1. Introduction
Soil moisture is a major component of climate systems and is driven by precipitation (Seneviratne et al. 2010). Through evaporation and transpiration by plants, soil moisture also affects the interaction between the land, atmosphere, and the water cycle as parts of the overall ecosystem (Fischer et al. 2007; Hong and Pan 2000; Liu et al. 2015; Luo et al. 2007; Xu et al. 2004). Moreover, soil moisture has a long memory of climate signals over the continents, making it comparatively predictable (Koster and Suarez 2001; Vautard et al. 2007). For example, the soil moisture memory in the midlatitudes is longer than that of tropics, and such memory is enhanced by any soil moisture–precipitation feedback (Liu et al. 2014). In addition, a number of researchers have revealed the influence of soil moisture on surface temperature (Alfaro et al. 2006; Durre et al. 2000; Hirschi et al. 2011; Liu et al. 2014; Shinoda and Yamaguchi 2003; Stéfanon et al. 2014; Wu and Zhang 2013; Zhang and Dong 2010) and precipitation (Alfieri et al. 2008; D’Odorico and Porporato 2004; Ho-Hagemann et al. 2015; Liu et al. 2014; Meng and Quiring 2010; Meng et al. 2017; Wu et al. 2002). These effects include two aspects: the antecedent influence and feedback effects of soil moisture. Recently, several investigations have presented novel results that have documented the relationships between soil moisture and both wildfires and outbreaks of dust storms (Kim and Choi 2015; Krueger et al. 2015). Note that an important feature of soil moisture is its strong relationship with drought, especially agricultural drought. Therefore, a number of studies have commonly used soil moisture to quantify agricultural drought (Andreadis et al. 2005; Dai 2011, 2013; Han et al. 2014; Mao et al. 2017; Martínez-Fernández et al. 2016; Mo 2008; Narasimhan and Srinivasan 2005; Nijssen et al. 2014; Panu and Sharma 2002; Sheffield and Wood 2007, 2008a,b; Sheffield et al. 2004; A. Wang et al. 2011; Wang 2005; X. Zhang et al. 2017) and even provide early warnings of drought (Ford et al. 2015).
In theory, the actual status of soil moisture at a particular site can be obtained from in situ measurements, which provide an ideal and very accurate data source. However, in situ measurements can only reflect local conditions and have poor temporal resolution. Therefore, various instruments have been developed to provide datasets with higher spatial and temporal resolution (Ventura et al. 2010; Zreda et al. 2012). In addition, a variety of datasets have also been created based on measurements from different station networks (Bell et al. 2013; Bosch et al. 2006; Chrisman and Zreda 2013; Dente et al. 2012; Dong and Tang 2015; Dorigo et al. 2011; Entin et al. 2000; Robock et al. 2000, 2005; Zreda et al. 2012), of which the best known are observational datasets from 2000 (Robock et al. 2000) and 2011 (Dorigo et al. 2011). Nevertheless, the temporal continuity of soil moisture data collected over long periods of time is only satisfactory in a few regions, such as in the United States and in Europe. In general, observed soil moisture records with long-term time series at large scales are still very limited and could not match the needs for research and operational monitoring.
Alternatively, as science and technology have developed in recent years, land surface model (LSM) simulations and remote sensing products (such as the Advanced Scatterometer and Advanced Microwave Scanning Radiometer of the Earth Observing System; Albergel et al. 2009; Brocca et al. 2011; Draper et al. 2009; Gruhier et al. 2008; Jackson et al. 2010; J. Zhang et al. 2015) have shown great potential for supplying reliable estimates of soil moisture. However, evident weaknesses in estimates of soil moisture based on satellite data have seriously prevented its widespread application. Particular problems include the thin superficial soil layer representation (between 1 mm and 2 cm), nonstationary time series, and large uncertainties generated during retrieval. In comparison, model simulations are often considered to be very good resources when analyzing variability in soil moisture for drought monitoring. For instance, the Global Land Data Assimilation System (GLDAS) provides products related to soil moisture at high spatial and temporal resolutions worldwide. The GLDAS is an offline (uncoupled to the atmosphere) terrestrial modeling system developed by the collaboration of the National Oceanic and Atmospheric Administration National Centers for Environmental Prediction and the National Aeronautics and Space Administration Goddard Space Flight Center; the goal of GLDAS is to generate high-quality land surface fields globally (Rodell et al. 2004). The main feature of GLDAS is that it makes use of advanced land surface modeling and data assimilation techniques combined with the integration of satellite-based remote sensing data and in situ observations (Rodell et al. 2004). The model-derived fields from this project supply useful information for several types of operational work, such as weather and climate prediction, applications related to water resources, and water cycle investigations. The high temporal and spatial resolutions allow the long-term soil moisture data from these simulations to provide excellent alternatives to ground-based field observations. Additionally, another example of an LSM is the Australian Community Atmosphere Biosphere Land Exchange (CABLE) model, which is one of the main missions in the development of the Australian Community Climate and Earth-System Simulator (ACCESS; Zhang et al. 2011).
As a result of the advantages of LSMs, several studies have been conducted to validate the accuracy of GLDAS soil moisture products. Chen et al. (2013) found that the four GLDAS models had a tendency to systematically underestimate soil moisture in superficial layer (0–5 cm) but simulated the soil moisture very well for deeper soil layers (20–40 cm). Bi and Ma (2015) found that the temporal variation of observed soil moisture was well represented by all of the four LSMs in GLDAS over the Tibetan Plateau. Li et al. (2015) used GLDAS model products as a proxy for in situ soil moisture data and validated the soil moisture products of Aquarius, an operational active/passive L-band satellite sensor, against model simulations from GLDAS. Bi et al. (2016) found that GLDAS-2 barely outperformed GLDAS-1 over the Tibetan Plateau. Grings et al. (2015) conducted an evaluation strategy that did not require in situ observations but instead used a land surface model (GLDAS) with two satellite-based soil moisture products in the Argentinean Pampas. Additionally, assessments were also conducted to evaluate the performance of CABLE. Previous studies, such as Wang and McGregor (2003) and Y.-P. Wang et al. (2011), evaluated the CABLE model against ground-based observations. Zhang et al. (2009) and Zhang et al. (2011) further explored the performance of an early model version in capturing the observed features of surface energy, water, and carbon fluxes using 50-yr offline simulations.
