et al. 2018) by providing a geographically linked signature for δ18O and δ2H values in the hydrological and ecological cycles.
In China, the precipitation isoscape is impacted by varied moisture regimes for the monsoon- and westerlies-dominated regions (Yao et al. 2013; Liu et al. 2014; Chiang et al. 2020; Chen et al. 2019; J. Chen et al. 2021). In the southeastern part of China, the precipitation is controlled by the East Asian monsoon and the Indian monsoon, and marine moisture near its evaporation source contributes to precipitation relatively enriched in 18O and 2H in the coastal region (Cai and Tian 2016; Zhou et al. 2019; Zhang et al. 2021), although some extreme events like typhoons may also impact the precipitation isoscape (F. Chen et al. 2021). In the northwestern region of China, where midlatitude westerlies dominate all year round, long-distance moisture advection may lead to heavy isotope depleted precipitation due to rainout, but the inland arid climate may strongly enhance the below-cloud raindrop evaporation, leading to precipitation enriched in heavy isotopes in summer (Pang et al. 2011; Wang et al. 2016a; S. Wang et al. 2017; Li et al. 2019). In addition, the Qinghai–Tibet Plateau, with mean elevation of about 4000 m above sea level, presents an elevation-dependent precipitation isoscape jointly influenced by the monsoon and westerlies moisture (Yao et al. 2013; Tian et al. 2007; Gao et al. 2018; Shi et al. 2020a).
As a key reference in examining the atmospheric processes using isotope methods (West et al. 2010), a high-resolution monthly precipitation isoscape reflecting the long-term climatology is necessary in China (Zhang and Wang 2016). Applications of δ18O and δ2H to trace meteorological input into groundwater and water supplies, agricultural water use, biosecurity, food provenance, and forensics need better than an annual resolution because of the seasonality of processes like plant growth, recharge, animal migration, and irrigation (Bowen and Good 2015; Hobson and Wassenaar 2019; Bowen et al. 2022). Availability of such isoscapes is an underpinning resource that enables stable isotope methods to be used by anyone who do not have the interest or resource to invest in local rainfall sampling (Liu et al. 2008; P. Zhao et al. 2019). In addition, isotope-enabled GCMs are increasingly being used to improve the representation of vapor transport, water cycling, and convective process in climate modeling, and the capacity to predict seasonality of precipitation isotopes is one measure of their success (Sturm et al. 2010; Hu et al. 2018). The high-resolution isoscape represents a resource to assist in evaluating the performance of isotope-enabled GCMs (e.g., the Stable Water Isotope Intercomparison Group Project) for a large and climatologically diverse part of the globe (Yang et al. 2017; Peng et al. 2020).
The existing high-resolution global annual and monthly precipitation isoscapes, such as the Online Isotopes in Precipitation Calculator (OIPC; Bowen and Wilkinson 2002) and the Regionalized Cluster-based Water Isotope Prediction (RCWIP; Terzer et al. 2013), are highly dependent on the Global Network of Isotopes in Precipitation (GNIP). With a lack of observations in western China these existing global precipitation isoscape products, which use only about 30 Chinese sites (Liu et al. 2014; Zhang and Wang 2016), are poor at predicting precipitation isotopes in the remote areas of western China (Shi et al. 2020a,b). So far there is no nationwide monthly precipitation isoscape to represent the obvious seasonality in China, although several precipitation isoscapes on an annual scale were created using geographic information systems (Zhang and Wang 2016). For example, Liu et al. (2008) determined an annual mean precipitation isoscape of δ18O values using observations at 55 sites, with latitude and elevation selected as the predictor variables. Similar annual mean precipitation isoscapes were later developed (Li et al. 2011; Yang et al. 2014; P. Zhao et al. 2019), with site numbers ranging from 62 (Li et al. 2011) to 126 (P. Zhao et al. 2019). More recently the growth in studies producing independent precipitation isotope measurements across China (Zhang and Wang 2016) has provided the increased data density to enable an updated and improved nationwide monthly precipitation isoscape to be developed.
China encompasses a wide geographical and climatological range, and therefore it is not effective to use a universal regression equation to describe the key regimes for different climate zones (Araguás-Araguás et al. 1998; Kong et al. 2019). Regionalized regression may be practical, but the sharp discontinuity caused by using different regression equations for different regions should be avoided as rarely represents the actual transition between different climate zones (Hollins et al. 2018). A regionalized fuzzy clustering method (Terzer et al. 2013; Terzer-Wassmuth et al. 2021) was previously used in creating a global precipitation isoscape, where fractional membership to each cluster was considered and the cluster weighed values were used, and the performance seems much better than previous universal regressions. As a state-of-the-art option, the fuzzy clustering is practical for a large spatial domain with a number of sampling sites (Terzer-Wassmuth et al. 2021), and can be applied to regions with more sparse data (Shi et al. 2020a).
In this study, we compiled an integrated database of δ18O and δ2H in precipitation across China, and then produced an updated monthly precipitation isoscape product (named C-Isoscape for short) using a fuzzy clustering approach (Terzer et al. 2013; Terzer-Wassmuth et al. 2021). In so doing, unrealistic transitions between zones with spatially discontinuous isotopic distribution can be avoided. The updated monthly precipitation isoscape products will provide a new precipitation isotope end-member resource for further application in water isotope-related studies across China.
2. Materials and methods
a. Isotope data compilation
We compiled the δ18O and δ2H data in precipitation at 223 sampling sites across China (Fig. 1; see also Table S1 in the online supplemental material), published in articles and/or online data repositories, during past decades (Zhang 1989; Xu et al. 2006; Tian et al. 2007, 2015; Zhao et al. 2009, 2011, 2018; L. Zhao et al. 2019; Chen et al. 2010, 2017; Liu et al. 2010, 2014; Peng et al. 2010; Pang et al. 2011, 2015; Xie et al. 2011; Wang et al. 2012, 2015, 2016c; X. Wang et al. 2017; Wang et al. 2018; L. Wang et al. 2019; S. Wang et al. 2019; Wang et al. 2020; Luo et al. 2013; Ren et al. 2013; Yao et al. 2013; Jin et al. 2015, 2019; Cui et al. 2017; Tang et al. 2017; Zhao 2017, 2018; Chen and Li 2018; IAEA 2020; Jia et al. 2018; Rao et al. 2018; Sun et al. 2018, 2019, 2020; Yao 2018; Yang et al. 2018; Zhou and Li 2018; Jiao et al. 2019; Qiu et al. 2019; Wu et al. 2019; G. Zhu et al. 2019; X. Zhu et al. 2019; Zhu et al. 2020; Shi et al. 2020; Shi et al. 2021a; Zhan et al. 2020). The site elevations range from 2 to 5170 m above sea level. In addition, to improve the estimation for the boundary regions, 48 sites from neighboring countries are also included (Fig. 1 and Table S1) (Wushiki 1977; Kurita et al. 2004; Yamanaka et al. 2007; Kumar et al. 2010; IAEA 2020; Juhlke et al. 2019; Burnik Šturm et al. 2017; Malygina et al. 2019; Adhikari et al. 2020; Acharya et al. 2020). For most sites used in this study, across or outside China, the sampling period is longer than one year (see Table S2). The isotope ratios are expressed as a δ-notation, that is, δ = (Rsample/Rstandard − 1) × 1000‰, where R is the ratio of the heavy isotope (18O or 2H) to light isotope (16O or 1H) in a sample relative to a standard called the Vienna Standard Mean Ocean Water (VSMOW). For most sites, the uncertainty of δ18O and δ2H measurements is better than 0.3‰ and 1‰, respectively.
Map showing the sampling period for each station for precipitation isotopes analyzed in this study.
Citation: Journal of Climate 35, 11; 10.1175/JCLI-D-21-0451.1
The δ18O and δ2H values were derived for the 12 months of the year, for each station (where possible), by calculating a precipitation amount weighted average value from observations undertaken on monthly, daily, or event-based frequencies (Table S2). For the references without precipitation amount data, we used gridded daily precipitation from the NOAA Climate Prediction Center (CPC) Global Unified Gauge-Based Analysis of Daily Precipitation (Chen et al. 2008), which is similar to Putman et al. (2019). If the 12 monthly averages, weighted by precipitation amount, were shown as numerical values in the original publications, the weighted mean data were directly compiled into our database; in these cases, the measured data for each monthly, daily, or event-based samples were usually not available. To ensure the accuracy of the isotope data, we did not extract any data from the published figures, which means that the sites without original or weighted mean numerical values in publications are not included in the 223 sampling sites. A total of 2609 (2155 in China) monthly average data points for δ18O and 2316 (1852 in China) for δ2H were compiled (Figs. S1 and S2).
