Abstract

Drought frequently occurs in North China and is one of the most damaging disasters in this region, and drought also brings considerable challenges the world-famous South-to-North Water Diversion (SNWD) project. However, it is difficult to identify the drought-induced water deficit at a regional scale. Gravity Recovery and Climate Experiment (GRACE) satellites monitor temporal variations in the Earth's gravitational potential and provide quality data sets for water storage analysis. In this study, we quantify the water deficit over North China by focusing on a recent drought event, the 2009/10 drought, and identifying its onset, persistence, and recovery. The results indicate that GRACE can successfully capture temporal variations in total water storage (TWS). TWS shows a declining trend, reaching a low point during the 2009/10 drought with a water storage deficit of up to 25 km3 (∼22 mm). Groundwater storage shows a similar pattern, with a trend of −6.97 mm/yr. Together with the water deficit, vegetation growth is substantially restricted, as indicated by a reduction in the leaf area index. The amount of water transferred by the SNWD project may ease the water crises in North China.

INTRODUCTION

The global climate system has significantly changed in recent years, leading to an increased frequency of extreme weather and other disaster events (Leng et al. 2015; Qin et al. 2015). As a typical weather-related phenomenon, drought causes various problems such as the shortage of water resources (Lehner et al. 2006), crop damage (Deng et al. 2011), and ecological deterioration (Lewis et al. 2011), thereby imposing a direct threat to long-term security and social stability. Recently, drought has become one of the dominant factors limiting regional economic and social developments under the combined impacts of climate change and intensified human activities (Feng et al. 2014). With increasing water demand, population explosion, and uncertain water supply in the context of climate change, drought is expected to become more frequent and severe (Smith & Katz 2013). Therefore, it is imperative to pay special attention to drought events.

Drought frequently occurs in most areas of China and accounts for 35% of all economic losses from disasters (Gao & Yang 2009; Ye et al. 2012). North China is an area with the most severe water shortage, particularly in arid and semi-arid regions (Feng et al. 2014). It has long been a center of population, industry, and agriculture, leading to the over-exploitation of groundwater. The per-capita water resources have inevitably kept falling. The problems of water shortage have arisen in part because of agricultural intensification in northern China, but mostly because of rapidly growing urban and industrial demand (Webber et al. 2008). Moreover, this area has shown significant sensitivity to drought events (Wei et al. 2003; Ju et al. 2006) which exacerbates the water shortage.

To ease the growing water shortage, China implemented the South-to-North Water Diversion (SNWD) project by transferring water from the Yangtze River to the Hai, Huai and Yellow River basins (Chen et al. 2013; Rogers et al. 2016). The middle and the eastern routes have been in service and provide water to hundreds of millions of people in North China. However, the success of the SNWD project is still being debated due to its high cost of reservoir and canal construction and the social and ecological effects (Barnett et al. 2015; Rogers et al. 2016). Moreover, the operation of the SNWD project faces substantial challenges due to climate change (Li et al. 2015; Zou et al. 2016). Although the primary goal of the SNWD is to supply water mainly for urban and industrial sectors, it is flexible to allocate a considerable amount of water for agriculture and ecosystems when extreme events occur (Liu & Zheng 2002). Therefore, further demonstration and research is needed to evaluate its resilience to frequent extreme events, such as droughts.

During 2009/2010, a large-scale significant drought swept across North China, causing a serious water shortage in industry and agriculture as well as restrictions on vegetation growth (Barriopedro et al. 2012). The 2009/10 drought seriously affected both human and livestock populations due to drinking water shortage. A few studies have focused on the drought in terms of meteorology, ecology, and economy. For example, Gao & Yang (2009) indicated that the La Niña event of 2008–2009 increased the differences in temperature and atmospheric pressure between the Indo-Pacific Oceans and the Asian continent, causing severe winter-time droughts in northern China. The drought might be the main driving force behind the decreasing trend in vegetation activity in North China: the summer droughts in 2007 and 2009 reduced the vegetation cover by more than 13% (Wu et al. 2014). Moreover, the drought led to price fluctuation of agricultural products in North China, despite the minor impact on the main agricultural products (Lin et al. 2013).

However, few of these studies have investigated this drought event from a hydrological perspective. The state of water storage in an area of interest is a direct hydrological response to the degree of drought, and water storage anomalies can affect the hydrological cycle (Li et al. 2012). Regional-scale water storage can be well quantified using data from the Gravity Recovery and Climate Experiment (GRACE). GRACE employs twin satellites to monitor the variation of Earth gravity and the temporal change of Earth's gravity field is assumed to be dominated by terrestrial water storage change. GRACE data have been successfully applied for water resources analysis in many areas such as central North America (Wang et al. 2012) and North China (Feng et al. 2013).

In this study, we aim to explore the drought condition of North China during the past decade, especially focusing on the 2009/10 drought, and to discuss whether GRACE can capture ‘typical’ drought conditions in North China. We identify the response of surface water, groundwater, soil moisture, and vegetation to the drought. Moreover, we roughly evaluate the amount of water transferred by the SNWD in remediation of the drought.

This paper is organized as follows. The section below describes the study area, data sets, and methods. This is followed by a section presenting the results for precipitation anomalies and temporal and spatial changes in water storage. Response of vegetation of coverage to the 2009/10 drought is also described. A discussion section follows and finally conclusions are drawn.

DATA AND METHODS

Study area

The region of interest in this study is North China (Figure 1), which frequently experiences drought events. North China covers an area of about 1.16 million km2, is located in the region between 35–45° N and 110–125° E, and has a climate predominantly influenced by the Asian monsoon. This region is in a semi-arid environment with annual precipitation of around 500 mm (Figure 1), with most precipitation occurring in summer, annual relative humidity of 53.6%, and wind speed of 2.9 m s−1 (Feng et al. 2012). North China is an important area of grain production (Barriopedro et al. 2012); the main land cover (39.5%) is cropland, with 33.6% grassland and 18.1% forest. Agricultural irrigation in the region is heavily reliant on groundwater.

Figure 1

Long-time mean annual precipitation of North China and the distribution of ground measurements for the groundwater level (triangles). The location of the Hai River (HR) basin and Liao River (LH) basin is shown in the inset.

Figure 1

Long-time mean annual precipitation of North China and the distribution of ground measurements for the groundwater level (triangles). The location of the Hai River (HR) basin and Liao River (LH) basin is shown in the inset.