Given the satisfactory quality of the LSM soil moisture products, operational application of these products in drought monitoring is highly anticipated. A number of studies have been conducted to show the applicability of soil moisture simulations using LSMs in drought monitoring at regional or global scales. Nijssen et al. (2014) implemented the Global Drought Information System on the basis of multiple land surface models [the Variable Infiltration Capacity (VIC), Noah, and “Sacramento” models] and found that the Global Drought Information System is effective in providing data related to global drought conditions in near–real time. In the United States, Sheffield et al. (2004) demonstrated a drought analysis method using simulated soil moisture from the VIC LSM. They found that the VIC soil moisture simulations had the ability to capture temporal variability and major drought events over period from 1950 to 1999. Those authors indicated that this LSM-based drought analysis had several advantages when compared with observation-based drought indices, such as the consideration of small processes related to the formation of drought conditions due to the high spatial and temporal resolution of the data and the physical basis of an LSM. Mo (2008) evaluated soil moisture products from the North American Land Data Assimilation System (NLDAS) VIC and Noah models for their ability to identify drought over the United States. They concluded that products from the NLDAS VIC provided valuable alternatives for long-term observations of soil moisture. However, they also noted that discrepancies exist between VIC and Noah in detecting drought events, especially in the western interior of the United States. Shukla et al. (2011) evaluated the ability of an LSM (VIC model)-based drought-monitoring system and found that the drought-monitoring system had excellent performance in detecting the onset and recovery of drought conditions in Washington State. Over Europe, Cammalleri et al. (2015) found that ensemble model soil moisture products based on three land surface models [“LISFLOOD”, the Community Land Model (CLM), and TESSEL (Tiled ECMWF Scheme of Surface Exchanges over Land)] are more skillful than individual models in retrieving extreme drought events even though no substantial discrepancies were detected among soil moisture outputs from the three LSMs. In China, A. Wang et al. (2011) investigated soil moisture drought using an ensemble of four LSM models (VIC, CLM, Noah, and a hybrid of CLM3.5 with the VIC) for the period 1950–2006. They found that problems with soil moisture droughts have become more severe and droughts have become more prolonged and frequent during the past 57 years, especially for northeastern and central China, suggesting an increasing susceptibility of agriculture to drought. Li and Ma (2015) detected drought using soil moisture simulations from LSM (CLM3.5) and found that drought in China presented an increasing trend over the period from 1951 to 2008; in addition, drought characteristics at multiple time scales have also been documented. More recently, L. Zhang et al. (2017) evaluated the performance of LSMs in detecting near-real-time drought in China.
Despite various validations of LSMs in their performance related to drought monitoring, an intercomparison and evaluation of LSMs derived from different countries on drought monitoring across China is still lacking, particularly in many important regions. In this paper, two LSMs, Noah and CABLE, were selected to evaluate their ability to monitor drought in terms of satisfactory resolution, coverage, and excellent near-real-time operational performance in China, individually. Additionally, two commonly used drought indices, the standardized precipitation index (SPI) and the standardized precipitation evapotranspiration index (SPEI), were employed to represent actual drought conditions because they provides long time series. This paper focuses on three areas in China: northern China (NC), northwestern China (NWC), and southwestern China (SWC). The selection of these three key regions was mainly for three reasons. First, droughts most frequently occurred in northern China, followed by northwestern and southwestern China, during the past 50 years. Second, there are several pieces of evidence for increasing extreme droughts in northern and northwestern China since the 1980s (Ma and Fu 2006; Qian et al. 2011). Third, southwestern China is a region where drought events have become frequent since 2000, although it is classified as a humid area. The purpose of the study presented here is to examine the performance of soil moisture products derived from the Noah and CABLE models in representing drought situations over these three important areas of China using 50 years of data and to assess the results and provide suggestions on the strengths and drawbacks of LSM products for drought monitoring.
This paper is organized as follows: Section 2 briefly introduces the methods, soil moisture products, and drought indices used in this study; section 3 analyzes the comparison results of the soil moisture products in drought monitoring; and section 4 presents the conclusions and discussion.
2. Data and method
This analysis focuses on three regions of China: northern China (33.5°–42.5°N, 110.5°–120.5°E), northwestern China (32.5°–40.5°N, 102.5°–110.5°E), and southwestern China (22.5°–31.5°N, 98.5°–109.5°E; Fig. 1). Modeled soil moisture from the Noah (from GLDAS-2.0) and CABLE models spanning from January 1961 to December 2010 with 1° grid resolution and 1-month temporal frequency were employed. Additionally, in situ precipitation and temperature measurements from 537 stations produced by the National Meteorological Information Center at the China Meteorological Administration were also used to calculate the SPI and SPEI using the same time period.
Locations of the three regions: northern China (NC; 33.5°–42.5°N, 110.5°–120.5°E; red border), northwestern China (NWC; 32.5°–40.5°N, 102.5°–110.5°E; blue border), and southwestern China (SWC; 22.5°–31.5°N, 98.5°–109.5°E; green border).
Citation: Journal of Applied Meteorology and Climatology 57, 9; 10.1175/JAMC-D-17-0118.1
a. GLDAS products and forcing data
A global high-resolution archive of simulated products from GLDAS is available with the support of the Land Information System based on ingesting a significant number of observation-based fields (Kumar et al. 2006). Currently, two versions have been developed for GLDAS: GLDAS, version 1 (hereinafter GLDAS-1), and GLDAS, version 2 (hereinafter GLDAS-2). GLDAS-2 can be further divided into two components, GLDAS-2.0 and GLDAS-2.1, although this study focuses only on GLDAS-2.0 products. What makes GLDAS-2.0 more advanced than GLDAS-1 is that GLDAS-2.0 employed the “Global Meteorological Forcing Dataset,” which provides more accurate meteorological forcing from Princeton University that was derived by combining reanalysis with observations (Sheffield et al. 2006), thus producing more climatologically consistent datasets. Specifically, GLDAS-2.0 provides long-term datasets spanning 1948–2010 at a high temporal resolution of 3 h. In contrast, forcing datasets of GLDAS-1 are not constant and changed several times from 1979 to the present, especially in 1995–97, leading to abnormal trends (see Rui and Beaudoing 2014). In addition, improvements on other aspects were also conducted in GLDAS-2.0, such as providing more accurate data sources for land surface parameters, model version upgrades, and updated bottom layer temperature of the Noah model (Rui and Beaudoing 2014). Therefore, this study uses the data products from GLDAS-2.0 given the advantages of this version. Recently, GLDAS-2.0 drives four LSMs: Noah, CLM, Mosaic, and VIC. However, only a small part of the GLDAS-2.0-simulated data was obtainable from the official GLDAS-2.0 website (https://disc.sci.gsfc.nasa.gov/datasets?keywords=GLDAS); these simulations come from the Noah model. In this paper, we use the GLDAS-2.0/Noah products that cover from January 1961 to December 2010 with a spatial resolution of 1.0° and a temporal resolution of 1 month. Here, the monthly data were created from temporal averaging of the 3-h products. The Noah model has four soil layers: 0–10, 10–40, 40–100, and 100–200 cm. To better represent seasonal variations, the first three layers (0–100 cm) of GLDAS-2.0/Noah were used to compare with CABLE soil moisture and SPI/SPEI.
b. CABLE products and forcing data
CABLE is an LSM that was originally generated in Australia and has been improved continuously in its different versions (Zhang et al. 2011). Note that the version of CABLE used in this analysis is the same as the version applied by L. Zhang et al. (2017). The primary role of CABLE is to describe water, land surface energy, and carbon cycles realistically in the global weather and climate system. Five primary parts comprise its framework: soil and snow, radiation, surface flux, ecosystem, and canopy micrometeorology (L. Zhang et al. 2017). In terms of numerical schemes used in the model, aerodynamic transfer resistance, surface roughness length, and zero-plane displacement height are calculated employing the scheme proposed by Raupach (1994), while direct beam radiation and the net diffuse are computed according to the theory of Wang and Leuning (1998) and L. Zhang et al. (2017). The major features, history of the development of the CABLE model, and comparisons with observations at a number of sites can be found in Kowalczyk et al. (2006). In recent years, several studies have indicated that the model performs well and it is being used to provide drought-monitoring products operationally over China based on a China–Australia bilateral project that addresses climate change (Zhang et al. 2009; L. Zhang et al. 2017).