Since this is a monthly climatology, there is no information about the interannual variability or longer-term trends, which is the same as for the OIPC and RCWIP. According to a global assessment, at the Hong Kong site, which has the longest measurement period in China, there is no interannual trend (Terzer et al. 2013). The rest of the sampling stations generally have less than 20 years of observation, insufficient to accurately assess a long-term trend. As simulated using isotope-equipped general climate models across China (Yang et al. 2017), there was a slightly increasing trend in δ18O values during 1979–2007 (ranging from 0.09‰ to 0.23‰ per decade for each subregion), but the trends were not statistically significant for most models and were generally close to the analytical uncertainty for samples.
b. Climate regionalization using fuzzy clustering
Here we applied a fuzzy clustering method allowing an individual target location to be partially classified into more than one cluster (Terzer et al. 2013). At each location (climate grid or observation site), the sum of membership for all clusters theoretically equals to 1, and the membership for each cluster ranges from 0 to 1. See appendix for the algorithm of fuzzy clustering membership (Terzer et al. 2013; Kaufman and Rousseeuw 1990). We used the function “fanny” (fuzzy analysis clustering) of package “cluster” version 2.1.2 in R programming version 4.1.1. The fuzzifier constant (i.e., the sharpness of transition between two fuzzy clusters) was set as 1.5. Euclidean distances are used as the metric for calculating dissimilarities between observations, and the measurements are standardized by subtracting the variable’s mean and dividing by the variable’s mean absolute deviation before calculating the dissimilarities. The maximal number of iterations were set as 500, and default tolerance for convergence was 0.0001.
We used 27 variables in climate regionalization: 3 spatial (latitude, longitude, and elevation), 12 monthly air temperature, and 12 monthly precipitation amount parameters at 35 928 grids in China. The long-term mean air temperature and precipitation amount for each of the 12 months of the year were derived from WorldClim historical monthly weather data during the period 1970–2018. These monthly time series data were downscaled from the CRU TS version 4.03 (Harris et al. 2020) using WorldClim historical climate data version 2.1 (Fick and Hijmans 2017) for bias correction. The precipitation amount is usually not normally distributed, so a natural logarithmic transformation was applied to monthly precipitation (Johnson and Ingram 2004); as some monthly amounts were <1 mm, we added 1 mm to the original amount so that no negative values could be obtained after logarithmic transformation (Neter et al. 1996). To reduce the computer loading in fuzzy clustering, the original climate data product at a spatial resolution of 2.5′ (latitude) × 2.5′ (longitude) was resampled to 10′ × 10′ (i.e., 35 928 grids) in this study. The elevation at 10′ × 10′ resolution based on the Shuttle Radar Topography Mission (SRTM) (van Zyl 2001) was also acquired from the WorldClim version 2.1. As suggested in Terzer et al. (2013), if membership for any cluster at a specified grid point is less than 0.02, the fraction of this cluster can be redistributed to other clusters identified in the membership, and the sum of fractions at a specified grid always equals 1.
To determine the appropriate number of fuzzy clusters, the average silhouette value and the normalized Dunn’s partition coefficient (Dunn 1974; Kaufman and Rousseeuw 1990) were used. Generally, the optimal number of clusters is defined as the maximum of the average silhouette value and the normalized Dunn’s partition coefficient. The potential cluster setting from 3 to 10 clusters was tested in this study.
c. Best-fitting regressions and residual calibration
The observed precipitation isotopes data as well as the corresponding spatial (latitude, longitude, and elevation) and climate (air temperature and precipitation amount) parameters can be used to determine the best-fitting regressions for each cluster and month. The mean monthly air temperature and precipitation amount (Figs. S3 and S4) at the locations where the precipitation samples were collected and analyzed for δ18O and δ2H were extracted from the WorldClim historical monthly weather data version 2.1 with a spatial resolution of 2.5′ (latitude) × 2.5′ (longitude) during the period 1970–2018 (Fick and Hijmans 2017). Only the 223 Chinese sites were considered to determine the best-fitting regression equations. We applied the all-subsets regression to each cluster, and each climate cluster and month has a different regression model. The all-subsets regression was conducted using the function “regsubsets” (functions for model selection) of package “leaps” version 3.1 in R programming version 4.1.1. Unlike stepwise regression, all-subsets regression fits all possible combinations of the five or less independent variables from latitude, longitude, elevation, air temperature, and precipitation amount. Ignoring the cases where the δ18O and δ2H values were constant, 31 regression models were tested for each cluster and month (Table S3). Among all the possible regressions, we used adjusted r2 to determine the optimal equation. The isotope data at sampling sites in China with membership > 0.1 were selected to determine the fitting equation for each cluster.
Based on the optimal regression equations for each cluster and month, the precipitation isoscape can be predicted with input of the above-mentioned climatology of air temperature and precipitation during the period 1970–2018 at 10′ × 10′ resolution (i.e., 35 928 climate grid boxes) in China. The memberships for each cluster were used to combine the four layers of predictions into the integrated prediction.
To calibrate the integrated prediction, we reconducted the fuzzy clustering after adding the climate data of the 271 sites (including 223 Chinese sites and 48 sites from neighboring countries) to the above-mentioned 35 928 climate grids. The proportion of the additional sites is small (<1%) in the new fuzzy clustering, and the clustering solution did not change by very much for the 35 928 grids (r2 > 0.9999; see Fig. S5), so we assigned membership of these additional 271 sites within and outside China. Using the newly assigned membership and previously determined best-fitting regression equations, the predictions and the residuals (predictions minus observations) were calculated at the 271 sites. Then the residual layer, at 10′ × 10′ resolution, was interpolated from the residuals at the 271 sites using an inverse distance weight (IDW) method. Finally, we added the residual layer to the predicted δ18O and δ2H values at 10′ × 10′ resolution (Terzer et al. 2013).
The root-mean-square error (RMSE) and mean absolute error (MAE) were used to assess the precipitation isoscape modeling (Fick and Hijmans 2017).
d. Cross validation
To examine the accuracy of predictions, cross validation was conducted using a jackknife resampling method (Efron and Tibshirani 1994). The predictions can be recomputed from subsamples of the available data, where the subsamples are formed by discarding one site at a time from the full set of data. That is, when a 223-site network is considered to create a precipitation isoscape, the cross validation is conducted 223 times and only 222 sites are used at each time. The r2 values between the observations and two predictions are compared to understand the robustness of the regression in this study (Kabacoff 2015).
3. Results and discussion
a. Optimization of regression methods
According to the average silhouette value and normalized Dunn’s partition coefficient, the optimal number of climate clusters is set as 4 (Fig. 2; see also Table S4). The high fractional memberships for the four clusters are concentrated in the southeast (SE; Fig. 3a), northeast (NE; Fig. 3b), southwest (SW; Fig. 3c), and northwest (NW; Fig. 3d) China, respectively. According to the spatial distribution of cluster membership > 0.4, southeast China generally shows a tropical and subtropical monsoon climate (Wang and Zuo 2010) and has the highest mean precipitation and air temperature among the four clusters (Figs. 3e,i); northeast China roughly corresponds to a temperate monsoon climate (Wang and Zuo 2010) but is drier and colder than southeast China (Figs. 3f,j); southwest China includes a mountainous climate with high elevation and low temperature (Wang and Zuo 2010) and generally denotes the Qinghai–Tibet Plateau (Figs. 3g,k); northwest China generally corresponds to a temperate continental climate (Wang and Zuo 2010), with low-lying arid desert basins surrounded by high mountains (Figs. 3h,l). The spatial pattern is generally consistent with the comprehensive physical regionalization of China (Zhao 1983; Zheng et al. 2015). The best-fitting regression equations for each cluster and month are provided in Table S5.
Average silhouette value and normalized Dunn’s partition coefficient for 3–10 clusters.
Citation: Journal of Climate 35, 11; 10.1175/JCLI-D-21-0451.1
(a)–(d) Maps showing the membership of each cluster. The membership > 0.4 of four clusters is mainly concentrated in southeast (SE), northeast (NE), southwest (SW), and northwest (NW) China, respectively. Also shown is the monthly variation of (e)–(h) air temperature and (i)–(l) precipitation amount for each cluster with membership > 0.4 across China. The boxes represent the 25th–75th percentiles, and the line through the box represents the 50th percentile; the whiskers indicate the 90th and 10th and the points indicate the 95th and 5th percentiles.