North China contains plains, mountains, and plateaus, with a declining slope from northwest to southeast. The Inner Mongolian Plateau and the Tai-hang Mountain lie in the north and west of the area; the North China plain is in the center and southeast. The area contains drought-prone basins, i.e., the Hai River (HR) basin and part of the Yellow River basin (Qin et al. 2015). Due to the large population (∼168 million), the average per capita water resource is only 23% of the Chinese average. In the North China plain, more than 70% of fresh water comes from groundwater (Zheng et al. 2010), which means that groundwater plays an important role in local normal life, agriculture, and industry. Because of the uneven spatial–temporal distribution of water resources, the economic losses and ecological disruption caused by drought events can be more severe than in other regions.

Data sets

The data sets used in this study include GRACE data, simulation data from land surface models, and ground-based measurements. GRACE data have been successfully used to detect water storage changes, but an evaluation of this data source is required for a specific area of interest (Long et al. 2013). We employed simulation data from two land surface models (i.e., variable infiltration capacity (VIC) and NOAH (NCEP, OSU, Air Force and Office of Hydrology Land Surface Model)) and evaluated the water storage changes calculated from the GRACE and the simulations. Moreover, we isolated groundwater storage variations from GRACE data to detect groundwater change, which were also validated using ground-based measurements of groundwater levels.

GRACE data

The GRACE satellite mission was launched by the National Aeronautics and Space Administration (NASA) and the German Aerospace Center in March 2002. The GRACE project monitors temporal variations in the Earth's gravitational potential. After atmospheric and oceanic effects have been accounted for, the remaining signal on monthly to inter-annual timescales is mostly related to variations in terrestrial water storage. Although its spatial resolution (∼150,000 km2) and temporal resolution (10-day to monthly) are low in comparison with other satellites, GRACE has the attractive advantage that it senses water stored at all levels, including groundwater (Rodell et al. 2009). Many studies have evaluated the use of GRACE satellites to monitor the hydrologic impacts of droughts (Long et al. 2013) and long-term total water changes.

The Level-3 GRACE product used in this study was processed by the University of Texas Center for Space Research (CSR) using a Gaussian filter with a 300 km smoothing radius to remove the stripes observed in the spherical harmonic coefficient fields. Data from the German Research Centre for Geosciences (GFZ) and the NASA Jet Propulsion Laboratory (JPL) (http://grace.jpl.nasa.gov/data/) were also used. Atmospheric and oceanic circulations had already been removed from mass distributions, and a correction had been made (Houborg et al. 2010). Our GRACE time series are monthly data for the period from January 2003 to December 2012. Anomalous fields were obtained by subtracting out the multi-year mean field and converted to equivalent water heights including changes regarding surface water, soil moisture, and groundwater, with a spatial resolution of 1°. We also isolated groundwater changes by subtracting the soil moisture and canopy storage changes from the total water storage (TWS) anomalies (Castle et al. 2014) to compare with the groundwater storage change (GWC).

Simulation data

To evaluate the dryness of the 2009/10 drought and to validate the terrestrial water storage measurements of GRACE, water fluxes (i.e., runoff and evapotranspiration) and soil moisture from two land surface models were used in this study. The first is the VIC model (Liang et al. 1994). In this study, the VIC daily simulation data at 0.25° resolution were obtained from Zhang et al. (2014), who produced a long-term hydrological data set for China especially. The simulation has been successfully calibrated and validated using ground-measured streamflow and soil moisture (Xie et al. 2015), and remote-sensing evapotranspiration (Zhang et al. 2014).

To perform more extensive evaluations, we also used the NOAH simulated hydrological data from the Global Land Data Assimilation System (GLDAS) (Rodell et al. 2004). The NOAH model has more than a 30-year history (Chen et al. 1996) and has undergone continuous improvement (Fang et al. 2009; Ingwersen et al. 2011). To verify the GRACE measurements, we used the NOAH simulated data from GLDAS because the data were widely applied (Syed et al. 2008; Rodell et al. 2009; Long et al. 2013) and they have also been evaluated in North China with acceptable uncertainties (Feng et al. 2013; Huang et al. 2015).

Please note the VIC and the NOAH simulation data of water fluxes and soil moisture were from other studies, and we neither performed the simulations nor evaluated the data. Output daily data at 0.25° resolution were aggregated to monthly and one-degree scale to be compared with GRACE.

Ground-based measurements and other data

In this study, ground-based measurements of precipitation, groundwater, and surface water storage were used. Ground-based measured precipitation data from the Chinese Meteorological Administration were applied to derive gridded precipitation at a spatial resolution of 0.25°, and the gridded precipitation data have been extensively verified for runoff, evapotranspiration, and soil moisture (Zhang et al. 2014). These gridded precipitation data can be used to identify the spatial coverage of meteorological droughts.

In order to detect the impact of the drought on the groundwater system, groundwater level observations were acquired from 95 observation wells. The distribution of these wells is shown in Figure 1. Reservoir storage constitutes a major part of surface water, so water stored in reservoirs in the HR basin from 2003 to 2012 in the Hai River Water Resources Bulletin (HRWRB) were also used to examine this drought. Moreover, the data of annual groundwater withdrawal from the HRWRB were applied to reflect the human impacts on groundwater storage.

METHODS

Figure 2 provides a flowchart to depict the methods and data used in this study. Drought is often caused by insufficient precipitation, so we first characterized the 2009/10 drought over a longer perspective based on the 53-year (from 1960 to 2012) precipitation. The Standardized Precipitation Index (SPI) and the probability of yearly precipitation are used to represent the status of the drought in the 53 years. Then, we identified the water storage condition, including the TWS, surface water, and groundwater. In order to evaluate the GRACE data, we compared water storage changes from GRACE and the simulation data. Moreover, the groundwater storage calculated from GRACE was also evaluated using in situ observations. Here, we especially present the methods used to calculate the SPI, water storage changes, and groundwater storage.

Figure 2

Flowchart of methods and data used in this study.

Figure 2

Flowchart of methods and data used in this study.