This paper uses the CABLE simulations at 1.0° spatial resolution and 1-month temporal resolution for 1961–2010 in China. This level of spatial/temporal resolution and coverage is entirely the same as that of the GLDAS-2.0/Noah, making an evaluation between these two LSMs feasible and convenient. The soil analysis of this model is divided into six layers, with depth of 0–2.2, 2.2–8, 8–23.4, 23.4–64.3, 64.3–172.8, and 172.8–460 cm from the top to the bottom. The study focuses on the upper four layers with a total depth of 64.3 cm for consistency with the depth used in Noah. The meteorological forcing data used for CABLE are reanalysis-based meteorological data that were corrected by station observations. The forcing data can be characterized by two periods in terms of their correction procedures: 1948–2000 and 2001–10; these were developed by Ngo-Duc et al. (2005) and L. Zhang et al. (2017), respectively. Detailed descriptions of the data can be found in L. Zhang et al. (2017).
c. SPI
The SPI is a drought index developed by McKee et al. (1993). It is characterized by multiple time scales ranging from 1 to 36 months and has a simple principle that presents the probability of the occurrence of a precipitation deficit over a given time period (McKee et al. 1993). Despite the presence of drought events, the SPI also has the ability to detect floods and to determine their severity (Moreira et al. 2006). To calculate SPI, precipitation deficits should be quantified on the basis of long-term precipitation data for the desired period (Hayes et al. 1999). First, the cumulative probability of an observed precipitation event for a desired time scale is deduced after the long-term datasets are fitted to a gamma distribution (Z. Zhang et al. 2015). Then, by transforming the cumulative probability to a standard normal random variable, the SPI value can be calculated. The value of the SPI was classified into several grades in which positive values indicate wet spell and negative values illustrate dry spell. Based on previous studies, SPIs with short time scales are significantly correlated to agricultural drought while those with long time scales may represent hydrological drought. For the purpose of evaluating LSM products against drought in this study, SPI with a time scale of 3 months (SPI3) and 1 month (SPI1) can meet the need to represent the actual drought status at each station over the three domains in China analyzed here. Given that the soil depths for the two LSMs used here are not identical, the employment of different time scales for SPI (both are short scale) is indispensable because SPI1 may have a higher correlation with shallower soil depth than SPI3.
d. SPEI
The standardized precipitation evapotranspiration index was proposed as a drought index that is based on the difference between precipitation and potential evapotranspiration (PET; Vicente-Serrano et al. 2010a). The principle for calculating the index is similar to that for the SPI. The only difference between the two indices is that SPEI uses “climatic water balance” instead of precipitation as input, meaning that this index includes the role of temperature and ET. Therefore, the SPEI is more suitable for identifying soil moisture drought than indices that only consider precipitation. The advantages of being multiscalar and considering temperature make SPEI an excellent index for drought analysis. More details for computing procedure and related comparisons among other drought indicators can be found in a series of studies conducted by Vicente-Serrano et al. (2010a,b, 2011a,b, 2012). In recent years, SPEI has been widely used in drought studies (Chen and Sun 2015; Wang et al. 2015; Xu et al. 2015; Yu et al. 2014; Zuo et al. 2018). Here, SPEI with time scales of 1 month (SPEI1) and 3 months (SPEI3) was also employed to evaluate LSM-based drought products.
3. Results
a. Forcing data
The corresponding statistical scores of these three metrics are presented in Table 1 over the period between 1961 and 2010 for three regions in China: NC, NWC, and SWC. The observation-based precipitation data come from the National Meteorological Information Center at the China Meteorological Administration.
Statistical metrics of comparison between precipitation (mm month−1) forcing data from Noah and CABLE and in situ precipitation (mm month−1) for three regions in China: northern China, northwestern China, and southwestern China; R values are all significant at 99.9% confidence levels.
On average, the monthly mean in situ precipitation values were 47.39, 36.49, and 96.17 mm month−1 in NC, NWC, and SWC, respectively. Biases of both models for precipitation data in all of the regions were small in comparison with the mean rainfall although the forcing data underestimated precipitation except for CABLE in NWC, and the magnitude of biases was similar for both models in all three regions.
Precipitation forcing for Noah has lower RMSE values in the northern regions of China (NC and NWC) more consistently than those values in southern China. Similar to the bias, NWC has the lowest RMSE of the three regions for both models. Both sets of forcing data were significantly correlated with observations, with R values of 0.99 in all regions except for NWC, which had relatively lower R values (0.98 for both Noah and CABLE). Overall, regional mean statistics indicated that the accuracy of CABLE forcing data was similar to that of Noah, although slight discrepancies did exist.
Figure 2 shows the correlation coefficients between model forcing precipitation and observation-based fields for 1961–2010 on each grid. Generally, the correlations between observations and forcing data of both models have similar spatial patterns regardless of negligible differences over the three regions. To be specific, significant correlations were observed between observations and forcing for both models that were greater than 0.80 in all the three regions, indicating a high level of accuracy. Therefore, one can assume that the precipitation rates for both models may be used as credible datasets for forcing the LSMs.
Temporal correlations (areas significant at the 99.9% confidence level are colored) of precipitation between observations and LSM forcing data used in this study: observations vs the (left) Noah and (right) CABLE models.
