The NASMD is a high-quality observational soil moisture database that includes over 1,800 stations to support drought, land–atmosphere, and satellite/model validation studies.
Soil moisture is an important state variable in the climate system, as it stimulates land–atmosphere interactions by modifying energy and wetness fluxes in the boundary layer (Legates et al. 2011). Soil water content influences evapotranspiration and corresponding near-surface atmospheric moisture availability (Pal and Eltahir 2001). Dry soil can induce and amplify warm and dry conditions, especially during the summer, by reducing local evaporation and modifying patterns of moisture convergence/divergence (Namias 1991). Soil moisture has been connected with the partitioning of surface energy fluxes (Dirmeyer et al. 2000; Guo and Dirmeyer 2013), near-surface atmospheric temperature (Hirschi et al. 2011; Teuling et al. 2010), planetary boundary layer instability (Myoung and Nielsen-Gammon 2010; Gentine et al. 2013), and the onset and location of afternoon convective precipitation (Findell et al. 2011; Taylor et al. 2012). Soil moisture is used for drought monitoring and in drought early warning systems in Asia (Wang et al. 2011; Tei et al. 2013), Africa (Anderson et al. 2012; Yuan et al. 2013), Australia (Cai et al. 2009), Europe (Zampieri et al. 2009; Mozny et al. 2012), North America (Tang and Piechota 2009; Bolten et al. 2010), and South America (Markewitz et al. 2010). However, despite the importance of soil moisture and its utility for drought monitoring, there are relatively few in situ soil moisture observations, especially in comparison to precipitation and temperature observations.
The lack of in situ soil moisture measurements means that most studies of the interactions between soil moisture and the atmosphere, biosphere, and hydrosphere are based on models. For example, Schubert et al. (2004) investigated the causes of droughts in the United States Great Plains using a general circulation model forced with observed SSTs and found that approximately two-thirds of the low-frequency rainfall variability can be explained by land–atmosphere interactions (e.g., soil moisture), while the remaining variance can be attributed to SST anomalies. In contrast, an observation-based study by Findell and Eltahir (1997) attributed only ∼16% of the variance in summer precipitation to spring soil moisture conditions. Without accurate in situ soil moisture measurements, such discrepancies remain unresolved. GCMs have also been used to investigate soil moisture–climate interactions. The Global Land–Atmosphere Coupling Experiment (GLACE) used a 12-GCM ensemble to identify regions with strong soil moisture–climate (Koster et al. 2004, 2006). They found significant variations in the coupling strength amongst the GCMs and therefore multimodel ensembles are commonly used to study land–atmosphere interactions. Given the difficulty that GCMs and land surface models have in accurately simulating soil moisture, observed soil moisture (either in situ or satellite) is commonly used to initialize or constrain, through data assimilation, these models (e.g., Harrison et al. 2012; Kumar et al. 2012).
Soil moisture observations are also important for the validation of satellite soil moisture retrievals from missions such as the Soil Moisture Ocean Salinity (SMOS) satellite. Al Bitar et al. (2012) evaluated SMOS soil moisture estimates with in situ soil moisture observations from Soil Climate Network (SCAN) and Snowpack Telemetry (SNOTEL) observation networks in several regions of the United States. Their results demonstrated that the accuracy of SMOS-derived soil moisture varied significantly from site to site, necessitating the validation of satellite-derived soil moisture in a variety of regions. Collow et al. (2012) evaluated the accuracy of SMOS-derived soil moisture with in situ measurements in the United States Great Plains. They concluded that the lack of uniform soil moisture measurements makes evaluating SMOS difficult and therefore additional stations are needed to provide a more robust evaluation of satellite-derived soil moisture. Of course, it should be noted that there are significant scaling issues involved in comparing in situ soil moisture measurements (a point) to satellite-derived soil moisture (50-km pixel) and most in situ networks are not sufficiently dense to adequately resolve soil moisture variability within each satellite pixel.