SPI

The severity of a drought can be quantified with a drought index. The SPI was used to reflect the meteorological drought, which was proposed by McKee et al. (1993) and is a widely used drought index. The index is a statistical monthly indicator that compares the accumulated precipitation during a period of specific months with the long-term cumulative rainfall distribution for an accumulated period (Nam et al. 2015). The timescales of SPI vary from 1 month to 24 months. Small timescale SPI (1 or 6 months) has quick fluctuations (McKee et al. 1993).

Total water storage changes

As is the case with many other satellite data sources, uncertainties in GRACE are inevitably caused by atmosphere, sensor, and other factors. The GRACE data therefore need evaluation for the area of interest. We calculated the monthly TWS changes () from two independent sources: the model simulations (i.e., from NOAH and VIC) and the GRACE data (Famiglietti et al. 2011). As the GRACE monthly data represent the mass anomaly, the difference of the GRACE anomaly in two successive months is equivalent to the monthly water storage changes (Wang et al. 2014): 
formula
(1)
where the subscript i stands for the ith month and represents the ith month TWS anomaly.
With the model simulation data (from NOAH and VIC), the water storage changes can be computed based on the monthly basin-scale water balance (Syed et al. 2008): 
formula
(2)
where P, E, and R denote precipitation, evapotranspiration, and runoff, respectively.

Therefore, the agreement of water storage changes calculated from Equations (1) and (2) is a useful indicator for the accuracy of GRACE in capturing the TWS change, because the model simulation and GRACE are independent approaches (Syed et al. 2008). Please note there is an optional approach for the evaluation of GRACE: TWS is equal to the sum of groundwater storage, surface water storage, and soil moisture storage (Famiglietti et al. 2011). In contrast to this approach, Equation (2) results in less uncertainty because it requires data only from land surface modeling and the associated fluxes (i.e., P, E, and R) are from the same modeling. However, the land surface modeling may still hold uncertainty, and we will return to this in the section ‘Uncertainties’.

Groundwater storage

Groundwater is an important part of TWS in North China. To detect groundwater storage changes during recent years, the storage variation is discussed. There are two methods for calculation of GWC. The first method is based on ground measurement by multiplying the measured groundwater level anomalies by the specific yield of each well (Huang et al. 2015): 
formula
(3)
where Hi represents the groundwater level measured in situ for the ith month and stands for the specific yield. In this study, the value of for each site was based on the specific yield map according to Huang et al. (2015).
The other method for GWC computation is subtraction of soil water storage from the GRACE TWS changes: 
formula
(4)
where G is the GWC, S and M denote the GRACE total water anomalies and the soil moisture changes simulated by the hydrologic model, respectively. The C and W represent canopy water storage and surface water (i.e., water storage in reservoirs), respectively. Both C and W were ignored in this study as they do not have obvious changes comparing with S and M. The surface water storage change will be presented in the section ‘Surface and groundwater storage’.

Through the two methods, groundwater storage is obtained, against which the GRACE data can be compared and the TWS changes calculated.

RESULTS

The precipitation extreme of the 2009/10 droughts

Precipitation is a direct indicator of drought. We used monthly precipitation data to analyze the water balance input during 2009 and 2010 (Figure 3) and diagnosed the dryness. As illustrated in Figure 3, the regional average accumulated precipitation is less than the climatological mean values calculated for the period 1960–2012. Especially in the summer and the fall of 2009, the precipitation only accounts for 78% of the climatological mean. The spring of 2010 is slightly wet due to a near-normal monsoon season (Barriopedro et al. 2012). The regional precipitation deficit reaches 14 mm throughout 2009/10 and 47 mm from May 2009 to April 2010.

Figure 3

(a) Accumulated monthly precipitation during 2009/10 compared with the climatological mean; (b) monthly departure from the climatological mean.

Figure 3

(a) Accumulated monthly precipitation during 2009/10 compared with the climatological mean; (b) monthly departure from the climatological mean.

To characterize the severity of this drought, 53-year monthly precipitation data (from 1960 to 2012) were used to calculate the SPI. Three timescales of SPI are shown in Figure 4(a), indicating different drought situations. Meteorological and soil moisture conditions respond to precipitation anomalies on relatively short timescales, whereas streamflow, reservoirs, and groundwater respond to long-term precipitation anomalies on the order of 6 to 24 months or longer. According to the SPI classification (Nam et al. 2015; Qin et al. 2015), the 12-month SPI (approximately −1.0) indicates a moderate drought during May 2009 to April 2010, the 1-month SPI represents a severe drought in August and October 2009, and the 6-month SPI indicates a severe drought from October to December 2009 with the lowest SPI value of approximately −1.63. Overall, there is an obvious drought event in North China from May 2009 to April 2010.

Figure 4

(a) SPI on three timescales (1 month, 6 months, and 12 months) for 2003–2012; (b) probability of the hydrological year's precipitation. Bars are the rank of the annual precipitation for 1960–2012.

Figure 4

(a) SPI on three timescales (1 month, 6 months, and 12 months) for 2003–2012; (b) probability of the hydrological year's precipitation. Bars are the rank of the annual precipitation for 1960–2012.

In addition to the SPI, the probability of yearly precipitation can also reflect the water input conditions with respect to North China. To compute the probability, we first defined the hydrological year as being the period between this May and the following April. We sorted the 53 years of precipitation from high to low and calculated the probability of each year using the Weibull equation (Helsel 2002). As shown in Figure 4(b), the precipitation of 2009 was ranked 43rd, and the probability of precipitation during this drought period was only about 84%, indicating that 2009 was a severely dry episode during the 53 years, which is consistent with the SPI results.

Response of water storage changes to the extreme drought

Total water storage

The lack of water input (i.e., precipitation) during the drought period likely induces a decrease in water storage. The GRACE-derived water storage anomaly is the result of water storage relative to the long-term average value. As shown in Figure 5(a), the GRACE data from CSR, JPL, and GFZ have similar trends and match quite well. Overall, there is a notable decrease of TWS in North China from 2003 to 2013, indicating recurrence of the drought. The TWS anomalies in 2009 and 2010 are below zero with a mean value of approximately −21 mm and a minimum value of −40 mm, which means that water storage is less than normal. The storage shows a small increase in the winter of 2009 and spring of 2010, and this trend is consistent with the precipitation change.

Figure 5

(a) TWS anomalies in North China from 2003 to 2012; (b) comparison of water storage changes from GRACE (average of CSR, GFZ, and JPL), VIC, and NOAH.