Citation: Journal of Applied Meteorology and Climatology 57, 9; 10.1175/JAMC-D-17-0118.1
b. Simulated products
SPI and SPEI are two widely used drought indices, and one of the most remarkable advantages for them is the flexible time scales. However, there is discrepancy between the two indices in terms of input elements. SPI only uses precipitation whereas SPEI considers both precipitation and PET. Theoretically, the main driver for drought is the shortage of rainfall. Nevertheless, the impact of evapotranspiration (ET) is significant when using soil moisture for agricultural drought monitoring. In addition, soil moisture at different depths has various time scales. Therefore, a sensitivity analysis to identify a more satisfactory drought index at a reasonable time scale is inevitable before the soil moisture drought assessment. Figure 3 shows the spatial patterns of the correlation between the drought indicators at different time scales (SPI1, SPI3, SPEI1, and SPEI3) and the simulated standardized soil moisture anomalies (SSMA) generated by the two LSMs for the period 1961–2010. In general, high correlations are found in time scales of 3 months rather than 1 month for both drought indices and both models. Furthermore, SPEI3 presents a better agreement with soil moisture of both models than SPI3. Therefore, SPEI3 data are used as the reference for the evaluation of LSM-based soil moisture drought products. When referring to SPEI3, Noah provided stronger correlations than CABLE with significantly positive values in SWC and similar correlations in NWC and NC. Despite slight differences, the two LSMs exhibit regional coherence, with maximum correlation values that occurred in the western, middle, and southwestern sections of NC, NWC, and SWC, respectively. Of the three regions, the highest correlations obtained were located in SWC, followed by NWC; the lowest correlation coefficients appeared in NC. Although Spennemann et al. (2015) pointed out the fact that precipitation served as the main driver of simulated soil moisture anomalies, inconsistencies of correlation patterns between Figs. 2 and 3 reveal that the discrepancies between LSMs also play a crucial role in simulations. Therefore, significant correlations presented by Noah may be associated with the fact that Noah outperforms CABLE in simulating soil moisture.
Correlations (areas significant at the 99.9% confidence level are colored) between drought indices (SPEI or SPI) at different time scales and the SSMA from the two land surface models for 1961–2010: (a) SPEI1, (b) SPEI3 (c) SPI1, and (d) SPI3 vs the (left) Noah and (right) CABLE models.
Citation: Journal of Applied Meteorology and Climatology 57, 9; 10.1175/JAMC-D-17-0118.1
Furthermore, the spatial patterns of the correlations with SPEI3 for different seasons are also presented in Fig. 4. Overall, the most obvious feature presented by Fig. 4 is a significant discrepancy between warm and cold seasons for both models, with higher correlations in warm seasons, such as June–August and September–November, and lower values in cold seasons (December–February and March–May). In addition, note that model discrepancy is also apparent. This difference between models is mainly reflected in that CABLE shows a more heterogeneous distribution than Noah when seasons change to the cold half of a year. Specifically, the correlation difference of Noah between cold seasons and warm seasons was significant higher than that of CABLE in the northern part of China (NC and NWC) while seasonal discrepancy in the southern part of China (SWC) was fairly small in both models. In warm seasons, NC and NWC show higher correlation values in Noah than in CABLE, while the corresponding values of Noah are evidently small when compared with those of CABLE in cold seasons. However, correlation coefficients in SWC are generally higher than those in NC and NWC over all seasons for Noah. This information suggests that in addition to the seasonal discrepancies between the two models, correlation values from the three regions are also characterized in different ways. The possible reason for the seasonal difference from both models might be related to frozen processes of soil moisture in autumn and winter that cause the simulations to not represent the actual situation of drought in the northern part of China. Meanwhile, the warm climate of SWC reduces the opportunity of frozen conditions affecting the results, making this region more consistent with soil moisture variations.
Seasonal correlations (areas significant at 99.9% confidence levels are colored) between SPEI3 and SSMA from the two land surface models for 1961–2010: SPEI3 vs the (a) Noah and (b) CABLE models.
Citation: Journal of Applied Meteorology and Climatology 57, 9; 10.1175/JAMC-D-17-0118.1
Figure 5 shows the areally averaged temporal evolution of the SPEI3 and the two SSMAs generated by Noah and CABLE for the period from 1961 to 2010. At the same time, correlation coefficients of these time series data for NC, NWC, and SWC are presented in Table 2 to quantify the relationship between monthly time series of the SPEI3 and SSMA. Among the three key regions in China, SWC shows a highest correlations (R > 0.55) for both models, followed by NWC with correlations around 0.55, and correlation coefficients in NC are fairly lower than 0.50. This result agrees with the fact illustrated by Fig. 3b. Note that very good agreement was observed between Noah and SPEI3 in SWC, while the time series generated by CABLE experienced small departures from those generated by Noah. In addition, there are particular periods with opposite behavior between Noah/CABLE and SPEI3 in Fig. 5, such as 1968, 1983, 1987, 2003, and 2010 in NC; 1984, 1998, and 2010 in NWC; and 1975, 1981, and 1997 in SWC. In particular, these periods were found almost in cold seasons for NC and NWC, which can be explained as the frozen processes of soil moisture mentioned in Fig. 4.
Temporal series of SPEI3 along with the SSMA from the Noah and CABLE models, for (a) NC, (b) NWC, and (c) SWC.
Citation: Journal of Applied Meteorology and Climatology 57, 9; 10.1175/JAMC-D-17-0118.1
Correlation coefficients between monthly time series of SPEI3 and SSMA from the Noah and CABLE models for NC, NWC, and SWC. (Correlation coefficients are all significant at 99.9% confidence levels.)
Another notable phenomenon is that a different mean was found for 1961–64 when compared with the rest of the period for Noah and CABLE over NC. In contrast, the changing point was not found in times series of SPEI. To detect the reasons, time series of the areal averaged precipitation (observation data, forcing data for Noah and CABLE) were investigated in NC and no discrepancies were found between 1961–64 and 1965–2010. Thus we can deduce that the turning point was related to the uncertainty in the simulations hydrological processes including evapotranspiration. Despite the fact that both LSMs were able to capture periods of excess and deficit, the magnitudes shown in SWC were greater than those of NC and NWC for both LSMs, particularly for the period between 2001 and 2010 for CABLE. The large magnitude of CABLE between 2001 and 2010 can probably be explained by the different correction schemes for forcing data over the two periods of 1961–2000 and 2001–10. Specifically, meteorological forcing data described in Ngo-Duc et al. (2005) were applied in CABLE for the period of 1961–2000; this suite of datasets was developed based on the National Centers for Environmental Prediction–National Center for Atmospheric Research reanalysis and with error corrected by the Climate Research Unit (CRU) data for all variables. However, forcing data for the subsequent period of 2001–10 were produced by L. Zhang et al. (2017) with a slightly different procedure. That procedure treated precipitation differently than other variables by adjusting it with rainfall information from the Climate Prediction Center Merged Analysis of Precipitation (Xie et al. 2003) while other variables were rectified by data according to the method used by Ngo-Duc et al. (2005). In other words, the key point needed to distinguish data quality when comparing between 1961–2000 and 2001–10 is to determine whether precipitation is treated separately using the advantage of field observations although the basic datasets that needed to be rectified are the same. The unique processing procedure for rainfall mainly considered the importance of precipitation on the formation and development of drought, given that CABLE would be run at the Institute of Arid Meteorology as a drought-monitoring tool.
c. Case studies in drought monitoring
In this section, we evaluate the models’ ability in capturing the spatial patterns of drought events in NC, NWC, and SWC by comparing with SPEI3. Droughts can be categorized into five grades on the basis of the SPEI values. Accordingly, to assess drought-monitoring performance, soil moisture products were also transformed into drought grades that are based on percentile values, which matched those of SPEI (Table 3). As illustrated in Fig. 4, the correlations between SPEI3 and soil moisture from both models have higher values in autumn for NC and NWC, whereas the correlations were significant in spring for SWC. Therefore, an autumn drought event over NC and NWC in 2002 and a spring drought event over SWC in 2010 were selected.