Citation: Journal of Climate 35, 11; 10.1175/JCLI-D-21-0451.1
Approximately 77% of the residuals are concentrated in the range between ±3.5‰ for predicted δ18O, and the residuals between ±1.5‰ represent 42% (Fig. 4a). A consistently low MAE and RMSE all year round is considered to be ideal (Terzer et al. 2013). On an intra-annual scale (Fig. 4b), the model results during warmer/wetter months from April to October show lower MAE and RMSE, and the minimum is in May (1.9‰ for MAE and 2.5‰ for RMSE). Southeast China shows the lowest RMSE (Fig. 4b), indicating a relatively good prediction in the wet/warm region. In contrast, in many months southwest China presents the largest RMSE, showing that it is more difficult to model the precipitation isoscape for the Qinghai–Tibet Plateau due to the complex orography and uneven spatial coverage of available data (Yao et al. 2013). Compared with a previous regional isoscape study in the Qinghai–Tibet Plateau (Shi et al. 2020a), here more mountainous sites (>20 sites more) were compiled, which may improve the isoscapes at the high elevations. Winter usually exhibits a large RMSE, which may be associated with the large spatial variability in precipitation isotopes during winter (Liu et al. 2014). The cross-validation also suggests the model presents a good prediction (Figs. 4c,d). The r2 value between observed and predicted δ18O is 0.67 (p < 0.0001), and the result between observed and cross-validation predicted δ18O is very close (r2 = 0.61, p < 0.0001). The seasonal variability for δ2H (Table S6) is similar to that for δ18O.
(a) Residual probability distribution of predicted δ18O. (b) Monthly variations of root-mean-square error (RMSE) of predicted δ18O for each cluster with membership > 0.1. (c) Correlations between observed δ18O and predicted δ18O. (d) Correlations between observed δ18O and cross-validation predicted δ18O. The dashed lines denote the best-fitting line, and the solid lines denote the y = x line.
Citation: Journal of Climate 35, 11; 10.1175/JCLI-D-21-0451.1
b. Predicted monthly isoscapes
The monthly precipitation isoscapes of δ18O across China (C-Isoscape) are presented in Fig. 5. In winter (January; Fig. 5a), there is a decreasing trend in δ18O from the southeastern coast to the northwestern inland (Lee et al. 2012). The precipitation enriched in 18O is distributed in southeast China, and the regions with precipitation relatively depleted in 18O include the northwestern and northeastern margin of China as well as the northern Qinghai–Tibet Plateau. When the monsoon starts in late spring (May; Fig. 5e) (Yao et al. 2013; Yu et al. 2016), the region with precipitation enriched in 18O moves northward from southeast China to the central part of China, and the southeastern coast receives precipitation relatively depleted in 18O; this is due to the moisture sources and convective process from the lower latitudes (Cai and Tian 2016; Tang et al. 2017; Ruan et al. 2019). In summer (June; Fig. 5f), when monsoon precipitation is concentrated, the region with precipitation enriched in 18O moves to northwest China, especially in the arid low-lying desert basins where subcloud evaporation becomes significant as temperatures increase (Lee et al. 2012; Wang et al. 2021); in addition, the southern Qinghai–Tibet Plateau emerges as a region with precipitation depleted in heavy isotopes, which is consistent with the northward transport of moisture from the Indian monsoon (Yao et al. 2013). The region of precipitation relatively enriched in 18O settles in northwest China for June to September, and then joins the relatively depleted value region in the Qinghai–Tibet Plateau in autumn (October; Fig. 5j). In contrast, the southeastern coast generally shows an enriching 18O trend from autumn to winter. The spatial pattern for δ2H (Fig. S6) is similar to that for δ18O.
C-Isoscape maps showing δ18O in precipitation for each month.
Citation: Journal of Climate 35, 11; 10.1175/JCLI-D-21-0451.1
In the summer months, oceanic moisture from the Indian Ocean, the South China Sea, and the western Pacific Ocean is delivered to the continent, especially in the southeast portion of China; while in winter months, the contributions of oceanic sources are weak (Peng et al. 2010; Xie et al. 2011; Yao et al. 2013; Cai and Tian 2016). The westerlies moisture flow is always dominant in the northwest portion of China (S. Wang et al. 2017; Kong et al. 2019; Shi et al. 2021a). Regarding the main moisture sources across China, we can broadly divide it into the westerlies and monsoon dominated regions, and the boundary between westerlies and monsoon is usually from the middle Qinghai–Tibet Plateau to northeast China (Chen et al. 2019; J. Chen et al. 2021). The isotopic patterns in China are associated with these seasonal patterns in moisture paths.
Although many studies present Chinese precipitation isoscapes on an annual basis (e.g., Liu et al. 2008; P. Zhao et al. 2019), the monthly variation has not been previously examined nationwide, partly due to the limited spatial coverage of data (Zhang and Wang 2016). In the annual mean precipitation isoscapes, the regions with relatively depleted values are usually located at the Qinghai–Tibet Plateau in southwest China as well as the northern margin of northwest and northeast China. When comparing the spatial pattern of the previously generated annual precipitation isoscapes (e.g., Liu et al. 2008) with the C-Isoscape, the relatively depleted value regions are consistent.
We also provide the maps of interpolated residuals of δ18O and δ2H for each month in Figs. S7 and S8. Positive or negative residuals mean that the predicted δ18O and δ2H is underestimated or overestimated, respectively, in comparison to the actual value. Generally, strongly positive residuals (>4‰ for δ18O) are mainly seen in some sporadic areas in western China and in northeast China almost in all months (Figs. S7 and S8), which may be associated with the lack of data available in the most northerly extent of this region, or in bordering Russia (Figs. S1 and S2). In contrast, the negative residuals are mainly located at the western margin of China (Figs. S7 and S8). Due to the data limitations, we cannot comment further on the interannual variability of the isoscape. However, when multiple years of observations are available, the standard deviations of precipitation isotopes can be used to understand the interannual uncertainty. Generally, the uncertainty increases from the coast to inland, indicating the regions under a relatively arid background usually exhibit a strong interannual variability (Figs. S9 and S10). For the relatively arid regions, long-term observation is still needed to scientifically understand the atmospheric processes controlling precipitation isotopes (Liu et al. 2014; Shi et al. 2021a).
Generally, the single-peak curves for δ18O and δ2H in precipitation can be seen for the northern part of China, which means that the precipitation is enriched in 18O and 2H in summer and depleted in winter (Fig. 5 and Fig. S6). The southern boundary of areas with correlation coefficient > 0.9 between δ18O (or δ2H) and air temperature occurs at approximately 35°N for the western part of China, and approximately 40°N for the eastern part of China (Fig. 6a and Fig. S11a). In this study, the temperature effect (positive correlation between δ18O or δ2H value and temperature; Dansgaard 1964) generally corresponds to the regions influenced by the westerlies or regions where winter precipitation is in the form of snow. In contrast, the amount effect (negative correlation between δ18O or δ2H and precipitation amount; Dansgaard 1964) exists for the region mainly impacted by the monsoon. Some studies have tried to identify more specific moisture sources, for example, dividing the westerlies dominated region into westerlies and Arctic and dividing the monsoon dominated region into the Indian Ocean, South Pacific, and North Pacific (e.g., Araguás-Araguás et al. 1998; Kong et al. 2019). However, the actual transition zones between these regions may be vast, and interannual variability in their position also cannot be ignored (Goldsmith et al. 2017). As for the regions showing negative correlation between δ18O (or δ2H) and precipitation amount in China (Fig. 6b and Fig. S11b), the western and eastern sections indeed present different seasonal variability in precipitation isotopes (Fig. 5): for the western section of regions showing amount effect directly impacted by the Indian monsoon moisture, precipitation is 18O and 2H enriched in spring and depleted in late summer (Yao et al. 2013); for the eastern section where East Asian monsoon moisture is also strong, the isotopic peak in spring cannot be detected and a U-shaped isotopic variability is seen (Ruan et al. 2019).
Map showing the correlation coefficient between δ18O and (a) air temperature or (b) precipitation amount.
Citation: Journal of Climate 35, 11; 10.1175/JCLI-D-21-0451.1
The regression equation between δ18O and δ2H in precipitation, known as the meteoric water line, is usually used as a reference reflecting regional evaporation condition and other hydrological information in isotopic studies (Craig 1961; Putman et al. 2019). The slope and intercept of local meteoric water lines vary geographically (Crawford et al. 2014) and usually depart from the global mean as δ2H = 8δ18O + 10 to some degree (Craig 1961). In this study, 1815 pairs of monthly δ18O and δ2H compiled in China were used to determine a best-fitting meteoric water line of δ2H = 7.8δ18O + 8.7 (r2 = 0.97, p < 0.01, n = 1815) using least squares regression, δ2H = 7.9δ18O + 10.3 using precipitation weighted least squares regression also adopted in GNIP (IAEA 2020), or δ2H = 8.1δ18O + 11.6 using precipitation weighted reduced major axis regression as a less sensitive method to outliers recommended by Crawford et al. (2014). This equation is generally consistent with previous nationwide assessments such as Zheng et al. (1983) (δ2H = 7.9δ18O + 8.2; 101 daily samples at 8 sites), Gu (2011) (δ2H = 7.7δ18O + 7.0; annual means at 20 sites), and Liu et al. (2014) (δ2H = 7.5δ18O + 1.0; 928 monthly samples at 31 sites). However, the observation-based nationwide meteoric water is biased by data availability, and the arid periods and remote regions are not well sampled (Liu et al. 2014). Based on C-Isoscape, we determined a new nationwide meteoric water line as δ2H = 7.4δ18O + 5.5 (r2 = 0.93, p < 0.01, n = 431 136) using least squares regression, δ2H = 7.6δ18O + 7.1 using precipitation weighted least squares regression, or δ2H = 8.0δ18O + 10.2 using precipitation weighted reduced major axis regression. These meteoric water lines represent the whole domain equally, rather than being biased by data availability. However, because the regions where data availability are poor are predominately arid, the difference between the sample-based and C-Isoscape-based national meteoric water lines is smaller for the precipitation weighted regressions than for the ordinary least squares regression.