Figure 5

(a) TWS anomalies in North China from 2003 to 2012; (b) comparison of water storage changes from GRACE (average of CSR, GFZ, and JPL), VIC, and NOAH.

There will be uncertainties in the GRACE data, so we verified the data by comparing with the TWS changes from the NOAH and VIC simulations. To make the comparison, the average GRACE values from CSR, JPL, and GFZ were computed. From Figure 5(b), the series of GRACE agrees well with the values from VIC and NOAH, although of GRACE displays larger amplitude. The correlation coefficient between GRACE and NOAH is 0.53 and the correlation of GRACE with VIC is 0.52, whereas the correlation between VIC and NOAH is about 0.85, suggesting a certain degree of consistency between the three sources of data. The smaller magnitude of VIC likely results from the difference in the input forcing data as well as the model formulations. Specifically, the input forcing data for VIC were created by interpolating ground observations (Zhang et al. 2014), which is different from the data for NOAH in GLDAS. However, the two models (NOAH and VIC) give similar pattern for with a high correlation. The outputs from the two models show lower correlation with GRACE. The reason is mainly due to the fact that both models lack explicit groundwater formulation. GRACE captures TWS variations including surface water and groundwater. Moreover, the two models are inclined rather for natural-world hydrology simulations, while the GRACE data contain human activity impacts (e.g., irrigation and withdraw groundwater).

The spatial distributions of TWS anomalies for this drought event are presented in Figure 6. From May 2009 to April 2010, the southern regions including Shanxi, Shandong, and Hebei provinces suffered a much more severe drought than the north. Although the spatial distribution is uneven, TWS is still below zero and the south of North China is the main affected area.

Figure 6

Spatial distributions of TWS anomalies between May 2009 and April 2010.

Figure 6

Spatial distributions of TWS anomalies between May 2009 and April 2010.

Furthermore, we computed water storage deficit for 2009/10. First, a 120-month climatology (January 2003 to December 2012) was computed based on the GRACE TWS changes time series by averaging the TWS changes values of each month of the GRACE record. This climatology (in unit of mm) represents the characteristic variability of water storage and serves as normal water storage conditions (Thomas et al. 2014). Second, water storage deficit in 2009/10 is defined as the volume of water required to return to normal water storage conditions. From Figure 7, we can see that drought events mainly occur in the south of North China, where the water resources storage deficits are apparently less than other regions. The regional average water storage deficits are up to 22 mm, about 25.5 km3 relative to the normal water storage condition.

Figure 7

TWS deficit relative to the normal water storage conditions during the 2009/10 drought (from May 2009 to April 2010). The dotted line shows the seriously dry area.

Figure 7

TWS deficit relative to the normal water storage conditions during the 2009/10 drought (from May 2009 to April 2010). The dotted line shows the seriously dry area.

Surface and groundwater storage

Due to data availability, data for yearly reservoir storage were used to reflect surface water storage. According to the Water Resources Bulletin of Hai River Basin (http://www.hwcc.gov.cn/), the number of reservoirs slightly increased from 137 in 2003 to 146 in 2012, so the TWS of reservoirs increased from 61.1 km3 in 2003 to 95.81 km3 in 2012 (Figure 8). To derive the surface water storage changes, we used the average storage of the reservoirs. Long-term average water storage is about 0.16 mm, but the storage reaches its lowest levels in 2009 (∼0.13 mm) and 2010 (∼0.14 mm), reflecting the influence of the drought.

Figure 8

Surface water storage (bars) and equivalent water thickness changes (line).

Figure 8

Surface water storage (bars) and equivalent water thickness changes (line).

Groundwater is a vital source of fresh water for agriculture, industry, domestic use, and ecosystems in North China (Feng et al. 2013). To quantify the influence of droughts on groundwater storage, in addition to the GRACE data, we used ground observations from 95 wells. Figure 9(a) presents the average variations of groundwater levels of the 95 wells. There is a gradual decline of approximately −0.41 m/yr, despite substantial uncertainties. For the 95 wells, trends in the groundwater level range from −2.5 to 2.0 m/yr, and the decreases mainly appear in the south of North China (Figure 9(b)).

Figure 9

(a) Groundwater level changes from 2005 to 2013 in North China. The shaded area shows the uncertainties (95% confidence intervals). (b) The trend of the groundwater level for each gauge.

Figure 9

(a) Groundwater level changes from 2005 to 2013 in North China. The shaded area shows the uncertainties (95% confidence intervals). (b) The trend of the groundwater level for each gauge.

Figure 10(a) shows the GWC derived from the in situ observations and GRACE over North China, and groundwater storage is described as the equivalent water height. Both of these data sets indicate a decreasing trend, of −4.68 mm/yr for GRACE and −6.97 mm/yr for ground observations. This difference may be attributable to the uncertainties within GRACE and ground observations and the spatial representation of the 95 ground observations. The leakage errors and measurements error are also plotted in the shaded areas in Figure 10(a). Despite such differences, the changes in groundwater storage from GRACE and ground observations have a strong correlation, with a Pearson correlation coefficient of 0.71.

Figure 10

Groundwater storage changes derived from GRACE and ground observations: (a) North China; (b) HR basin; and (c) LH basin. The shaded area shows the estimation errors for the GRACE-derived groundwater changes.

Figure 10

Groundwater storage changes derived from GRACE and ground observations: (a) North China; (b) HR basin; and (c) LH basin. The shaded area shows the estimation errors for the GRACE-derived groundwater changes.

Besides, as there are many in situ groundwater (GW)-level measurements located in the HR basin and the Liao River (LH) basin, we separately estimated the in situ and GRACE-derived groundwater storage (GWS) anomalies over the two subregions (Figure 10(b) and 10(c)). The results show that groundwater storage in the HR basin has decreased significantly with a trend of −9.02 mm/yr from GRACE-derived and −14.8 mm/yr from groundwater observations. While the LH basin has a less pronounced annual trend of groundwater storage, the groundwater anomalies are negative values during 2009/10 which may be caused by this severe drought event.

Response of soil moisture and vegetation coverage to the extreme drought

In addition to water storage depletion, the 2009/10 drought caused negative impacts on vegetation growth (Wang et al. 2015; Zhang et al. 2016). Wu et al. (2014) indicated that this drought probably reduced the normalized difference vegetation index by 6.68% in 2009 in the Beijing–Tianjin sand source region.