Drought grades in terms of soil moisture percentile.
Between September and November of 2002, severe drought spread over NC to NWC and had significant socioeconomic impacts. Figure 6 shows the evolution of the drought patterns. As illustrated by SPEI3, the drought weakened from September to November with narrowed scale and decreased intensity. In particular, drought intensity was dramatically high, with severe to extreme grades over southwestern NWC and eastern NC in September. Then the intensity declined in southwestern NWC, and both the intensity and scale diminished in eastern NC. Over southwestern part of NWC, the drought scale from Noah generally agrees with SPEI3, while both the scale and intensity have significant discrepancies between CABLE and SPEI3. Over NC, both models overestimated the intensity. Overall, Noah presents spatial patterns that are much more similar to the SPEI3 for this drought event.
Spatial patterns of autumn droughts in 2002 monitored by SPEI3 and the Noah and CABLE models.
Citation: Journal of Applied Meteorology and Climatology 57, 9; 10.1175/JAMC-D-17-0118.1
Between March and May of 2010, SWC experienced a spring drought. Although dryness was the most dominant feature in cold seasons over the regions of SWC, this spring drought event was catastrophic. Figure 7 illustrates the evolution of this drought event. Similarly, drought weakened from March to May shown by SPEI3, with decreasing intensity and scale over middle-southern SWC. Evolution of drought intensity is generally similar for both the Noah and CABLE models. Nevertheless, evident discrepancies exist in drought scale, in that the drought center from CABLE is deviated from that of SPEI3. In terms of representing the 2010 drought spatially, the Noah soil moisture pattern appears to be more reliable.
As in Fig. 6, but for spring droughts in 2010.
Citation: Journal of Applied Meteorology and Climatology 57, 9; 10.1175/JAMC-D-17-0118.1
4. Conclusions
Soil moisture not only serves as an indicator of excess and deficit soil water but also is regarded as one of the common drought indices used to detect and quantify drought, particularly agricultural drought. Therefore, high-quality land surface model products are important and are urgently needed for application in drought monitoring at regional scales. In this context, an assessment of the simulated soil moisture generated by CABLE and Noah (from GLDAS-2.0) was carried out in terms of performance in drought monitoring over northern, northwestern, and southwestern China for the period 1961–2010.
The main conclusions are as follow: 1) The precipitation forcing data of both models were very accurate when compared with local observations as based on statistical metrics, although both models underestimate precipitation with a negative bias except for CABLE over NWC. 2) GLDASv2/Noah outperformed CABLE, especially in SWC, and is thus a useful tool for capturing drought conditions and their variability, but both simulations show good correlations with SPEI3 in NC, NWC, and SWC. 3) The 2002 autumn drought and 2010 spring drought conditions were selected to compare the model drought categories with the SPEI3 drought category, and GLDASv2/Noah was slightly better than CABLE.
This study distinguishes itself from other analyses in that it employs soil moisture products from new versions of two land surface models, making long-term drought monitoring and assessment possible. Then, the SPEI index with 3-month time scale was selected to quantify the actual status of drought when a comparison was conducted between model products and observations. The choice of SPEI3 was based on a sensitivity analysis that compares the correlations between the two drought indicators of different time scales (SPI1, SPI3, SPEI1, and SPEI3) and SSMA generated by the two LSMs to identify a more satisfactory drought index and a certain time scale. SPEI is a valuable metric for representing drought because of its multiple time scales and consideration of both precipitation and evapotranspiration. In this sense, it can capture soil moisture characteristics better than SPI.
Further research should be done to consider the microaspects of soil moisture. In comparing the results obtained by Noah with those obtained by CABLE, it is seen that the choice of the LSM clearly has a great impact on soil moisture outputs during drought detection and monitoring when the accuracy of atmospheric forcing is similar in both models. In addition, another problem with regard to depth of soil moisture used in this study should be further explored because the number of soil layers and depth of each layer are not identical between the two models. We chose depths from both models and tried to match the need for identically modeled soil depths. Uncertainties generated by depth discrepancy may still exist, however, requiring further discussion in the next study.
The use of simulated soil moisture values for drought research, drought monitoring, and other applications is steadily growing. This paper provides a robust validation result that encourages making use of model-derived simulations as a new data source for drought investigations. Furthermore, other kinds of data sources and tools are expected to be developed in the future.
Acknowledgments
This work was supported by the China Special Fund for Meteorological Research in the Public Interest (Major projects) (GYHY201506001) and National Natural Science Foundation of China (NSFC) (41775093). The data used in this study were acquired as part of the mission of NASA’s Earth Science Division and are archived and distributed by the Goddard Earth Sciences Data and Information Services Center. The authors declare no conflicts of interest.
REFERENCES
Albergel, C., C. Rüdiger, D. Carrer, J.-C. Calvet, N. Fritz, V. Naeimi, Z. Bartalis, and S. Hasenauer, 2009: An evaluation of ASCAT surface soil moisture products with in-situ observations in southwestern France. Hydrol. Earth Syst. Sci., 13, 115–124, https://doi.org/10.5194/hess-13-115-2009.
Alfaro, E. J., A. Gershunov, and D. Cayan, 2006: Prediction of summer maximum and minimum temperature over the central and western United States: The roles of soil moisture and sea surface temperature. J. Climate, 19, 1407–1421, https://doi.org/10.1175/JCLI3665.1.
Alfieri, L., P. Claps, P. D’Odorico, F. Laio, and T. M. Over, 2008: An analysis of the soil moisture feedback on convective and stratiform precipitation. J. Hydrometeor., 9, 280–291, https://doi.org/10.1175/2007JHM863.1.
Andreadis, K. M., E. A. Clark, A. W. Wood, A. F. Hamlet, and D. P. Lettenmaier, 2005: Twentieth-century drought in the conterminous United States. J. Hydrometeor., 6, 985–1001, https://doi.org/10.1175/JHM450.1.
Bell, J. E., and Coauthors, 2013: U.S. Climate Reference Network soil moisture and temperature observations. J. Hydrometeor., 14, 977–988, https://doi.org/10.1175/JHM-D-12-0146.1.
Bi, H., and J. Ma, 2015: Evaluation of simulated soil moisture in GLDAS using in-situ measurements over the Tibetan Plateau. 2015 IEEE Int. Geoscience and Remote Sensing Symp., Milan, Italy, IEEE, 4825–4828, https://doi.org/10.1109/IGARSS.2015.7326910.
Bi, H., J. Ma, W. Zheng, and J. Zeng, 2016: Comparison of soil moisture in GLDAS model simulations and in situ observations over the Tibetan Plateau. J. Geophys. Res. Atmos., 121, 2658–2678, https://doi.org/10.1002/2015JD024131.