Despite the importance of soil moisture in the climate system, relatively little work has been done to assemble and homogenize in situ soil moisture measurements and to utilize these measurements for investigating land–atmosphere interactions. Robock et al. (2000) developed the Global Soil Moisture Data Bank, providing soil moisture observations from 25 stations in the United States. The Global Soil Moisture Data Bank has since been incorporated into the International Soil Moisture Network (ISMN, www.ipf.tuwien.ac.at/insitu). ISMN is a global database of in situ soil moisture observations, containing data from 47 networks and more than 1,900 stations located in North America, Europe, Asia, and Australia (Dorigo et al. 2011).
Development of the North American Soil Moisture Database (NASMD, http://soilmoisture.tamu.edu/) began in 2011 with funding from the National Science Foundation to support the study of land–atmosphere interactions. The NASMD was developed to provide harmonized and quality-controlled soil moisture data for scientists and decision makers. For example, these data have utility for 1) improving our understanding of land–atmosphere interactions (Ford et al. 2014b, 2015a,b); 2) developing seasonal to decadal climate forecasting tools (Ford and Quiring 2013, 2014a); 3) calibrating, validating, and improving the physical parameterizations in regional and global land surface models (Xia et al. 2015a,b); 4) developing and validating satellite-derived soil moisture algorithms (Ford et al. 2014a); and 5) monitoring and detecting climate variability and change in this key hydrological variable (Khong et al. 2015).
Although ISMN and NASMD are similar in that their primary purpose is to aggregate, quality control, and disseminate soil moisture measurements, there are a number of important differences. The first is geographic focus; ISMN is a global database, while NASMD is focused on North America. A second difference is station density. The goal of the NASMD was to develop the densest possible network of in situ soil moisture in North America. A great deal of time was invested in uncovering soil moisture networks and datasets that had not been previously published or utilized in land–atmosphere studies. In some cases this involved digitizing soil moisture data that were only previously available in hardcopy (e.g., Khong et al. 2015). The NASMD has integrated data from 33 observation networks and two shorter-term soil moisture monitoring campaigns comprising over 1,800 observation sites in the United States, Canada, and Mexico. Although the NASMD includes data from some of the same networks as ISMN, it includes many networks that are not part of ISMN and so NASMD has approximately twice as many stations in North America as ISMN. A third difference between the two databases is that NASMD was initially designed to be a retrospective database for studying land–atmosphere interactions. ISMN was developed to support satellite calibration and validation activities for SMOS and it supports near-real-time updates (Dorigo et al. 2011). Finally, observations from networks measuring soil moisture at subdaily scales are aggregated to a daily resolution in the NASMD, while ISMN provides hourly data. Both NASMD and ISMN are heterogeneous in terms of measurement technique, measurement depth, spatial extent, and degree of automation. In addition, both ISMN and NASMD apply automated quality control algorithms to the all of the soil moisture measurements they receive.
Much of our understanding of land–atmosphere interactions has been informed by land surface and regional climate models; however, these models are difficult to validate because of the lack of observations. Soil moisture databases, like NASMD and ISMN, provide data for validation of land surface model output (Zhang and Wegehenkel 2006; Jiang et al. 2009; Meng and Quiring 2010; Tang et al. 2012) and satellite soil moisture retrievals (Jackson et al. 2012; Al Bitar et al. 2012; Rowlandson et al. 2012; Collow et al. 2012). Therefore, soil moisture databases are important for increasing our understanding of the climate system. This paper describes how development of the NASMD, including quality control/quality assurance, standardization, and collection of metadata. The utility of the NASMD is demonstrated through an analysis of the inter- and intraannual variability of soil moisture from multiple networks, and we conclude the article by highlighting some new developments and soil moisture products.
NASMD DATA.
Soil moisture data sources.
The NASMD has integrated data from 33 observation networks and two shorter-term soil moisture monitoring campaigns comprising over 1,800 observation sites in the United States and Canada (Table 1). Several other networks, representing over 500 observation sites in the United States, Canada, and Mexico, have agreed to contribute data in the near future. Figure 1 shows the location of sites currently available in the NASMD (soilmoisture.tamu.edu). Once we receive data from all of the networks that have agreed to provide it, the NASMD will have sites in all 50 states and six Canadian provinces, covering a wide range of soil texture, land cover, elevation, and climate conditions. The NASMD is currently the largest collection of in situ soil moisture observations in North America and there are >8 million observations in the database.