c. Comparison with global monthly isoscapes
As seen in previous global studies (e.g., Bowen and Revenaugh 2003; Terzer-Wassmuth et al. 2021), precipitation isoscapes for western China usually exhibit a larger uncertainty than seen for eastern China. Here we calculate the difference between the C-Isoscape in this study and previously published global precipitation isoscapes (OIPC and RCWIP) (Fig. 7, Figs. S12–S14). Among the four subregions, southeast China usually shows a minimal difference between the three precipitation isoscapes (Fig. 7), indicating that they are all good at modeling this region, with its relatively weak seasonal variability in precipitation isotopes (Fig. 5).
Monthly distributions of RCWIP minus C-Isoscape for precipitation δ18O.
Citation: Journal of Climate 35, 11; 10.1175/JCLI-D-21-0451.1
In northwest China, the OIPC and RCWIP usually overestimate the precipitation δ18O and δ2H in winter and underestimate them in summer (Fig. 7, Figs. S12–S14). This is associated with the very limited observations in the GNIP database, with only two sites (Hotan and Urumqi) west of 100°E in northwest China (Zhang and Wang 2016; Liu et al. 2014), which cannot adequately represent the spatial variability in arid central Asia. Under the extremely arid climate of northwest China, falling raindrops may experience strong evaporation below the cloud base, and precipitation collected near surface is enriched in 18O and 2H (Zhang and Wang 2018; Wang et al. 2021; Pang et al. 2011). In addition, the strong orographic influence of the Tianshan Mountains leads to a dramatic rain-shadow effect resulting in scarce precipitation on the leeward side, which is enriched in heavy isotopes due to below-cloud evaporation, except in winter when a slight isotopic rain-shadow effect leads to a depletion in heavy isotopes due to rainout (Wang et al. 2016b,c); this subtle spatial variation requires more spatial coverage to resolve in a precipitation isoscape than provided by GNIP.
In southwest China, the two global precipitation isoscapes usually predict relatively depleted values than the C-Isoscape from this study, and underestimate the summer δ18O and δ2H values (e.g., Figs. 7g,h). In addition, there is only one GNIP site (Lhasa) on the Qinghai–Tibet Plateau, and the lack of observations causes the large uncertainty in OIPC and RCWIP. The observation work in recent decades provides more precipitation isotope data for this plateau region (Yao et al. 2013).
Although researchers have contributed considerable observations of precipitation δ18O and δ2H (not included in GNIP) in the past two decades (Zhang and Wang 2016), there are still clear spatial gaps in arid and inaccessible areas. Due to the great spatial diversity and seasonal variability in precipitation isotopes across western China (Fig. 5 and Fig. S6), including northwest and southwest China, a denser observation network is still needed. It should also be noted that the spatial coverage of precipitation isotopes is constrained by the availability of local providers (Ye et al. 2019), and many studies are only short term, not capturing interannual variation (Table S2). To produce high-quality datasets lasting for decades requires stations to be attended (e.g., meteorological or hydrological stations). However, in the current nationwide meteorological (e.g., Zhang et al. 2020) and hydrological (e.g., Cong et al. 2010) observation network, western China, dominated by deserts and mountain ranges, continues to have many monitoring blind spots. Actually, it is not only in these unmonitored locations that additional data would be valuable. The new precipitation isoscapes presented here, including the residuals in Figs. S7 and S8, will help to identify where the isoscape is most regionally representative, and where it would be most valuable to optimize the measurement network (Hatvani et al. 2021).
4. Conclusions
During the past decades, a large number of precipitation δ18O and δ2H measurements have been conducted across China, but there is no high-resolution nationwide monthly precipitation isoscape in China. Here we compiled a nationwide database including 223 sampling sites, and applied a regionalized fuzzy clustering method to produce a monthly precipitation isoscape product, C-Isoscape. The moisture transportation path, controlled by the westerlies and the monsoon, results in different spatial and seasonal diversity of precipitation isotopes. Compared with previous global precipitation isoscape products such as OIPC and RCWIP, the C-Isoscape usually shows precipitation more enriched in heavy isotopes in summer and more depleted in winter for northwest China, while the C-Isoscape precipitation is more 18O and 2H enriched in most months for southwest China. Providing a geographic perspective on isotope fractionation, the monthly precipitation isoscapes interpolated in this study can be regarded as a long-term reference for isotopic diagnostics in atmospheric processes.
Acknowledgments.
The authors greatly thank all the researchers for providing the isotope data in precipitation across China, especially the IAEA/WMO Global Network of Isotopes in Precipitation, Tandong Yao (Institute of Tibetan Plateau Research, Chinese Academy of Sciences), Xianfang Song (Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences), Lide Tian (Yunnan University), Liangcheng Tan (Institute of Earth Environment, Chinese Academy of Sciences), Xiaofan Zhu (Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences), Xue Qiu (Lanzhou City University), Xiaoyan Wang (Weinan Normal University), Fenli Chen (Northwest Normal University), Liangju Zhao (Northwest University), Yanlong Kong (Institute of Geology and Geophysics, Chinese Academy of Sciences), Congjian Sun (Shanxi Normal University) and Huawu Wu (Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences). The authors also thank Liwei Wang, Lihong Duan, Yijie Xia (Northwest Normal University), and Shaohua Dang (Tongji University) for help in data processing and discussion. The research is supported by the National Natural Science Foundation of China (41971034 and 41701028), the Foundation for Distinguished Young Scholars of Gansu Province (20JR10RA112), and the Northwest Normal University (NWNU-LKZD2021-04).
Data availability statement.
The C-Isoscape product version 1.0 (the GeoTIFF and text formats) and the publicly available precipitation isotope input can be accessed from https://github.com/isoscape/cisoscape1.0. The WorldClim historical monthly weather data version 2.1 are acquired from https://www.worldclim.org/data/worldclim21.html. The CPC Global Unified Gauge-based Analysis of Daily Precipitation is available at https://downloads.psl.noaa.gov/Datasets/cpc_global_precip. OIPC version 3.2 is available at http://wateriso.utah.edu/waterisotopes. RCWIP version 1.00 is available at http://www-naweb.iaea.org/napc/ih/IHS_resources_rcwip.html.
APPENDIX
Fuzzy Clustering
REFERENCES
Acharya, S., X. Yang, T. Yao, and D. Shrestha, 2020: Stable isotopes of precipitation in Nepal Himalaya highlight the topographic influence on moisture transport. Quat. Int., 565, 22–30, https://doi.org/10.1016/j.quaint.2020.09.052.
Adhikari, N., J. Gao, T. Yao, Y. Yang, and D. Dai, 2020: The main controls of the precipitation stable isotopes at Kathmandu, Nepal. Tellus, 72B, 1445379, https://doi.org/10.1080/16000889.2020.1721967.
Allen, S. T., J. W. Kirchner, and G. R. Goldsmith, 2018: Predicting spatial patterns in precipitation isotope (δ2H and δ18O) seasonality using sinusoidal isoscapes. Geophys. Res. Lett., 45, 4859–4868, https://doi.org/10.1029/2018GL077458.
Araguás-Araguás, L., K. Froehlich, and K. Rozanski, 1998: Stable isotope composition of precipitation over Southeast Asia. J. Geophys. Res., 103, 28 721–28 742, https://doi.org/10.1029/98JD02582.
Bowen, G. J., and B. H. Wilkinson, 2002: Spatial distribution of δ18O in meteoric precipitation. Geology, 30, 315–318, https://doi.org/10.1130/0091-7613(2002)030<0315:SDOOIM>2.0.CO;2.
Bowen, G. J., and J. Revenaugh, 2003: Interpolating the isotopic composition of modern meteoric precipitation. Water Resour. Res., 39, 1299, https://doi.org/10.1029/2003WR002086.
Bowen, G. J., and S. P. Good, 2015: Incorporating water isoscapes in hydrological and water resource investigations. Wiley Interdiscip. Rev.: Water, 2, 107–119, https://doi.org/10.1002/wat2.1069.