Bosch, D. D., V. Lakshmi, T. J. Jackson, M. Choi, and J. M. Jacobs, 2006: Large scale measurements of soil moisture for validation of remotely sensed data: Georgia soil moisture experiment of 2003. J. Hydrol., 323, 120–137, https://doi.org/10.1016/j.jhydrol.2005.08.024.
Brocca, L., and Coauthors, 2011: Soil moisture estimation through ASCAT and AMSR-E sensors: An intercomparison and validation study across Europe. Remote Sens. Environ., 115, 3390–3408, https://doi.org/10.1016/j.rse.2011.08.003.
Cammalleri, C., F. Micale, and J. Vogt, 2015: On the value of combining different modelled soil moisture products for European drought monitoring. J. Hydrol., 525, 547–558, https://doi.org/10.1016/j.jhydrol.2015.04.021.
Chen, H., and J. Sun, 2015: Changes in drought characteristics over China using the standardized precipitation evapotranspiration index. J. Climate, 28, 5430–5447, https://doi.org/10.1175/JCLI-D-14-00707.1.
Chen, Y., K. Yang, J. Qin, L. Zhao, W. Tang, and M. Han, 2013: Evaluation of AMSR-E retrievals and GLDAS simulations against observations of a soil moisture network on the central Tibetan Plateau. J. Geophys. Res. Atmos., 118, 4466–4475, https://doi.org/10.1002/jgrd.50301.
Chrisman, B., and M. Zreda, 2013: Quantifying mesoscale soil moisture with the cosmic-ray rover. Hydrol. Earth Syst. Sci., 17, 5097–5108, https://doi.org/10.5194/hess-17-5097-2013.
Dai, A., 2011: Drought under global warming: A review. Wiley Interdiscip. Rev.: Climate Change, 2, 45–65, https://doi.org/10.1002/wcc.81.
Dai, A., 2013: Increasing drought under global warming in observations and models. Nat. Climate Change, 3, 52–58, https://doi.org/10.1038/nclimate1633.
Dente, L., Z. Vekerdy, J. Wen, and Z. Su, 2012: Maqu network for validation of satellite-derived soil moisture products. Int. J. Appl. Earth Obs. Geoinf., 17, 55–65, https://doi.org/10.1016/j.jag.2011.11.004.
D’Odorico, P., and A. Porporato, 2004: Preferential states in soil moisture and climate dynamics. Proc. Natl. Acad. Sci. USA, 101, 8848–8851, https://doi.org/10.1073/pnas.0401428101.
Dong, L., and S. Tang, 2015: A multi-scale soil moisture and temperature regularly automatic monitoring network aim at multi-satellite data validation in Tibet Plateau. Proc. SPIE, 9637, 96371G, https://doi.org/10.1117/12.2194429.
Dorigo, W. A., and Coauthors, 2011: The International Soil Moisture Network: A data hosting facility for global in situ soil moisture measurements. Hydrol. Earth Syst. Sci., 15, 1675–1698, https://doi.org/10.5194/hess-15-1675-2011.
Draper, C. S., J. P. Walker, P. J. Steinle, R. A. M. de Jeu, and T. R. H. Holmes, 2009: An evaluation of AMSR-E derived soil moisture over Australia. Remote Sens. Environ., 113, 703–710, https://doi.org/10.1016/j.rse.2008.11.011.
Durre, I., J. M. Wallace, and D. P. Lettenmaier, 2000: Dependence of extreme daily maximum temperatures on antecedent soil moisture in the contiguous United States during summer. J. Climate, 13, 2641–2651, https://doi.org/10.1175/1520-0442(2000)013<2641:DOEDMT>2.0.CO;2.
Entin, J. K., A. Robock, K. Y. Vinnikov, S. E. Hollinger, S. X. Liu, and A. Namkhai, 2000: Temporal and spatial scales of observed soil moisture variations in the extratropics. J. Geophys. Res., 105, 11 865–11 877, https://doi.org/10.1029/2000JD900051.
Fischer, E. M., S. I. Seneviratne, P. L. Vidale, D. Lüthi, and C. Schär, 2007: Soil moisture–atmosphere interactions during the 2003 European summer heat wave. J. Climate, 20, 5081–5099, https://doi.org/10.1175/JCLI4288.1.
Ford, T. W., D. B. McRoberts, S. M. Quiring, and R. E. Hall, 2015: On the utility of in situ soil moisture observations for flash drought early warning in Oklahoma, USA. Geophys. Res. Lett., 42, 9790–9798, https://doi.org/10.1002/2015GL066600.
Grings, F., C. A. Bruscantini, E. Smucler, F. Carballo, M. E. Dillon, E. A. Collini, M. Salvia, and H. Karszenbaum, 2015: Validation strategies for satellite-based soil moisture products over Argentine Pampas. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 8, 4094–4105, https://doi.org/10.1109/JSTARS.2015.2449237.
Gruhier, C., P. de Rosnay, Y. Kerr, E. Mougin, E. Ceschia, J. C. Calvet, and P. Richaume, 2008: Evaluation of AMSR-E soil moisture product based on ground measurements over temperate and semi-arid regions. Geophys. Res. Lett., 35, L10405, https://doi.org/10.1029/2008GL033330.
Guo, Z., P. A. Dirmeyer, Z.-Z. Hu, X. Gao, and M. Zhao, 2006: Evaluation of the Second Global Soil Wetness Project soil moisture simulations: 2. Sensitivity to external meteorological forcing. J. Geophys. Res., 111, D22S03, https://doi.org/10.1029/2006JD007845.
Han, E., W. T. Crow, T. Holmes, and J. Bolten, 2014: Benchmarking a soil moisture data assimilation system for agricultural drought monitoring. J. Hydrometeor., 15, 1117–1134, https://doi.org/10.1175/JHM-D-13-0125.1.
Hayes, M. J., M. D. Svoboda, D. A. Wilhite, and O. V. Vanyarkho, 1999: Monitoring the 1996 drought using the standardized precipitation index. Bull. Amer. Meteor. Soc., 80, 429–438, https://doi.org/10.1175/1520-0477(1999)080<0429:MTDUTS>2.0.CO;2.
Hirschi, M., and Coauthors, 2011: Observational evidence for soil-moisture impact on hot extremes in southeastern Europe. Nat. Geosci., 4, 17–21, https://doi.org/10.1038/ngeo1032.
Ho-Hagemann, H. T. M., S. Hagemann, and B. Rockel, 2015: On the role of soil moisture in the generation of heavy rainfall during the Oder flood event in July 1997. Tellus, 67A, 28661, https://doi.org/10.3402/tellusa.v67.28661.
Hong, S. Y., and H. L. Pan, 2000: Impact of soil moisture anomalies on seasonal, summertime circulation over North America in a regional climate model. J. Geophys. Res., 105, 29 625–29 634, https://doi.org/10.1029/2000JD900276.