To further investigate the impact of this drought, we calculated the average leaf area index (LAI) within the growing season (from May to October) for three types of land cover (grass, crop, and forest), as LAI is an important indicator of crop growth and plant productivity (Liang et al. 2015). As shown in Figure 11(c), LAI reaches its lowest level during 2009. Especially for crop land, LAI in 2009 is less than its multi-year mean of approximately 0.11. An area of more than 0.3 million km2 of North China shows a substantial LAI reduction.

Figure 11

(a) LAI reduction in 2009 (A, B, and C stand for three different parts of North China); (b) soil moisture deficits in 2009; (c) time series of LAI corresponding to three types of land cover; and (d) soil moisture and LAI during the plant growth season (May–October).

Figure 11

(a) LAI reduction in 2009 (A, B, and C stand for three different parts of North China); (b) soil moisture deficits in 2009; (c) time series of LAI corresponding to three types of land cover; and (d) soil moisture and LAI during the plant growth season (May–October).

Vegetation growth is more sensitive to soil moisture storage than the TWS generally, so we analyzed the correlation between LAI and soil moisture in space and time. The reductions of LAI and soil moisture were calculated. As shown in Figure 11(a) and 11(b), LAI and soil moisture have similar distribution in space. Both show significant reductions in the northwest of North China (Part A), but have minor increases in Part B. Their distributions are not so consistent in the southeast (Part C) where there are intensive human activities (e.g., crop cultivation and groundwater extraction). Overall, LAI and soil moisture storage show great reductions in the 2009/10 drought event. Figure 11(d) presents the time series of the LAI and soil moisture change during the plant growth season (May–October). We can see that the change of LAI agrees well with soil moisture in time, and their Pearson correlation coefficient is 0.74. Moreover, both of them reach the low points in 2009. Therefore, the LAI reduction is consistent with the distribution of soil moisture to some degree in space and time.

Thus, vegetation growth is substantially constrained during this drought, consequently reducing agricultural production.

Implications for the SNWD project

The SNWD project supplies water resources from the Yangtze River basin to North China, and it is expected to transfer approximately 27.8 km3 of water per year (http://www.nsbd.gov.cn/zx/gczz/201106/t20110630_188241.html). A large part of water transferred by SNWD goes to urban areas for domestic and industrial uses, and a small proportion is being directly supplied to agriculture and ecosystem use. The original plan of the SNWD project was not based on frequent droughts, but it can essentially recharge water storage in North China if a drought event occurs (Liu & Zheng 2002). Moreover, the SNWD project is expected to effectively reduce the exploitation of local water resources (Zou et al. 2016) and to reserve water for agricultural irrigation during a drought.

In this study, we demonstrated that the 2009/10 drought was a severe episode with precipitation ranking 84%, and the water storage deficit being about 22 mm (∼25 km3). Therefore, the SNWD project can probably replenish the water deficit at this level of drought and relieve the drought impact to a certain degree. Certainly, the efficiency of the SNWD in combating drought will depend on the water configuration strategy. However, the amount of water transfer by the SNWD is not a constant; it depends on precondition of water resource regions and requirement of receiving water regions. During the summer monsoon rainy season in South China, the SNWD is expected to provide a large amount of water resources to replenish the surface water and groundwater storage in North China when a drought event occurs. Combining with the water configuration strategy for the SNWD, local water management is essential to alleviate water shortage (Barnett et al. 2015).

In relieving the stress of water resources and combating extreme weather events, the SNWD project requires additional evaluations of water quality regarding surface and groundwater and the effect on ecosystems (Zhu et al. 2008; Tang et al. 2014). The water quality in the source water for the SNWD may run the risk of failing to meet the required standard (Xin et al. 2015). Moreover, the eastern route of the SNWD will probably cause the discharge into the estuary to fall below critical levels in some months and future water transfer projects along the lower Yangtze River will further compound the problem (Chen et al. 2013).

DISCUSSION

Further evidence and impact of the drought

Climate change in North China during past decades can be characterized as an increase in air temperature and a decrease in precipitation (Ming et al. 2015). Moreover, the frequency and intensity of drought over North China has significantly increased during the last five decades (Qin et al. 2015), mainly caused by the dramatic decrease in precipitation (Xu et al. 2015). In this study, we focus on the 2009/10 drought event in the context of the environmental changes in the past decade. Given the SPI values and the probability of precipitation, this drought was a severe event. The drought started in May 2009 and ended in April 2010, as shown by Barriopedro et al. (2012). In contrast to existing studies focusing on the drought from the viewpoint of meteorology or ecology, we addressed this drought event from a hydrological perspective in order to analyze the influence on water storage, which is essential for ecosystem and agricultural production.

With decreasing precipitation, water storage depletion has taken place during the past decade in North China (Moiwo et al. 2013). In this study, we found that surface water storage reached a low level in 2009 and 2010. The responsiveness of the groundwater system is important for hydrological drought development (Van Loon & Laaha 2015). The groundwater level has been declining at a rate of ∼0.3 m/yr since 1960 (Cao et al. 2016).

Human activities have a significant influence on water storage change in North China, such as urban water use and irrigation. Groundwater recharge depends primarily on the amount of irrigation water pumped from aquifers, and the net water use (recharge minus irrigation) is responsible for the decline of the groundwater level (Kropp et al. 2015). Figure 12 shows total groundwater withdrawals from 2003 to 2012. Although the groundwater withdrawals have continuously decreased during the past decade (∼0.6 km3/yr), they primarily contributed to the groundwater decline in North China, because there is no significant increase trend in the net recharge (Figure 5(b)). Similar results were also shown in Zheng et al. (2010). However, the water deficit during the 2009/10 drought is dominated by inadequate precipitation input, so that groundwater storage remains at a low level during the period (Figure 10). Moreover, our study shows that the rate of groundwater level decline is approximately 0.41 m/yr from 2005 to 2013, besides which the depletion rate of groundwater has been accelerating during past decades globally (Wada et al. 2010), which may be attributable to the groundwater overuse and the reoccurrence of drought events.

Figure 12

Groundwater withdrawals from 2003 to 2012 in the HR basin.

Figure 12

Groundwater withdrawals from 2003 to 2012 in the HR basin.