Bowen, G. J., L. I. Wassenaar, and K. A. Hobson, 2005: Global application of stable hydrogen and oxygen isotopes to wildlife forensics. Oecologia, 143, 337–348, https://doi.org/10.1007/s00442-004-1813-y.
Bowen, G. J., J. S. Guo, and S. T. Allen, 2022: A 3-D groundwater isoscape of the contiguous USA for forensic and water resource science. PLOS ONE, 17, e0261651, https://doi.org/10.1371/journal.pone.0261651.
Burnik Šturm, M., O. Ganbaatar, C. C. Voigt, and P. Kaczensky, 2017: First field-based observations of δ2H and δ18O values of event-based precipitation, rivers and other water bodies in the Dzungarian Gobi, SW Mongolia. Isotopes Environ. Health Stud., 53, 157–171, https://doi.org/10.1080/10256016.2016.1231184.
Cai, Z., and L. Tian, 2016: Atmospheric controls on seasonal and interannual variations in the precipitation isotope in the East Asian monsoon region. J. Climate, 29, 1339–1352, https://doi.org/10.1175/JCLI-D-15-0363.1.
Chen, C. J., and T. Y. Li, 2018: Geochemical characteristics of cave drip water respond to ENSO based on a 6-year monitoring work in Yangkou Cave, Southwest China. J. Hydrol., 561, 896–907, https://doi.org/10.1016/j.jhydrol.2018.04.061.
Chen, F., M. Zhang, S. Wang, X. Qiu, and M. Du, 2017: Environmental controls on stable isotopes of precipitation in Lanzhou, China: An enhanced network at city scale. Sci. Total Environ., 609, 1013–1022, https://doi.org/10.1016/j.scitotenv.2017.07.216.
Chen, F., and Coauthors, 2019: Westerlies Asia and monsoonal Asia: Spatiotemporal differences in climate change and possible mechanisms on decadal to sub-orbital timescales. Earth-Sci. Rev., 192, 337–354, https://doi.org/10.1016/j.earscirev.2019.03.005.
Chen, F., C. Huang, Q. Lao, S. Zhang, C. Chen, X. Zhou, X. Lu, and Q. Zhu, 2021: Typhoon control of precipitation dual isotopes in southern China and its palaeoenvironmental implications. J. Geophys. Res. Atmos., 126, e2020JD034336, https://doi.org/10.1029/2020JD034336.
Chen, J., J. Cao, and Y. Huang, 2010: The hydrogen and oxygen isotope composition of precipitation in the Xiamen coastal area. J. Mar. Sci., 28, 11–17, https://doi.org/10.3969/j.issn.1001-909X.2010.01.002.
Chen, J., W. Huang, S. Feng, Q. Zhang, X. Kuang, J. Chen, and F. Chen, 2021: The modulation of westerlies–monsoon interaction on climate over the monsoon boundary zone in East Asia. Int. J. Climatol., 41, E3049–E3064, https://doi.org/10.1002/joc.6903.
Chen, M., W. Shi, P. Xie, V. B. S. Silva, V. E. Kousky, R. W. Higgins, and J. E. Janowiak, 2008: Assessing objective techniques for gauge-based analyses of global daily precipitation. J. Geophys. Res., 113, D04110, https://doi.org/10.1029/2007JD009132.
Chiang, J. C., M. J. Herman, K. Yoshimura, and I. Y. Fung, 2020: Enriched East Asian oxygen isotope of precipitation indicates reduced summer seasonality in regional climate and westerlies. Proc. Natl. Acad. Sci. USA, 117, 14 745–14 750, https://doi.org/10.1073/pnas.1922602117.
Cong, Z., J. Zhao, D. Yang, and G. Ni, 2010: Understanding the hydrological trends of river basins in China. J. Hydrol., 388, 350–356, https://doi.org/10.1016/j.jhydrol.2010.05.013.
Craig, H., 1961: Isotopic variations in meteoric waters. Science, 133, 1702–1703, https://doi.org/10.1126/science.133.3465.1702.
Crawford, J., C. E. Hughes, and S. Lykoudis, 2014: Alternative least squares methods for determining the meteoric water line, demonstrated using GNIP data. J. Hydrol., 519, 2331–2340, https://doi.org/10.1016/j.jhydrol.2014.10.033.
Cui, J., L. Tian, T. W. Biggs, and R. Wen, 2017: Deuterium-excess determination of evaporation to inflow ratios of an alpine lake: Implications for water balance and modeling. Hydrol. Processes, 31, 1034–1046, https://doi.org/10.1002/hyp.11085.
Dansgaard, W., 1964: Stable isotopes in precipitation. Tellus, 16, 436–468, https://doi.org/10.3402/tellusa.v16i4.8993.
Dayem, K. E., P. Molnar, D. S. Battisti, and G. H. Roe, 2010: Lessons learned from oxygen isotopes in modern precipitation applied to interpretation of speleothem records of paleoclimate from eastern Asia. Earth Planet. Sci. Lett., 295, 219–230, https://doi.org/10.1016/j.epsl.2010.04.003.
Dunn, J. C., 1974: Some recent investigations of a new fuzzy partitioning algorithm and its application to pattern classification problems. J. Cybern., 4 (2), 1–15, https://doi.org/10.1080/01969727408546062.
Efron, B., and R. J. Tibshirani, 1994: An Introduction to the Bootstrap. Chapman and Hall/CRC, 456 pp.
Falster, G., B. Konecky, M. Madhavan, S. Stevenson, and S. Coats, 2021: Imprint of the Pacific Walker circulation in global precipitation δ18O. J. Climate, 34, 8579–8597, https://doi.org/10.1175/JCLI-D-21-0190.1.
Fick, S. E., and R. J. Hijmans, 2017: WorldClim 2: New 1-km spatial resolution climate surfaces for global land areas. Int. J. Climatol., 37, 4302–4315, https://doi.org/10.1002/joc.5086.
Gao, J., Y. He, V. Masson-Delmotte, and T. Yao, 2018: ENSO effects on annual variations of summer precipitation stable isotopes in Lhasa, southern Tibetan Plateau. J. Climate, 31, 1173–1182, https://doi.org/10.1175/JCLI-D-16-0868.1.
Goldsmith, Y., and Coauthors, 2017: Northward extent of East Asian monsoon covaries with intensity on orbital and millennial timescales. Proc. Natl. Acad. Sci. USA, 114, 1817–1821, https://doi.org/10.1073/pnas.1616708114.
Gu, W., 2011: Isotope Hydrology. Science Press, 1113 pp.
Harris, I., T. J. Osborn, P. Jones, and D. Lister, 2020: Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Sci. Data, 7, 109, https://doi.org/10.1038/s41597-020-0453-3.
Hatvani, I. G., M. Leuenberger, B. Kohán, and Z. Kern, 2017: Geostatistical analysis and isoscape of ice core derived water stable isotope records in an Antarctic macro region. Polar Sci., 13, 23–32, https://doi.org/10.1016/j.polar.2017.04.001.
Hatvani, I. G., and Coauthors, 2021: Geostatistical evaluation of the design of the precipitation stable isotope monitoring network for Slovenia and Hungary. Environ. Int., 146, 106263, https://doi.org/10.1016/j.envint.2020.106263.
Hobson, K. A., and L. I. Wassenaar, 2019: Tracking Animal Migration with Stable Isotopes. 2nd ed. Elsevier, 253 pp., https://doi.org/10.1016/C2017-0-01125-4.
Hobson, K. A., K. Doward, K. J. Kardynal, and J. N. McNeil, 2018: Inferring origins of migrating insects using isoscapes: A case study using the true armyworm, Mythimna unipuncta, in North America. Ecol. Entomol., 43, 332–341, https://doi.org/10.1111/een.12505.
Hollins, S. E., C. E. Hughes, J. Crawford, D. I. Cendón, and K. T. Meredith, 2018: Rainfall isotope variations over the Australian continent—Implications for hydrology and isoscape applications. Sci. Total Environ., 645, 630–645, https://doi.org/10.1016/j.scitotenv.2018.07.082.
Hu, J., J. Emile-Geay, J. Nusbaumer, and D. Noone, 2018: Impact of convective activity on precipitation δ18O in isotope-enabled general circulation models. J. Geophys. Res. Atmos., 123, 13 595–13 610, https://doi.org/10.1029/2018JD029187.
IAEA, 2020: Water isotope system for data analysis, visualization and electronic retrieval. International Atomic Energy Agency, accessed 29 January 2020, https://nucleus.iaea.org/wiser/index.aspx.
Jia, W., X. Ma, X. Xu, R. Yuan, D. Ding, and G. Zhu, 2018: Composition of stable isotope in precipitation and its influences by different vapor sources in the eastern Qilian Mountains. J. Mt. Sci., 15, 2207–2217, https://doi.org/10.1007/s11629-018-4844-2.