List of selected networks (21 out of 34) that are part of the NASMD including the number of stations, period of record, and the depths at which soil moisture is measured.
Map of observation sites that have contributed data to the NASMD.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Map of observation sites that have contributed data to the NASMD.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Map of observation sites that have contributed data to the NASMD.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Metadata.
Metadata have been collected for each site, including location, county, state, parent observation network, depths at which soil moisture are measured, type of soil moisture sensor, and the sampling frequency. In addition, soil characteristics such as bulk density, texture, percent sand/silt/clay, and hydraulic conductivity are reported at each depth that soil moisture is measured. Soil texture information from site-specific soil surveys are available for just over 1,000 of the stations included in the NASMD (∼69% of all of the stations). Soil characteristics for the remaining sites are obtained from the National Cooperative Soil Survey (NRCS) Soil Survey Geographic Database (SSURGO).
The NASMD uses the land cover classification scheme provided by the Environmental Protection Agency’s National Land Cover Dataset (NLCD) 2001 (www.epa.gov/mrlc/classification.html) to identify the land cover at each site. If land cover information is reported by the parent observation network, it is used to identify the relevant land cover type in the NLCD classification system. Approximately 500 sites (approximately 36% of NASMD sites) provide land use and land cover (LULC) information. For the remaining sites, LULC has determined by NASDM staff using either site photos or using high-resolution satellite imagery such as Google Earth. Table 2 lists all of the parameters reported in the NASMD metadata as well as the metadata sources.
List of observation site properties that are included in the NASMD metadata as well as the unit and source of each parameter.
If information on soil characteristics were not collected at the site, these parameters were estimated from the United States Department of Agriculture’s SSURGO database (Reybold and TeSelle 1989). SSURGO provides soil texture and hydraulic parameter information at multiple column depths for the entire contiguous United States. In addition, Leib et al. (2003) evaluated several different soil moisture sensor estimates under alfalfa crop and showed that although the sensor trends were similar, the magnitudes of sensor estimates varied considerably. The authors concluded that sensor-specific calibration is necessary to obtain a high degree of soil-moisture-estimation accuracy. Thus, sensor change or recalibration dates are included in the NASMD metadata if these were available from the observation network.
Data integration and harmonization.
Soil moisture observation networks use many different types of sensors to measure soil wetness. The NASMD reports all data as volumetric soil water content (q), which represents the ratio of the volume of water in a given soil column to the total soil column volume. Here q is influenced significantly by site-specific characteristics such as soil texture and land cover and thus should not be directly compared across space. Many of the networks that have been incorporated into the NASMD measure soil wetness at subhourly to daily time scales. However, all of the soil moisture observations in the NASMD have been resampled to daily resolution because our goal is to provide a harmonized soil moisture dataset (i.e., harmonized with respect to measurement units, time step, metadata, and QC procedures) to support a variety of applications. Resampling all the data to daily resolution results in a loss of information and it may not be ideal for all applications. For example, Ford et al. (2015a) uses soil moisture data from the morning to examine whether afternoon convective precipitation occurs preferentially over wet or dry soils. Their study would not have been possible with daily data. However, for many applications such as drought monitoring and model/satellite validation, daily data are appropriate and, in many cases, preferable.
Figure 2 illustrates the NASMD data processing procedure. First, raw data from all observation networks are ingested and, if necessary, data are converted to q and resampled to daily resolution. All data that are provided to the NASMD are evaluated by our quality control (QC) algorithm to identify dubious or questionable values. These values are flagged and/or removed from the dataset. Small gaps (<10 days) in the data are then filled using the procedure described in Ford and Quiring (2014b) and the data (and flags) are stored in the network database (Fig. 3). Users are able to access all of the quality-controlled data for each station as well as the station metadata through the web interface (soilmoisture.tamu.edu).