Uncertainties

North China is an interesting area for many researchers for its special climate features. Although a few studies have used GRACE data to detect groundwater depletion (Feng et al. 2013) or long-term changes of TWS, this is the first study to quantify the water deficit during a typical drought (2009/10) over North China. We only focused on the specific period to explore thoroughly water storage changes as well as vegetation response to drought event. Quantifying the drought severity and the spatial and temporal distribution of water storage deficit have valuable implications for evaluating the world-famous SNWD Project.

As mentioned above, Level-3 gridded GRACE data were adopted in this study, but these data may have bias and leakage errors. Newly released CSR and JPL mascon solutions may provide advantages relative to the spherical harmonics, including reduced leakage from land to ocean, increased signal amplitudes, and easily empirical postprocessing for hydrologists (Scanlon et al. 2016). Like many other studies (Syed et al. 2008; Famiglietti et al. 2011), simulations with land surface models were used to verify the GRACE data, but these simulations themselves hold uncertainties because human activities such as groundwater pumping and irrigation were not well considered in the models (Pan et al. 2017). Nevertheless, the verification in this study is acceptable because of consistent TWS changes between the simulations and GRACE.

CONCLUSIONS

In this study, the hydrological effects of the 2009/10 drought in North China are discussed using multi-source data, including satellite data, ground measurements, and model simulations. On the basis of the precipitation data, the shortage of precipitation was 47 mm from May 2009 to April 2010: this event is regarded as a severe drought on the basis of the SPI time series. Moreover, the probability of precipitation during this period was about 84% in the past 52 years, also indicating a notable drought event, consistent with the SPI analysis. There was a declining trend in TWS for the past decade based on GRACE data, and the regional deficit of water storage was approximately 22 mm (∼25 km3) in 2009/10. The relatively dry area is located in the south of North China. Furthermore, both groundwater storage and TWS have a decreasing trend, while the surface water reached its lowest level in 2009. Due to limited water availability, vegetation growth was constrained. More than a 0.3 million km2 area of North China shows a substantial LAI reduction, and this reduction is consistent with the distribution of soil moisture to some degree in space and time. Thus, this drought event has led to observable hydrological effects as well as suppression of vegetation growth in North China. The SNWD project may ease the water storage deficit in North China for this level of drought intensity.

GRACE data have attractive advantages for large-scale drought and flood-potential detection (Reager & Famiglietti 2009; Houborg et al. 2010; Li et al. 2012). However, the effective spatial resolution of GRACE is about 150,000 km2 at best (Swenson et al. 2006), so these data may not be suitable for small-scale issues. Use of multi-source data, including satellite data, ground measurement, and model simulations, is an effective strategy to quantify both drought intensity and water deficits.

The primary objective of the middle and the eastern routes of the SNWD project is to meet the demand of urban and industrial water use in North China, thereby relaxing the groundwater over-exploitation. Moreover, this project is expected to reserve relatively more water for agricultural irrigation which is assumed as the main agent for the groundwater declines. Therefore, it would be interesting to examine the effectiveness of the SNWD project in groundwater recovery and identify the benefit for the agriculture and ecosystem, especially under climate change.

ACKNOWLEDGEMENTS

This study was supported by grants from the National Natural Science Foundation of China (No. 41471019, 61661136006) and the National Key Research and Development Program of China (No. 2016YFA0600104).