Jiao, Y., C. Liu, X. Gao, Q. Xu, Y. Ding, and Z. Liu, 2019: Impacts of moisture sources on the isotopic inverse altitude effect and amount of precipitation in the Hani Rice Terraces region of the Ailao Mountains. Sci. Total Environ., 687, 470–478, https://doi.org/10.1016/j.scitotenv.2019.05.426.
Jin, K., W. Rao, J. Sun, and W. Guo, 2015: Isotopic characteristics and source of precipitation in the Ordos desert area. Yellow River, 37, 31–45, https://doi.org/10.3969/j.issn.1000-1379.2015.03.009.
Jin, Z., Y. Wang, F. Li, L. Qian, Y. Hu, and Y. Shi, 2019: Stable isotopes and chemical characteristics of precipitation in Hangzhou and Huzhou, East China. Environ. Sci. Pollut. Res. Int., 26, 23 717–23 729, https://doi.org/10.1007/s11356-019-05712-y.
Johnson, K. R., and B. L. Ingram, 2004: Spatial and temporal variability in the stable isotope systematics of modern precipitation in China: Implications for paleoclimate reconstructions. Earth Planet. Sci. Lett., 220, 365–377, https://doi.org/10.1016/S0012-821X(04)00036-6.
Juhlke, T. R., C. Meier, R. van Geldern, K. A. Vanselow, J. Wernicke, J. Baidulloeva, J. A. C. Barth, and S. M. Weise, 2019: Assessing moisture sources of precipitation in the western Pamir Mountains (Tajikistan, central Asia) using deuterium excess. Tellus, 71B, 1601987, https://doi.org/10.1080/16000889.2019.1601987.
Kabacoff, R. I., 2015: R in Action: Data Analysis and Graphics with R. 2nd ed. Manning, 608 pp.
Kaufman, L., and P. J. Rousseeuw, 1990: Finding Groups in Data: An Introduction to Cluster Analysis. John Wiley, 342 pp.
Kong, Y., K. Wang, J. Li, and Z. Pang, 2019: Stable isotopes of precipitation in China: A consideration of moisture sources. Water, 11, 1239, https://doi.org/10.3390/w11061239.
Kumar, B., and Coauthors, 2010: Isotopic characteristics of Indian precipitation. Water Resour. Res., 46, W12548, https://doi.org/10.1029/2009WR008532.
Kurita, N., N. Yoshida, G. Inoue, and E. A. Chayanova, 2004: Modern isotope climatology of Russia: A first assessment. J. Geophys. Res., 109, D03102, https://doi.org/10.1029/2003JD003404.
Lachniet, M. S., D. E. Lawson, H. Stephen, A. R. Sloat, and W. P. Patterson, 2016: Isoscapes of δ18O and δ2H reveal climatic forcings on Alaska and Yukon precipitation. Water Resour. Res., 52, 6575–6586, https://doi.org/10.1002/2016WR019436.
Lee, J. E., C. Risi, I. Fung, J. Worden, R. A. Scheepmaker, B. Lintner, and C. Frankenberg, 2012: Asian monsoon hydrometeorology from TES and SCIAMACHY water vapor isotope measurements and LMDZ simulations: Implications for speleothem climate record interpretation. J. Geophys. Res., 117, D15112, https://doi.org/10.1029/2011JD017133.
Li, Y., M. Zhang, S. Wang, Z. Li, and X. Li, 2011: Spatial distribution of δ18O in China’s precipitation based on a secondary variable of temperature. Prog. Geogr., 30, 1387–1394, https://doi.org/10.11820/dlkxjz.2011.11.008.
Li, Z., and Coauthors, 2019: Water resources in inland regions of central Asia: Evidence from stable isotope tracing. J. Hydrol., 570, 1–16, https://doi.org/10.1016/j.jhydrol.2019.01.003.
Liu, J., X. Song, G. Yuan, X. Sun, X. Liu, and S. Wang, 2010: Characteristics of δ18O in precipitation over eastern monsoon China and the water vapor sources. Chin. Sci. Bull., 55, 200–211, https://doi.org/10.1007/s11434-009-0202-7.
Liu, J., X. Song, G. Yuan, X. Sun, and L. Yang, 2014: Stable isotopic compositions of precipitation in China. Tellus, 66B, 22567, https://doi.org/10.3402/tellusb.v66.22567.
Liu, Z., L. Tian, X. Chai, and T. Yao, 2008: A model-based determination of spatial variation of precipitation δ18O over China. Chem. Geol., 249, 203–212, https://doi.org/10.1016/j.chemgeo.2007.12.011.
Luo, W., S. Wang, and X. Xie, 2013: A comparative study on the stable isotopes from precipitation to speleothem in four caves of Guizhou, China. Chem. Erde, 73, 205–215, https://doi.org/10.1016/j.chemer.2012.05.002.
Maher, B. A., and R. Thompson, 2012: Oxygen isotopes from Chinese caves: Records not of monsoon rainfall but of circulation regime. J. Quat. Sci., 27, 615–624, https://doi.org/10.1002/jqs.2553.
Malygina, N. S., A. N. Eirikh, and N. Y. Kurepina, 2019: Isotopic composition of precipitation in Altai foothills: Observation and interpolation data. Bull. Tomsk Polytech. Univ. Geol. Assets Eng., 330, 44–54, http://izvestiya.tpu.ru/archive/article/view/91/41.
Neter, J. W., M. H. Kutner, C. J. Nachtsheim, and W. Wasserman, 1996: Applied Linear Statistical Models. 4th ed. Times Mirror Higher Education Group, 720 pp.
Pang, S., S. Zhao, R. Wen, and Z. Liu, 2015: Spatial and temporal variation of stable isotopes in precipitation in the Haihe River basin. Chin. Sci. Bull., 60, 1218–1226, https://doi.org/10.1360/N972014-01040.
Pang, Z., Y. Kong, K. Froehlich, T. Huang, L. Yuan, Z. Li, and F. Wang, 2011: Processes affecting isotopes in precipitation of an arid region. Tellus, 63B, 352–359, https://doi.org/10.1111/j.1600-0889.2011.00532.x.
Peng, P., X. J. Zhang, and J. Chen, 2020: Bias correcting isotope-equipped GCMs outputs to build precipitation oxygen isoscape for eastern China. J. Hydrol., 589, 125153, https://doi.org/10.1016/j.jhydrol.2020.125153.
Peng, T.-R., C.-H. Wang, C.-C. Huang, L.-Y. Fei, C.-T. A. Chen, and J.-L. Hwong, 2010: Stable isotopic characteristic of Taiwan’s precipitation: A case study of western Pacific monsoon region. Earth Planet. Sci. Lett., 289, 357–366, https://doi.org/10.1016/j.epsl.2009.11.024.
Putman, A. L., R. P. Fiorella, G. J. Bowen, and Z. Cai, 2019: A global perspective on local meteoric water lines: Meta-analytic insight into fundamental controls and practical constraints. Water Resour. Res., 55, 6896–6910, https://doi.org/10.1029/2019WR025181.
Qiu, X., and Coauthors, 2019: The test of the ecohydrological separation hypothesis in a dry zone of the northeastern Tibetan Plateau. Ecohydrology, 12, e2077, https://doi.org/10.1002/eco.2077.
Rao, W., W. Zhang, B. Yong, H. Tan, K. T. Meredith, K. Jin, F. Zheng, and S. Wang, 2018: Identifying the source of atmospheric moisture over arid deserts using stable isotopes (2H and 18O) in precipitation. Hydrol. Processes, 32, 436–449, https://doi.org/10.1002/hyp.11431.
Ren, W., T. Yao, X. Yang, and D. R. Joswiak, 2013: Implications of variations in δ18O and δD in precipitation at Madoi in the eastern Tibetan Plateau. Quat. Int., 313–314, 56–61, https://doi.org/10.1016/j.quaint.2013.05.026.
Ruan, J., H. Zhang, Z. Cai, X. Yang, and J. Yin, 2019: Regional controls on daily to interannual variations of precipitation isotope ratios in southeast China: Implications for paleomonsoon reconstruction. Earth Planet. Sci. Lett., 527, 115794, https://doi.org/10.1016/j.epsl.2019.115794.
Sánchez-Murillo, R., and C. Birkel, 2016: Groundwater recharge mechanisms inferred from isoscapes in a complex tropical mountainous region. Geophys. Res. Lett., 43, 5060–5069, https://doi.org/10.1002/2016GL068888.
Shi, X., C. Risi, T. Pu, J.-L. Lacour, Y. Kong, K. Wang, Y. He, and D. Xia, 2020: Variability of isotope composition of precipitation in the southeastern Tibetan Plateau from the synoptic to seasonal time scale. J. Geophys. Res. Atmos., 125, e2019JD031751, https://doi.org/10.1029/2019JD031751.