Jackson, T. J., and Coauthors, 2010: Validation of Advanced Microwave Scanning Radiometer soil moisture products. IEEE Trans. Geosci. Remote Sens., 48, 4256–4272, https://doi.org/10.1109/TGRS.2010.2051035.
Kim, H., and M. Choi, 2015: Impact of soil moisture on dust outbreaks in East Asia: Using satellite and assimilation data. Geophys. Res. Lett., 42, 2789–2796, https://doi.org/10.1002/2015GL063325.
Koster, R. D., and M. J. Suarez, 2001: Soil moisture memory in climate models. J. Hydrometeor., 2, 558–570, https://doi.org/10.1175/1525-7541(2001)002<0558:SMMICM>2.0.CO;2.
Kowalczyk, E., Y. P. Wang, R. M. Law, H. L. Davies, J. L. McGregor, and G. Abramowitz, 2006: The CSIRO Atmosphere Biosphere Land Exchange (CABLE) model for use in climate models and as an offline model. CSIRO Marine & Atmospheric Research Tech. Paper 013, 37 pp., www.cmar.csiro.au/e-print/open/kowalczykea_2006a.pdf.
Krueger, E. S., T. E. Ochsner, D. M. Engle, J. D. Carlson, D. Twidwell, and S. D. Fuhlendorf, 2015: Soil moisture affects growing-season wildfire size in the southern Great Plains. Soil. Sci. Soc. Amer. J., 79, 1567–1576, https://doi.org/10.2136/sssaj2015.01.0041.
Kumar, S. V., and Coauthors, 2006: Land information system: An interoperable framework for high resolution land surface modeling. Environ. Modell. Software, 21, 1402–1415, https://doi.org/10.1016/j.envsoft.2005.07.004.
Li, D., T. Zhao, J. Shi, R. Bindlish, T. J. Jackson, B. Peng, M. An, and B. Han, 2015: First evaluation of Aquarius soil moisture products using in situ observations and GLDAS model simulations. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 8, 5511–5525, https://doi.org/10.1109/JSTARS.2015.2452955.
Li, M., and Z. Ma, 2015: Soil moisture drought detection and multi-temporal variability across China. Sci. China Earth Sci., 58, 1798–1813, https://doi.org/10.1007/s11430-015-5076-8.
Liu, D., G. Wang, R. Mei, Z. Yu, and M. Yu, 2014: Impact of initial soil moisture anomalies on climate mean and extremes over Asia. J. Geophys. Res. Atmos., 119, 529–545, https://doi.org/10.1002/2013JD020890.
Liu, D., Z. Yu, and J. Zhang, 2015: Diagnosing the strength of soil temperature in the land atmosphere interactions over Asia based on RegCM4 model. Global Planet. Change, 130, 7–21, https://doi.org/10.1016/j.gloplacha.2015.03.007.
Luo, Y., E. H. Berbery, K. E. Mitchell, and A. K. Betts, 2007: Relationships between land surface and near-surface atmospheric variables in the NCEP North American regional reanalysis. J. Hydrometeor., 8, 1184–1203, https://doi.org/10.1175/2007JHM844.1.
Ma, Z., and C. Fu, 2006: Some evidence of drying trend over northern China from 1951 to 2004. Chin. Sci. Bull., 51, 2913–2925, https://doi.org/10.1007/s11434-006-2159-0.
Mao, Y., Z. Wu, H. He, G. Lu, H. Xu, and Q. Lin, 2017: Spatio-temporal analysis of drought in a typical plain region based on the soil moisture anomaly percentage index. Sci. Total Environ., 576, 752–765, https://doi.org/10.1016/j.scitotenv.2016.10.116.
Martínez-Fernández, J., A. González-Zamora, N. Sánchez, A. Gumuzzio, and C. M. Herrero-Jiménez, 2016: Satellite soil moisture for agricultural drought monitoring: Assessment of the SMOS derived Soil Water Deficit Index. Remote Sens. Environ., 177, 277–286, https://doi.org/10.1016/j.rse.2016.02.064.
McKee, T. B., N. J. Doesken, and J. Kleist, 1993: The relationship of drought frequency and duration to time scales. Eighth Conf. on Applied Climatology, Anaheim, CA, Amer. Meteor. Soc., 179–184.
Meng, L., and S. M. Quiring, 2010: Examining the influence of spring soil moisture anomalies on summer precipitation in the U.S. Great Plains using the Community Atmosphere Model version 3. J. Geophys. Res., 115, D21118, https://doi.org/10.1029/2010JD014449.
Meng, W.-G., Y.-X. Zhang, J.-N. Li, G.-F. Dai, H.-R. Li, and Y.-Y. Huang, 2017: Sensitivity of mesoscale convective systems and associated heavy rainfall to soil moisture over south China. J. Trop. Meteor., 23, 91–102.
Mo, K. C., 2008: Model-based drought indices over the United States. J. Hydrometeor., 9, 1212–1230, https://doi.org/10.1175/2008JHM1002.1.
Moreira, E. E., A. A. Paulo, L. S. Pereira, and J. T. Mexia, 2006: Analysis of SPI drought class transitions using loglinear models. J. Hydrol., 331, 349–359, https://doi.org/10.1016/j.jhydrol.2006.05.022.
Narasimhan, B., and R. Srinivasan, 2005: Development and evaluation of soil moisture deficit index (SMDI) and evapotranspiration deficit index (ETDI) for agricultural drought monitoring. Agric. For. Meteor., 133, 69–88, https://doi.org/10.1016/j.agrformet.2005.07.012.
Ngo-Duc, T., J. Polcher, and K. Laval, 2005: A 53-year forcing data set for land surface models. J. Geophys. Res., 110, D06116, https://doi.org/10.1029/2004JD005434.
Nijssen, B., and Coauthors, 2014: A prototype global drought information system based on multiple land surface models. J. Hydrometeor., 15, 1661–1676, https://doi.org/10.1175/JHM-D-13-090.1.
Panu, U. S., and T. C. Sharma, 2002: Challenges in drought research: Some perspectives and future directions. Hydrol. Sci. J., 47, S19–S30, https://doi.org/10.1080/02626660209493019.
Qian, T., A. Dai, K. E. Trenberth, and K. W. Oleson, 2006: Simulation of global land surface conditions from 1948 to 2004. Part I: Forcing data and evaluations. J. Hydrometeor., 7, 953–975, https://doi.org/10.1175/JHM540.1.
Qian, W., X. Shan, and Y. Zhu, 2011: Ranking regional drought events in China for 1960–2009. Adv. Atmos. Sci., 28, 310–321, https://doi.org/10.1007/s00376-009-9239-4.
Raupach, M. R., 1994: Simplified expressions for vegetation roughness length and zero-plane displacement as functions of canopy height and area index. Bound.-Layer Meteor., 71, 211–216, https://doi.org/10.1007/BF00709229.