REFERENCES

REFERENCES
Barnett
J.
,
Rogers
S.
,
Webber
M.
,
Finlayson
B.
&
Wang
M.
2015
Sustainability: transfer project cannot meet China's water needs
.
Nature
52
,
295
297
.
Barriopedro
D.
,
Gouveia
C. M.
,
Trigo
R. M.
&
Wang
L.
2012
The 2009/10 drought in China: possible causes and impacts on vegetation
.
Journal of Hydrometeorology
13
(
4
),
1251
1267
.
Cao
G.
,
Scanlon
B. R.
,
Han
D.
&
Zheng
C.
2016
Impacts of thickening unsaturated zone on groundwater recharge in the North China plain
.
Journal of Hydrology
537
,
260
270
.
Castle
S. L.
,
Thomas
B. F.
,
Reager
J. T.
,
Rodell
M.
,
Swenson
S. C.
&
Famiglietti
J. S.
2014
Groundwater depletion during drought threatens future water security of the Colorado River Basin
.
Geophysical Research Letters
41
(
16
),
5904
5911
.
Chen
F.
,
Mitchell
K.
,
Schaake
J.
,
Xue
Y.
,
Pan
H.
,
Koren
V.
,
Duan
Y.
,
Ek
M.
&
Betts
A.
1996
Modeling of land-surface evaporation by four schemes and comparison with FIFE observations
.
Journal of Geophysical Research
101
,
7251
7568
.
Chen
D.
,
Webber
M.
,
Finlayson
B.
,
Barnett
J.
,
Chen
Z.
&
Wang
M.
2013
The impact of water transfers from the lower Yangtze River on water security in Shanghai
.
Applied Geography
45
,
303
310
.
Deng
X. Z.
,
Lin
Y. Z.
,
Ge
Q. S.
&
Zhao
Y. H.
2011
Impact of drought on price fluctuation of agricultural production across the North China region
.
Resources Science
33
(
4
),
766
772
.
Famiglietti
J. S.
,
Lo
M.
,
Ho
S. L.
,
Bethune
J.
,
Anderson
K. J.
,
Syed
T. H.
,
Swenson
S. C.
,
de Linage
C. R.
&
Rodell
M.
2011
Satellites measure recent rates of groundwater depletion in California's Central Valley
.
Geophysical Research Letters
38
(
3
),
doi: 10.1029/2010gl046442
.
Fang
H.
,
Beaudoing
H. K.
,
Rodell
M.
,
Teng
W. L.
&
Vollmer
B. E.
2009
Global Land Data Assimilation System (GLDAS) products, services and application from NASA Hydrology Data and Information Services Center (HDISC)
. In:
ASPRS 2009 Annual Conference
,
8–31 March
,
Baltimore, MD
,
USA
.
Feng
J.
,
Yan
D.
,
Li
C.
,
Bao
S.
&
Gao
Y.
2012
Assessing the impact of climate variability on potential evapotranspiration during the past 50 years in North China
.
Journal of Food, Agriculture and Environment
10
(
3–4
),
1392
1398
.
Feng
W.
,
Zhong
M.
,
Lemoine
J.-M.
,
Biancale
R.
,
Hsu
H.-T.
&
Xia
J.
2013
Evaluation of groundwater depletion in North China using the Gravity Recovery and Climate Experiment (GRACE) data and ground-based measurements
.
Water Resources Research
49
(
4
),
2110
2118
.
Feng
J.
,
Yan
D.
,
Li
C.
,
Yu
F.
&
Zhang
C.
2014
Assessing the impact of climatic factors on potential evapotranspiration in droughts in north China
.
Quaternary International
336
,
6
12
.
Gao
H.
&
Yang
S.
2009
A severe drought event in northern China in winter 2008–2009 and the possible influences of La Niña and Tibetan Plateau
.
Journal of Geophysical Research
114
(
D24
),
10.1029/2009jd012430
.
Helsel
H. R.
2002
Statistical methods in water resources
.
Techniques of Water-Resources Investigations of the US Geological Survey
,
Chapter A3, Book 4. US Geological Survey
.
Houborg
R.
,
Rodell
M.
&
Lawrimore
J.
2010
Using Enhanced GRACE Water Storage Data to Improve Drought Detection by the U.S. and North American Drought Monitors
,
25–30 July
.
IGARSS
,
Honolulu, HI
,
USA
.
Huang
Z.
,
Pan
Y.
,
Gong
H.
,
Yeh
P. J. F.
,
Li
X.
,
Zhou
D.
&
Zhao
W.
2015
Subregional-scale groundwater depletion detected by GRACE for both shallow and deep aquifers in North China plain
.
Geophysical Research Letters
42
(
6
),
1791
1799
.
Ingwersen
J.
,
Steffens
K.
,
Högy
P.
,
Warrach-Sagi
K.
,
Zhunusbayeva
D.
,
Poltoradnev
M.
,
Gäbler
R.
,
Wizemann
H. D.
,
Fangmeier
A.
,
Wulfmeyer
V.
&
Streck
T.
2011
Comparison of Noah simulations with eddy covariance and soil water measurements at a winter wheat stand
.
Agricultural and Forest Meteorology
151
(
3
),
345
355
.
Ju
J.
,
Lu
J.
&
Ren
J.
2006
The effect of interdecadal variations of Arctic Oscillation on aridzation in North China
.
Plateau Meteorology
25
(
1
),
74
81
.
Kropp
J. P.
,
Yang
X.
,
Chen
Y.
,
Pacenka
S.
,
Gao
W.
,
Zhang
M.
,
Sui
P.
&
Steenhuis
T. S.
2015
Recharge and groundwater use in the North China plain for six irrigated crops for an eleven year period
.
Plos One
10
(
1
),
e0115269
.
Lehner
B.
,
Döll
P.
,
Alcamo
J.
,
Henrichs
T.
&
Kaspar
F.
2006
Estimating the impact of global change on flood and drought risks in Europe: a continental, integrated analysis
.
Climatic Change
75
(
3
),
273
299
.
Leng
G.
,
Tang
Q.
&
Rayburg
S.
2015
Climate change impacts on meteorological, agricultural and hydrological droughts in China
.
Global and Planetary Change
126
,
23
34
.
Lewis
S. L.
,
Brando
P. M.
,
Phillips
O. L.
,
van der Heijden
G. M. F.
&
Nepstad
D.
2011
The 2010 Amazon drought
.
Science
331
(
6017
),
554
.
Li
B.
,
Rodell
M.
,
Zaitchik
B. F.
,
Reichle
R. H.
,
Koster
R. D.
&
van Dam
T. M.
2012
Assimilation of GRACE terrestrial water storage into a land surface model: evaluation and potential value for drought monitoring in western and central Europe
.
Journal of Hydrology
446–447
,
103
115
.
Li
L.
,
Zhang
L.
,
Xia
J.
,
Gippel
C. J.
,
Wang
R.
&
Zeng
S.
2015
Implications of modelled climate and land cover changes on runoff in the middle route of the South to North water transfer project in China
.
Water Resources Management
29
(
8
),
2563
2579
.
Liang
X.
,
Lettenmaier
D. P.
,
Wood
E. F.
&
Burges
S. J.
1994
A simple hydrologically based model of land surface water and energy fluxes for general circulation models
.
Journal of Geophysical Research
99
(
D7
),
14
.
Liang
L.
,
Di
L.
,
Zhang
L.
,
Deng
M.
,
Qin
Z.
,
Zhao
S.
&
Lin
H.
2015
Estimation of crop LAI using hyperspectral vegetation indices and a hybrid inversion method
.
Remote Sensing of Environment
165
,
123
134
.
Lin
Y.
,
Deng
X.
&
Jin
Q.
2013