NASMD data processing and integration schematic: an overview of the data flows, from the observation network to the user.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
NASMD data processing and integration schematic: an overview of the data flows, from the observation network to the user.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
NASMD data processing and integration schematic: an overview of the data flows, from the observation network to the user.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Soil moisture plots from 5- and 25-cm depths in Acme, Oklahoma, in 2000. There are two plots of the soil moisture data from each depth. (a),(c) The soil moisture data before it has been filled by the DAR procedure and (b),(d) the soil moisture data after infilling.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Soil moisture plots from 5- and 25-cm depths in Acme, Oklahoma, in 2000. There are two plots of the soil moisture data from each depth. (a),(c) The soil moisture data before it has been filled by the DAR procedure and (b),(d) the soil moisture data after infilling.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Soil moisture plots from 5- and 25-cm depths in Acme, Oklahoma, in 2000. There are two plots of the soil moisture data from each depth. (a),(c) The soil moisture data before it has been filled by the DAR procedure and (b),(d) the soil moisture data after infilling.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
QAQC ALGORITHM.
Quality control procedures are commonly applied to a wide variety of climate and environmental datasets such as air temperature and precipitation (Hubbard et al. 2005), solar radiation (Journée and Bertrand 2011), sea surface temperatures (Merchant et al. 2008), and ocean salinity (Ingleby and Huddleston 2007); however, considerably fewer have focused on quality control of in situ soil moisture (Illston et al. 2008; You et al. 2010; Dorigo et al. 2013). Soil moisture is one of the most difficult variables to validate with quality assurance tests because it is influenced by multiple meteorological variables and the physical and chemical properties of the soil (Fiebrich et al. 2010). Thus, a robust quality control procedure for soil moisture must include multiple, complementary data flagging techniques to assess both the magnitude and variability of the soil moisture data. Quality control algorithms are employed by a number of soil moisture monitoring networks. For example, the West Texas Mesonet’s QC method includes tests for absolute magnitude, measurement-to-measurement variability, observation persistence, and spatial coherence (Schroeder et al. 2005) and similar quality control algorithms are also used by the Oklahoma Mesonet (Shafer et al. 2000; McPherson et al. 2007) and the Nebraska Automated Weather Data Network (AWDN; Hubbard et al. 2005).
It is reasonable to ask whether it is necessary for the NASMD to develop a separate QC procedure when a number of the networks that contribute data to the NASMD already perform some type of QC. It is also reasonable to ask whether the errors in the soil moisture data will have a significant impact on the value of soil moisture data for practical applications and scientific studies. These questions were addressed by Xia et al. (2015c). They developed an automated soil moisture QC method for a subset of stations from the NASMD using the North American Land Data Assimilation System phase 2 (NLDAS-2) Noah soil porosity, soil temperature, and fraction of liquid and total soil moisture. Overall, the results of Xia et al. (2015c) demonstrate that it is necessary to perform QC on all of the soil moisture data, even for networks that already have their own QC routines. They also found that removing the soil moisture measurements that were flagged by their QC procedure had a significant positive impact on the agreement between NLDAS-2 model-simulated and observed soil moisture, especially in Alabama, North Carolina, and west Texas. Therefore, based on the results of Xia et al. (2015c), we conclude it is necessary to develop an appropriate quality control procedure for the NASMD and that this procedure should be applied to all stations.
Soil moisture is influenced by several physical, site-specific characteristics including soil properties (e.g., texture, organic matter, structure), land cover, and climate conditions (e.g., precipitation, temperature, solar radiation). Thus, the ideal QC method uses site-specific characteristics to assess the data validity. You et al. (2010) describe examples of such techniques employed by the AWDN when quality controlling soil moisture data such as the precipitation-and-irrigation-based method, which evaluates whether changes in soil moisture are accompanied by precipitation or irrigation. Similarly, the precipitation and irrigation amount test assesses whether a single time step rise in soil moisture exceeds the total precipitation plus irrigation amount (You et al. 2010). Such robust tests provide physically based methods of data validation instead of the purely statistical methods employed by many (Durre et al. 2010; Dorigo et al. 2011). However, these techniques can only be applied when ancillary data such as temperature, precipitation, irrigation (if applicable), and soil porosity are available. Unfortunately, the majority of stations in the NASMD do not have enough ancillary data to use physically based techniques and so only statistical-based approaches are used to QC the NASMD.