Robock, A., K. Y. Vinnikov, G. Srinivasan, J. K. Entin, S. E. Hollinger, N. A. Speranskaya, S. Liu, and A. Namkhai, 2000: The Global Soil Moisture Data Bank. Bull. Amer. Meteor. Soc., 81, 1281–1299, https://doi.org/10.1175/1520-0477(2000)081<1281:TGSMDB>2.3.CO;2.
Robock, A., M. Mu, K. Vinnikov, I. V. Trofimova, and T. I. Adamenko, 2005: Forty-five years of observed soil moisture in the Ukraine: No summer desiccation (yet). Geophys. Res. Lett., 32, L03401, https://doi.org/10.1029/2004GL021914.
Rodell, M., and Coauthors, 2004: The Global Land Data Assimilation System. Bull. Amer. Meteor. Soc., 85, 381–394, https://doi.org/10.1175/BAMS-85-3-381.
Rui, H., and H. Beaudoing, 2014: README document for the Global Land Data Assimilation System version 2 (GLDAS-2) products. NASA Goddard Earth Sciences Data and Information Services Center Rep., 22 pp., ftp://hydro1.sci.gsfc.nasa.gov/data/s4pa/GLDAS/README.GLDAS2.pdf.
Seneviratne, S. I., and Coauthors, 2010: Investigating soil moisture–climate interactions in a changing climate: A review. Earth-Sci. Rev., 99, 125–161, https://doi.org/10.1016/j.earscirev.2010.02.004.
Sheffield, J., and E. F. Wood, 2007: Characteristics of global and regional drought, 1950–2000: Analysis of soil moisture data from off-line simulation of the terrestrial hydrologic cycle. J. Geophys. Res., 112, D17115, https://doi.org/10.1029/2006JD008288.
Sheffield, J., and E. F. Wood, 2008a: Global trends and variability in soil moisture and drought characteristics, 1950–2000, from observation-driven simulations of the terrestrial hydrologic cycle. J. Climate, 21, 432–458, https://doi.org/10.1175/2007JCLI1822.1.
Sheffield, J., and E. F. Wood, 2008b: Projected changes in drought occurrence under future global warming from multi-model, multi-scenario, IPCC AR4 simulations. Climate Dyn., 31, 79–105, https://doi.org/10.1007/s00382-007-0340-z.
Sheffield, J., G. Goteti, F. H. Wen, and E. F. Wood, 2004: A simulated soil moisture based drought analysis for the United States. J. Geophys. Res., 109, D24108, https://doi.org/10.1029/2004JD005182.
Sheffield, J., G. Goteti, and E. F. Wood, 2006: Development of a 50-year high-resolution global dataset of meteorological forcings for land surface modeling. J. Climate, 19, 3088–3111, https://doi.org/10.1175/JCLI3790.1.
Shinoda, M., and Y. Yamaguchi, 2003: Influence of soil moisture anomaly on temperature in the Sahel: A comparison between wet and dry decades. J. Hydrometeor., 4, 437–447, https://doi.org/10.1175/1525-7541(2003)4<437:IOSMAO>2.0.CO;2.
Shukla, S., A. C. Steinemann, and D. P. Lettenmaier, 2011: Drought monitoring for Washington State: Indicators and applications. J. Hydrometeor., 12, 66–83, https://doi.org/10.1175/2010JHM1307.1.
Spennemann, P. C., J. A. Rivera, A. C. Saulo, and O. C. Penalba, 2015: A comparison of GLDAS soil moisture anomalies against standardized precipitation index and multisatellite estimations over South America. J. Hydrometeor., 16, 158–171, https://doi.org/10.1175/JHM-D-13-0190.1.
Stéfanon, M., P. Drobinski, F. D’Andrea, C. Lebeaupin-Brossier, and S. Bastin, 2014: Soil moisture–temperature feedbacks at meso-scale during summer heat waves over western Europe. Climate Dyn., 42, 1309–1324, https://doi.org/10.1007/s00382-013-1794-9.
Vautard, R., and Coauthors, 2007: Summertime European heat and drought waves induced by wintertime Mediterranean rainfall deficit. Geophys. Res. Lett., 34, L07711, https://doi.org/10.1029/2006GL028001.
Ventura, F., O. Facini, S. Piana, and P. Rossi Pisa, 2010: Soil moisture measurements: Comparison of instrumentation performances. J. Irrig. Drain. Eng., 136, 81–89, https://doi.org/10.1061/(ASCE)0733-9437(2010)136:2(81).
Vicente-Serrano, S. M., S. Begueria, and J. I. López-Moreno, 2010a: A multiscalar drought index sensitive to global warming: The Standardized Precipitation Evapotranspiration Index. J. Climate, 23, 1696–1718, https://doi.org/10.1175/2009JCLI2909.1.
Vicente-Serrano, S. M., S. Begueria, J. I. López-Moreno, M. Angulo, and A. El Kenawy, 2010b: A new global 0.5° gridded dataset (1901–2006) of a multiscalar drought index: Comparison with current drought index datasets based on the Palmer drought severity index. J. Hydrometeor., 11, 1033–1043, https://doi.org/10.1175/2010JHM1224.1.
Vicente-Serrano, S. M., S. Begueria, and J. I. López-Moreno, 2011a: Comment on “Characteristics and trends in various forms of the Palmer drought severity index (PDSI) during 1900–2008” by Aiguo Dai. J. Geophys. Res., 116, D19112, https://doi.org/10.1029/2011JD016410.
Vicente-Serrano, S. M., and Coauthors, 2011b: A multiscalar global evaluation of the impact of ENSO on droughts. J. Geophys. Res., 116, D20109, https://doi.org/10.1029/2011JD016039.
Vicente-Serrano, S. M., and Coauthors, 2012: Performance of drought indices for ecological, agricultural, and hydrological applications. Earth Interact., 16, https://doi.org/10.1175/2012EI000434.1.
Wang, A., D. P. Lettenmaier, and J. Sheffield, 2011: Soil moisture drought in China, 1950–2006. J. Climate, 24, 3257–3271, https://doi.org/10.1175/2011JCLI3733.1.
Wang, G. L., 2005: Agricultural drought in a future climate: Results from 15 global climate models participating in the IPCC 4th assessment. Climate Dyn., 25, 739–753, https://doi.org/10.1007/s00382-005-0057-9.
Wang, W., Y. Zhu, R. Xu, and J. Liu, 2015: Drought severity change in China during 1961–2012 indicated by SPI and SPEI. Nat. Hazards, 75, 2437–2451, https://doi.org/10.1007/s11069-014-1436-5.
Wang, Y.-P., and R. Leuning, 1998: A two-leaf model for canopy conductance, photosynthesis and partitioning of available energy I: Model description and comparison with a multi-layered model. Agric. For. Meteor., 91, 89–111, https://doi.org/10.1016/S0168-1923(98)00061-6.
Wang, Y.-P., and J. L. McGregor, 2003: Estimating regional terrestrial carbon fluxes for the Australian continent using a multiple-constraint approach II. The atmospheric constraint. Tellus, 55B, 290–304, https://doi.org/10