Economic effects of drought on agriculture in North China
.
International Journal of Disaster Risk Science
4
(
2
),
59
67
.
Liu
C.
&
Zheng
H.
2002
South-to-north water transfer schemes for China
.
International Journal of Water Resources Development
18
(
3
),
453
471
.
Long
D.
,
Scanlon
B. R.
,
Longuevergne
L.
,
Sun
A. Y.
,
Fernando
D. N.
&
Save
H.
2013
GRACE satellite monitoring of large depletion in water storage in response to the 2011 drought in Texas
.
Geophysical Research Letters
40
(
13
),
3395
3401
.
McKee
T. B.
,
Doesken
N. J.
&
Kliest
J.
1993
The relationship of drought frequency and duration to time scales
. In:
8th Conference of Applied Climatology
,
17–22 January
,
Anaheim, CA
,
USA
.
Ming
B.
,
Guo
Y.-q.
,
Tao
H.-b.
,
Liu
G.-z.
,
Li
S.-k.
&
Wang
P.
2015
SPEIPM-based research on drought impact on maize yield in North China plain
.
Journal of Integrative Agriculture
14
(
4
),
660
669
.
Nam
W.-H.
,
Hayes
M. J.
,
Svoboda
M. D.
,
Tadesse
T.
&
Wilhite
D. A.
2015
Drought hazard assessment in the context of climate change for South Korea
.
Agricultural Water Management
160
,
106
117
.
Pan
Y.
,
Zhang
C.
,
Gong
H.
,
Yeh
P. J. F.
,
Shen
Y.
,
Guo
Y.
,
Huang
Z.
&
Li
X.
2017
Detection of human-induced evapotranspiration using GRACE satellite observations in the Haihe River basin of China
.
Geophysical Research Letters
44
(
1
),
190
199
.
Reager
J. T.
&
Famiglietti
J. S.
2009
Global terrestrial water storage capacity and flood potential using GRACE
.
Geophysical Research Letters
36
(
23
),
doi: 10.1029/2009gl040826
.
Rodell
M.
,
Houser
P. R.
,
Jambor
U.
,
Gottschalck
J.
,
Mitchell
K.
,
Meng
C. J.
,
Arsenault
K.
,
Cosgrove
B.
,
Radakovich
J.
,
Bosilovich
M.
,
Entin
J. K.
,
Walker
J. P.
,
Lohmann
D.
&
Toll
D.
2004
The global land data assimilation system
.
Bulletin of the American Meteorological Society
85
(
3
),
381
394
.
Rodell
M.
,
Velicogna
I.
&
Famiglietti
J. S.
2009
Satellite-based estimates of groundwater depletion in India
.
Nature
460
(
7258
),
999
1002
.
Rogers
S.
,
Barnett
J.
,
Webber
M.
,
Finlayson
B.
&
Wang
M.
2016
Governmentality and the conduct of water: China's South-North water transfer project
.
Transactions of the Institute of British Geographers
41
(
4
),
429
441
.
Scanlon
B. R.
,
Zhang
Z.
,
Save
H.
,
Wiese
D. N.
,
Landerer
F. W.
,
Long
D.
,
Longuevergne
L.
&
Chen
J.
2016
Global evaluation of new Grace mascon products for hydrologic applications
.
Water Resources Research
52
(
12
),
9412
9429
.
Swenson
S.
,
Yeh
P. J. F.
,
Wahr
J.
&
Famiglietti
J.
2006
A comparison of terrestrial water storage variations from GRACE with in situ measurements from Illinois
.
Geophysical Research Letters
33
(
16
),
doi: 10.1029/2006gl026962
.
Syed
T. H.
,
Famiglietti
J. S.
,
Rodell
M.
,
Chen
J.
&
Wilson
C. R.
2008
Analysis of terrestrial water storage changes from GRACE and GLDAS
.
Water Resources Research
44
(
2
),
doi: 10.1029/2006wr005779
.
Thomas
A. C.
,
Reager
J. T.
,
Famiglietti
J. S.
&
Rodell
M.
2014
A GRACE-based water storage deficit approach for hydrological drought characterization
.
Geophysical Research Letters
41
(
5
),
1537
1545
.
Wada
Y.
,
van Beek
L. P. H.
,
van Kempen
C. M.
,
Reckman
J. W. T. M.
,
Vasak
S.
&
Bierkens
M. F. P.
2010
Global depletion of groundwater resources
.
Geophysical Research Letters
37
(
20
),
doi: 10.1029/2010gl044571
.
Wang
H.
,
Jia
L.
,
Steffen
H.
,
Wu
P.
,
Jiang
L.
,
Hsu
H.
,
Xiang
L.
,
Wang
Z.
&
Hu
B.
2012
Increased water storage in North America and Scandinavia from GRACE gravity data
.
Nature Geoscience
6
,
38
42
.
Wang
H.
,
Guan
H.
,
Gutiérrez-Jurado
H. A.
&
Simmons
C. T.
2014
Examination of water budget using satellite products over Australia
.
Journal of Hydrology
511
,
546
554
.
Wang
H.
,
Chen
A.
,
Wang
Q.
&
He
B.
2015
Drought dynamics and impacts on vegetation in China from 1982 to 2011
.
Ecological Engineering
75
,
303
307
.
Webber
M.
,
Barnett
J.
,
Finlayson
B.
&
Wang
M.
2008
Pricing China's irrigation water
.
Global Environmental Change
18
(
4
),
617
625
.
Wei
J.
,
Tao
S. Y.
&
Zhang
Q. Y.
2003
Analysis of drought in northern China based on the Palmer Severity Drought Index
.
Acta Geographica Sinica
58
(
S1
),
91
99
.
Wu
Z.
,
Wu
J.
,
He
B.
,
Liu
J.
,
Wang
Q.
,
Zhang
H.
&
Liu
Y.
2014
Drought offset ecological restoration program-induced increase in vegetation activity in the Beijing-Tianjin Sand Source Region, China
.
Environmental Science and Technology
48
(
20
),
12108
12117
.
Xin
X.-k.
,
Li
K.-f.
,
Finlayson
B.
&
Yin
W.
2015
Evaluation, prediction, and protection of water quality in Danjiangkou Reservoir, China
.
Water Science and Engineering
8
(
1
),
30
39
.
Xu
K.
,
Yang
D.
,
Yang
H.
,
Li
Z.
,
Qin
Y.
&
Shen
Y.
2015
Spatio-temporal variation of drought in China during 1961–2012: a climatic perspective
.
Journal of Hydrology
526
,
253
264
.
Ye
T.
,
Shi
P.
,
Wang
J. A.
,
Liu
L.
,
Fan
Y.
&
Hu
J.
2012
China's drought disaster risk management: perspective of severe droughts in 2009–2010
.
International Journal of Disaster Risk Science
3
(
2
),
84
97
.
Zhang
X.
,
Tang
Q.
,
Pan
M.
&
Tang
Y.
2014
A long-term land surface hydrologic fluxes and states dataset for China
.
Journal of Hydrometeorology
15
,
2067
2084
.
Zheng
C.
,
Liu
J.
,
Cao
G.
,
Kendy
E.
,
Wang
H.
&
Jia
Y.
2010
Can China cope with its water crisis? Perspectives from the North China plain
.
Groundwater
48
,
350
354
.
Zhu
Y. P.
,
Zhang
H. P.
,
Chen
L.
&
Zhao
J. F.
2008
Influence of the South–North Water Diversion Project and the mitigation projects on the water quality of Han River
.
Science of the Total Environment
406
(
1–2
),
57
68
.
Zou
J.
,
Zhan
C.
,
Xie
Z.
,
Qin
P.
&
Jiang
S.
2016
Climatic impacts of the middle route of the South-to-North water transfer project over the Haihe River basin in North China simulated by a regional climate model
.
Journal of Geophysical Research: Atmospheres
121
(
15
),
8983
8999
.