We have chosen to utilize the same QC algorithm for all stations in the NASMD because our goal is to provide a harmonized soil moisture dataset (i.e., harmonized with respect to measurement units, time step, metadata, and QC procedures). We believe that there is significant value in building a harmonized dataset and that it has the greatest utility (and is most appropriate) for those who are interested in using data from the NASMD since many applications require data from multiple soil moisture networks. We also make the original data available so that those who are interested in developing their own QC procedures, or using data that have not undergone QC, may do so.
Soil moisture data are processed before entering the NASMD QC algorithm to convert data from different observation networks into the same format. Because different observation networks measure soil moisture at different soil depths, data from each network are processed separately in the QC procedure. The algorithm was developed as a set of MATLAB scripts, which, together, take approximately 10–20 min to process a network’s worth of data, depending on the size and time length of the network observations. The NASMD QC procedure was developed similarly to the algorithms used successfully by the Oklahoma Mesonet, West Texas Mesonet, and AWDN. The algorithm includes tests for soil moisture range, persistence, magnitude, and variability.
Range/integrity test.
Volumetric soil water content represents the volume of water contained in a specified soil column divided by the total soil column volume and, thus, cannot be less than 0 or exceed 1. In reality, volumetric water content cannot physically exceed soil porosity, with the rare exception of supersaturated soils. The range test thus removes any values that do not fall within the range of 0 to 0.6. This test is similar to the range test that is described in Hubbard et al. (2005) and is used by the AWDN for soil moisture data validation. The integrity test is similar to that described by Durre et al. (2010) employed by the Global Historical Climatology Network (GHCN). This procedure checks for sampling day replication and validates the date labels as well as each site’s reported latitude and longitude. The integrity test’s purpose is to ensure that in subsequent QAQC steps, data will only be flagged or removed as a function of the data itself and not labeling or conversion issues.
Streak (persistence) test.
The streak test assesses soil moisture variability over time. Soil moisture observations are removed if the same value is recorded every day over a >10-day period. Considering soil moisture’s highly variable nature, sensor records of the same value over such a long period is assumed to be sensor failure and not valid data. The 10-day threshold was selected examining a range of thresholds ranging from 5 to 50 days using data from several networks and geographic regions. We acknowledge that the variability of daily volumetric water content is a strong function of climate and soil texture (and as sensor type). However, even when conditions are very wet or very dry, there are small variations in the soil moisture measurements (especially given that it is measured to the thousandth decimal). For example, even deep in the soil profile (100 cm) in an arid location like Walnut Gulch, Arizona, there are small day-to-day variations in soil moisture and rarely are the exact water content values reported on two consecutive days (results not shown).
Similar to the persistence test described by Shafer et al. (2000) used by the Oklahoma Mesonet, our streak test cannot discern when the sensor failed and thus the entire streak of similar values is flagged.
Deviance (magnitude) test.
The deviance test assesses if the absolute magnitude of a soil moisture measurement is valid based on previous measurements during that period of time. The method calculates the mean and standard deviation of a 30-day window surrounding each day of the year over all years of measurement. The daily measurement in question is then converted to a z score by subtracting the window mean and dividing by the window standard deviation. Daily observations are flagged if their respective z score is greater than 3. Essentially the procedure flags observations that deviate more than three times the standard deviation, calculated from average soil moisture conditions typical of that period of time. The deviance test is similar to the outlier check described by Durre et al. (2010) and used by the GHCN. Given that 3 standard deviations covers 99.73% of the distribution, it is possible that using this threshold will occasionally result in rejecting a true extreme value. However, based on our testing, this threshold appears to perform well and does not result in too many true measurements being rejected.