Shi, Y., S. Wang, M. Zhang, A. A. Argiriou, R. Guo, Y. Song, and X. Zhu, 2020a: Isoscape of δ18O in precipitation of the Qinghai-Tibet Plateau: Assessment and improvement. Water, 12, 3392, https://doi.org/10.3390/w12123392.
Shi, Y., S. Wang, M. Zhang, Y. Li, and Y. Song, 2020b: Applicability on the OIPC and RCWIP stable hydrogen and oxygen isotope data in precipitation across the Tianshan Mountains, Xinjiang. J. Glaciol. Geocryol., 42, 974–985, https://doi.org/10.7522/j.issn.1000-0240.2019.0075.
Shi, Y., Z. Jin, A. Wu, G. Li, and F. Li, 2021a: Stable isotopic characteristics of precipitation related to the environmental controlling factors in Ningbo, East China. Environ. Sci. Pollut. Res. Int., 28, 10 696–10 706, https://doi.org/10.1007/s11356-020-11332-8.
Shi, Y., S. Wang, L. Wang, M. Zhang, A. A. Argiriou, Y. Song, and S. Lei, 2021b: Isotopic evidence in modern precipitation for the westerly meridional movement in Central Asia. Atmos. Res., 259, 105698, https://doi.org/10.1016/j.atmosres.2021.105698.
Sturm, C., Q. Zhang, and D. Noone, 2010: An introduction to stable water isotopes in climate models: Benefits of forward proxy modelling for paleoclimatology. Climate Past, 6, 115–129, https://doi.org/10.5194/cp-6-115-2010.
Sun, C., Y. Chen, J. Li, W. Chen, and X. Li, 2019: Stable isotope variations in precipitation in the northwesternmost Tibetan Plateau related to various meteorological controlling factors. Atmos. Res., 227, 66–78, https://doi.org/10.1016/j.atmosres.2019.04.026.
Sun, C., W. Chen, Y. Chen, and Z. Cai, 2020: Stable isotopes of atmospheric precipitation and its environmental drivers in the Eastern Chinese Loess Plateau, China. J. Hydrol., 581, 124404, https://doi.org/10.1016/j.jhydrol.2019.124404.
Sun, Z., Y. Yang, J. Zhao, N. Tian, and X. Feng, 2018: Potential ENSO effects on the oxygen isotope composition of modern speleothems: Observations from Jiguan Cave, central China. J. Hydrol., 566, 164–174, https://doi.org/10.1016/j.jhydrol.2018.09.015.
Tang, Y., X. Song, Y. Zhang, D. Han, L. Ai, T. Zhao, and Y. Wang, 2017: Using stable isotopes to understand seasonal and interannual dynamics in moisture sources and atmospheric circulation in precipitation. Hydrol. Processes, 31, 4682–4692, https://doi.org/10.1002/hyp.11388.
Terzer, S., L. I. Wassenaar, L. J. Araguás-Araguás, and P. K. Aggarwal, 2013: Global isoscapes for δ18O and δ2H in precipitation: Improved prediction using regionalized climatic regression models. Hydrol. Earth Syst. Sci., 17, 4713–4728, https://doi.org/10.5194/hess-17-4713-2013.
Terzer-Wassmuth, S., L. I. Wassenaar, J. M. Welker, and L. J. Araguás-Araguás, 2021: Improved high-resolution global and regionalized isoscapes of δ18O, δ2H, and d-excess in precipitation. Hydrol. Processes, 35, e14254, https://doi.org/10.1002/hyp.14254.
Tian, C., P. Meng, J. Zhang, S. Sun, C. He, C. Jia, and J. Li, 2015: Seasonal characteristics of δD and δ18O in precipitation and its water vapor sources in the Xiaolangdi Reservoir area of Yellow River, China. Chin. J. Appl. Ecol., 26, 3579–3587, https://doi.org/10.13287/j.1001-9332.20150929.009.
Tian, L., T. Yao, K. MacClune, J. W. C. White, A. Schilla, B. Vaughn, R. Vachon, and K. Ichiyanagi, 2007: Stable isotopic variations in west China: A consideration of moisture sources. J. Geophys. Res., 112, D10112, https://doi.org/10.1029/2006JD007718.
Vachon, R. W., J. M. Welker, J. W. C. White, and B. H. Vaughn, 2010: Monthly precipitation isoscapes (δ18O) of the United States: Connections with surface temperatures, moisture source conditions, and air mass trajectories. J. Geophys. Res., 115, D21126, https://doi.org/10.1029/2010JD014105.
van Zyl, J. J., 2001: The Shuttle Radar Topography Mission (SRTM): A breakthrough in remote sensing of topography. Acta Astronaut., 48, 559–565, https://doi.org/10.1016/S0094-5765(01)00020-0.
Wang, H., J. Zhang, and Z. Liu, 2012: Indications of the hydrogen and oxygen isotopes in precipitation for climate change in Huanglong, Sichuan. Carsol. Sin., 31, 253–258, https://doi.org/10.3969/j.issn.1001-4810.2012.03.005.
Wang, J., and W. Zuo, 2010: Geographic Atlas of China. SinoMaps Press, 365 pp.
Wang, L., Y. Dong, D. Han, and Z. Xu, 2019: Stable isotopic compositions in precipitation over wet island in central Asia. J. Hydrol., 573, 581–591, https://doi.org/10.1016/j.jhydrol.2019.04.005.
Wang, Q., and Coauthors, 2018: The transfer of oxygen isotopic signals from precipitation to drip water and modern calcite on the seasonal time scale in Yongxing Cave, central China. Environ. Earth Sci., 77, 474, https://doi.org/10.1007/s12665-018-7607-z.
Wang, S., M. Zhang, Y. Che, F. Chen, and F. Qiang, 2016a: Contribution of recycled moisture to precipitation in oases of arid central Asia: A stable isotope approach. Water Resour. Res., 52, 3246–3257, https://doi.org/10.1002/2015WR018135.
Wang, S., M. Zhang, Y. Che, X. Zhu, and X. Liu, 2016b: Influence of below-cloud evaporation on deuterium excess in precipitation of arid central Asia and its meteorological controls. J. Hydrometeor., 17, 1973–1984, https://doi.org/10.1175/JHM-D-15-0203.1.
Wang, S., M. Zhang, C. E. Hughes, X. Zhu, L. Dong, Z. Ren, and F. Chen, 2016c: Factors controlling stable isotope composition of precipitation in arid conditions: An observation network in the Tianshan Mountains, central Asia. Tellus, 68B, 26206, https://doi.org/10.3402/tellusb.v68.26206.
Wang, S., M. Zhang, J. Crawford, C. E. Hughes, M. Du, and X. Liu, 2017: The effect of moisture source and synoptic conditions on precipitation isotopes in arid Central Asia. J. Geophys. Res. Atmos., 122, 2667–2682, https://doi.org/10.1002/2015JD024626.
Wang, S., Q. Wang, J. Wu, X. He, Y. Ding, L. Wang, and Z. Hu, 2019: Characteristics of stable isotopes in precipitation and moisture sources in the headwaters of the Yangtze River. Environ. Sci., 40, 2615–2623, https://doi.org/10.13227/j.hjkx.201811140.
Wang, S., R. Jiao, M. Zhang, J. Crawford, C. E. Hughes, and F. Chen, 2021: Changes in below-cloud evaporation affect precipitation isotopes during five decades of warming across China. J. Geophys. Res. Atmos., 126, e2020JD033075, https://doi.org/10.1029/2020JD033075.
Wang, X., Z. Li, R. Tayier, and S. Wang, 2015: Characteristics of atmospheric precipitation isotopes and isotopic evidence for the moisture origin in Yushugou River basin, eastern Tianshan Mountains, China. Quat. Int., 380–381, 106–115, https://doi.org/10.1016/j.quaint.2014.12.023.
Wang, X., A. Lu, C. Jiang, S. Wang, and H. Liu, 2017: Characteristics of δD and δ18O in precipitation and the moisture origin in Weinan. J. Arid Land Resour. Environ., 31, 122–128, https://doi.org/10.13448/j.cnki.jalre.2017.257.
Wang, Y., C. Hu, J. Ruan, and K. R. Johnson, 2020: East Asian precipitation δ18O relationship with various monsoon indices. J. Geophys. Res. Atmos., 125, e2019JD032282, https://doi.org/10.1029/2019JD032282.
West, J. B., G. J. Bowen, T. E. Dawson, and K. P. Tu, 2010: Isoscapes: Understanding Movement, Pattern, and Process on Earth through Isotope Mapping. Springer, 487 pp., https://doi.org/10.1007/978-90-481-3354-3..
Wu, H., X. Y. Li, J. Zhang, J. Li, J. Liu, L. Tian, and C. Fu, 2019: Stable isotopes of atmospheric water vapour and precipitation in the northeast Qinghai-Tibetan Plateau. Hydrol. Processes, 33, 2997–3009, https://doi.org/10.1002/hyp.13541.