Soil moisture at shallower depths exhibits considerable intraannual variability; however, interannual variability is typically much less. Table 3 displays average daily coefficient of variation (CV) at nine observation networks between measurement years, calculated using the same 30-day window employed in the deviance test. Minimum, maximum, and mean CV values are reported, representing the depth at which the lowest and highest variability was observed, as well as the average variability between all depths. The West Texas Mesonet data exhibit the highest variability from year to year; however, the maximum variability was still only 0.34, meaning that the data standard deviation is one-third of the 30-day window mean.
Daily coefficient of variation calculated for interannual variability using 30-day window. Minimum and maximum CV is reported for the depth at which the lowest and highest soil moisture variability are observed, respectively. The depths at which these measurements were made are shown in brackets for minimum and maximum CV. The number of measurement depths that were averaged is shown in brackets for mean CV.
Step test.
The step test assesses the change in magnitude between consecutive measurements. The procedure calculates the average and standard deviation of the difference between consecutive measurements for each site. Similar to the deviance test, each daily “step” value is converted into a z score using the magnitude difference average and standard deviation. Observations are flagged if their respective z score is greater than 3.
Observations that are flagged by the streak (persistence) test are immediately removed; however, the deviance and step tests are complimentary such that an observation is only removed if it is flagged by both tests. The deviance and step tests work off of each other because of the nature of extremely dry or wet events. The deviance test flags extreme events that diverge from the time period mean by more than three times the standard deviation. Thus, if an extreme precipitation event occurs, the likelihood of the deviance test flagging the corresponding soil moisture observation is increased. However, the aforementioned soil moisture observation is not removed if the step test shows that the next daily value is also considerably higher than normal. This is the case as the “normal” soil moisture response to heavy precipitation is a dramatic increase with a gradual decrease (Illston et al. 2008). Any anomalously high soil moisture observation is removed only if the previous and subsequent observations are at or below normal, suggesting a sensor failure instead of a large precipitation event.
Figure 4 shows 25-cm soil moisture data from October 2003 to September 2004 at Ashton, Kansas. The data that were not flagged by the deviance and step tests are shown in blue, while the data that were flagged are shown in red. The flagged data show a rapid soil moisture increase, well above the deviance test threshold, possibly the cause of a strong precipitation event. However, after a couple of days, soil moisture decreases nearly as rapidly. This is not the normal response of a drying soil and is more indicative of sensor failure. Thus, these data were also flagged by the step test and subsequently removed. A month later, soil moisture similarly increases rapidly, but subsequently decreases gradually in a way that is much more indicative of a drying soil. Since these data were not flagged by both the deviance and step tests, they were not removed.
Soil moisture data from the 25-cm depth at Ashton, Kansas. Data shown in blue were not flagged by the deviance and step tests while data in red were flagged.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Soil moisture data from the 25-cm depth at Ashton, Kansas. Data shown in blue were not flagged by the deviance and step tests while data in red were flagged.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Soil moisture data from the 25-cm depth at Ashton, Kansas. Data shown in blue were not flagged by the deviance and step tests while data in red were flagged.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
The four validation tests that constitute the NASMD QAQC algorithm—the range, streak, deviance, and step tests—flagged 0.67%, 0.52%, 4.40%, and 3.22%, respectively, of the >8 million values soil moisture data that have been assessed by the QC procedure. By way of comparison, the soil moisture quality control methodology developed by Dorigo et al. (2013) flagged approximately 13% of the soil moisture measurements in the ISMN.
APPLICATIONS.
Soil moisture exhibits substantial variability on daily, monthly, seasonal, and interannual time scales (Illston et al. 2004). Soil moisture persistence ranges from several days to several months depending on the soil layer depth and overlying climatic conditions (Georgakakos and Bae 1994; Wu et al. 2002; Wu and Dickinson 2004). Seasonal soil moisture anomalies have been shown in models to influence weather patterns and precipitation in subsequent seasons (Oglesby and Erickson 1989; Meng and Quiring 2010; Hirschi et al. 2011). We use in situ soil moisture observations from several networks contained within the NASMD to explore the variability of soil moisture on intraannual and interannual time scales.
Intraannual variability.