Wushiki, H., 1977: Deuterium content in the Himalayan precipitation at Khumbu district, observed in 1974/1975. J. Japan. Soc. Snow Ice, 39, 50–56, https://doi.org/10.5331/seppyo.39.Special_50.
Xie, L., G. Wei, W. Deng, and X. Zhao, 2011: Daily δ18O and δD of precipitations from 2007 to 2009 in Guangzhou, South China: Implications for changes of moisture sources. J. Hydrol., 400, 477–489, https://doi.org/10.1016/j.jhydrol.2011.02.002.
Xu, Q., S. Liu, S. An, C. Jiang, and X. Liu, 2006: Study on hydrogen and oxygen stable isotopes in precipitation in Wolong Nature Reserve, Sichuan Province. For. Res., 19, 679–686, https://doi.org/10.3321/j.issn:1001-1498.2006.06.002.
Yamanaka, T., M. Tsujimura, D. Oyunbaatar, and G. Davaa, 2007: Isotopic variation of precipitation over eastern Mongolia and its implication for the atmospheric water cycle. J. Hydrol., 333, 21–34, https://doi.org/10.1016/j.jhydrol.2006.07.022.
Yang, J., X. Qin, J. Wu, K. Duan, W. Du, and H. Liu, 2014: The application of modified BW method in studying spatial distribution of δ18O in precipitation over China. J. Glaciol. Geocryol., 36, 1430–1439, https://doi.org/10.7522/j.issn.1000-0240.2014.0170.
Yang, S., M. Zhang, S. Wang, Y. Liu, F. Qiang, and D. Qu, 2017: Interannual trends in stable oxygen isotope composition in precipitation of China during 1979–2007: Spatial incoherence. Quat. Int., 454, 25–37, https://doi.org/10.1016/j.quaint.2017.07.029.
Yang, X., M. E. Davis, S. Acharya, and T. Yao, 2018: Asian monsoon variations revealed from stable isotopes in precipitation. Climate Dyn., 51, 2267–2283, https://doi.org/10.1007/s00382-017-4011-4.
Yao, T., 2018: Environmental isotope data of the Tibetan Plateau (2005–2009). National Tibetan Plateau Data Center, accessed 14 October 2018, https://doi.org/10.11888/AtmosphericEnvironment.tpe.249292.db.
Yao, T., and Coauthors, 2013: A review of climatic controls on δ18O in precipitation over the Tibetan Plateau: Observations and simulations. Rev. Geophys., 51, 525–548, https://doi.org/10.1002/rog.20023.
Ye, T., and Coauthors, 2019: Improved population mapping for China using remotely sensed and points-of-interest data within a random forests model. Sci. Total Environ., 658, 936–946, https://doi.org/10.1016/j.scitotenv.2018.12.276.
Yu, W., and Coauthors, 2016: Short-term variability in the dates of the Indian monsoon onset and retreat on the southern and northern slopes of the central Himalayas as determined by precipitation stable isotopes. Climate Dyn., 47, 159–172, https://doi.org/10.1007/s00382-015-2829-1.
Zhan, L., J. Chen, C. Zhang, T. Wang, P. Xin, and L. Li, 2020: Fog interception maintains a major waterfall landscape in southwest China revealed by isotopic signatures. Water Resour. Res., 56, e2019WR025628, https://doi.org/10.1029/2019WR025628.
Zhang, J., and Coauthors, 2021: Coupled effects of moisture transport pathway and convection on stable isotopes in precipitation across the East Asian monsoon region: Implications for paleoclimate reconstruction. J. Climate, 34, 9811–9822, https://doi.org/10.1175/JCLI-D-21-0271.1.
Zhang, M., and S. Wang, 2016: A review of precipitation isotope studies in China: Basic pattern and hydrological process. J. Geogr. Sci., 26, 921–938, https://doi.org/10.1007/s11442-016-1307-y.
Zhang, M., and S. Wang, 2018: Precipitation isotopes in the Tianshan Mountains as a key to water cycle in arid central Asia. Sci. Cold Arid. Reg., 10, 27–37, https://doi.org/10.3724/SP.J.1226.2018.00027.
Zhang, S., 1989: Characteristics of hydrogen-oxygen isotope compositions of contemporarily atmospheric sedimentation in Shaanxi Province. Geol. Shaanxi, 7, 57–66, https://cnki.net/kcms/detail/detail.aspx?dbcode=CJFD&filename=SXDY198902005.
Zhang, Y., Y. Ren, G. Ren, and G. Wang, 2020: Precipitation trends over mainland China from 1961–2016 after removal of measurement biases. J. Geophys. Res. Atmos., 125, e2019JD031728, https://doi.org/10.1029/2019JD031728.
Zhao, J., B. Wei, and S. Xiao, 2009: Stable isotopic characteristics of atmospheric precipitation in Yichang, Hubei. Trop. Geogr., 29, 526–531, https://doi.org/10.3969/j.issn.1001-5221.2009.06.004.
Zhao, L., L. Yin, H. Xiao, G. Cheng, M. Zhou, Y. Yang, C. Li, and J. Zhou, 2011: Isotopic evidence for the moisture origin and composition of surface runoff in the headwaters of the Heihe River basin. Chin. Sci. Bull., 56, 406–415, https://doi.org/10.1007/s11434-010-4278-x.
Zhao, L., X. Liu, N. Wang, Y. Kong, Y. Song, Z. He, Q. Liu, and L. Wang, 2019: Contribution of recycled moisture to local precipitation in the inland Heihe River Basin. Agric. For. Meteor., 271, 316–335, https://doi.org/10.1016/j.agrformet.2019.03.014.
Zhao, P., 2018: Stable isotopic characteristics and influencing factors in precipitation in the monsoon region of Northern China. M.S. thesis, College of Urban and Environmental Sciences, Northwest University, 61 pp., http://cdmd.cnki.com.cn/Article/CDMD-10697-1018108988.htm.
Zhao, P., L. Tan, P. Zhang, S. Wang, B. Cui, D. Li, G. Xue, and X. Cheng, 2018: Stable isotopic characteristics and influencing factors in precipitation in the monsoon marginal region of northern China. Atmosphere, 9, 97, https://doi.org/10.3390/atmos9030097.
Zhao, P., Z. Guo, D. She, and X. Tang, 2019: Spatial distribution of the oxygen-18 in precipitation in China based on a new empirical model. J. Mt. Sci., 16, 2605–2614, https://doi.org/10.1007/s11629-019-5514-8.
Zhao, S., 1983: A new scheme for comprehensive physical regionalization in China. Acta Geogr. Sin., 38 (1), 1–10, https://doi.org/10.11821/xb198301001.
Zhao, W., 2017: Study on the isotopes and moisture source in precipitation in the Shule River Basin. Doctoral dissertation, Lanzhou University, 95 pp., http://cdmd.cnki.com.cn/Article/CDMD-10730-1017714128.htm.
Zheng, D., Q. Yang, and S. Wu, 2015: General Introduction to Physical Geography of China. Science Press, 767 pp.
Zheng, S., F. Hou, and B. Ni, 1983: Research on stable hydrogen and oxygen isotopes in precipitation of China. Chin. Sci. Bull., 28, 801–806, https://doi.org/10.1360/csb1983-28-13-801.
Zhou, H., X. Zhang, T. Yao, M. Hua, X. Wang, Z. Rao, and X. He, 2019: Variation of δ18O in precipitation and its response to upstream atmospheric convection and rainout: A case study of Changsha station, south-central China. Sci. Total Environ., 659, 1199–1208, https://doi.org/10.1016/j.scitotenv.2018.12.396.
Zhou, J. L., and T. Y. Li, 2018: A tentative study of the relationship between annual δ18O and δD variations of precipitation and atmospheric circulations—A case from Southwest China. Quat. Int., 479, 117–127, https://doi.org/10.1016/j.quaint.2017.05.038.
Zhu, G., H. Guo, D. Qin, H. Pan, Y. Zhang, W. Jia, and X. Ma, 2019: Contribution of recycled moisture to precipitation in the monsoon marginal zone: Estimate based on stable isotope data. J. Hydrol., 569, 423–435, https://doi.org/10.1016/j.jhydrol.2018.12.014.
Zhu, X., and Coauthors, 2019: Exploring the contribution of precipitation to water within the active layer during the thawing period in the permafrost regions of central Qinghai-Tibet Plateau by stable isotopic tracing. Sci. Total Environ., 661, 630–644, https://doi.org/10.1016/j.scitotenv.2019.01.064.
Zhu, X., and Coauthors, 2020: Long-distance atmospheric moisture dominates water budget in permafrost regions of the central Qinghai-Tibet Plateau. Hydrol. Processes, 34, 4280–4294, https://doi.org/10.1002/hyp.13871.