Soil moisture data from five networks, totaling 196 stations across the United States, are used to examine intraannual variability. Figure 5 shows a map of these networks, specifically the Atmospheric Radiation Measurement (ARM), the Delaware Environment Observing System (DEOS), the North Carolina Environment and Climate Observing Network (ECONET), the Illinois Climate Network (ICN), and the Michigan Automated Weather Network (MAWN). These networks were chosen for their data completeness and record length. Soil moisture measurement depths vary depending on the network (Table 1). To assess intraannual soil moisture variability, daily volumetric soil water content data from all sites within each network were plotted together spanning 1 January to 31 December. Soil moisture is occasionally measured in frozen soils for the MAWN and ICN networks, which typically results in unrealistically low soil moisture measurements. We identify frozen soils using daily soil temperature measurements from these networks. As described above, the NASMD QC algorithm does not utilize soil temperature data to flag observations because soil temperatures are not available for all networks. For the purposes of this analysis, daily soil moisture observations taken in frozen soils are removed before analyzing intraannual variability. Figures 6a–d show intraannual variability of near-surface (5 to 10 cm) soil moisture from ARM, DEOS, ICN, and MAWN networks, respectively. The ECONET network only measures soil moisture at 20 cm and therefore it is not included. Figures 6a, 6b, and 6c show that there is a strong seasonal cycle in the near-surface soil moisture at ARM, ICN, and DEOS stations. Soil moisture observations from the first three networks show maximum soil moisture values during the late winter and early spring and this is followed by a consistent period of drying during the summer. These patterns are similar to those found in Illinois by Hollinger and Isard (1994) and in Oklahoma by Illston et al. (2004).
Observing networks used for analysis of intraannual and interannual variability. Data from a total of 196 stations were used in the analysis.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Observing networks used for analysis of intraannual and interannual variability. Data from a total of 196 stations were used in the analysis.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Observing networks used for analysis of intraannual and interannual variability. Data from a total of 196 stations were used in the analysis.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Near-surface soil moisture from (a) ARM, (b) DEOS, (c) ICN, and (d) MAWN. Each line represents one station during 1 year of measurement.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Near-surface soil moisture from (a) ARM, (b) DEOS, (c) ICN, and (d) MAWN. Each line represents one station during 1 year of measurement.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Near-surface soil moisture from (a) ARM, (b) DEOS, (c) ICN, and (d) MAWN. Each line represents one station during 1 year of measurement.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Figure 6d shows the intraannual variability of soil moisture measured at 10 cm for the MAWN stations that are located in Michigan and Door County, Wisconsin. Data from MAWN exhibits higher interstation variability and less seasonal variation than the other four networks. One possible explanation is that land cover over MAWN network sites is more diverse than other networks. For example, the majority of ARM and ICN sites are surrounded by grassland landscapes while MAWN sites’ land cover ranges from mixed forest to bog. These patterns may also arise because of differences in the time period, soil characteristics, climate conditions, instrumentation, and depth of measurements between the networks.
Figures 7a–d show similar plots as Fig. 6 only for deeper measurements of soil moisture. Data are plotted from the ARM network at 35 cm, the ECONET network at 20 cm, the ICN network at 50 cm, and the MAWN network at 25 cm. Seasonal variability of deeper soil moisture is similar to that of the near surface (Fig. 6); however, the signal is somewhat dampened. This agrees with the findings of Wu et al. (2002), who found that the amplitude of Illinois soil moisture response to precipitation variability decreased with soil depth. In general, soil moisture at three of the four networks shows considerable seasonal variability. Our results illustrate that soil moisture can vary significantly within networks and between networks owing to the large number of factors that influence soil moisture.
Middle layer soil moisture from (a) ARM 35 cm, (b) ECONET 20 cm, (c) ICN 50 cm, and (d) MAWN 25 cm. Each line represents one station during 1 year of measurement.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Middle layer soil moisture from (a) ARM 35 cm, (b) ECONET 20 cm, (c) ICN 50 cm, and (d) MAWN 25 cm. Each line represents one station during 1 year of measurement.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Middle layer soil moisture from (a) ARM 35 cm, (b) ECONET 20 cm, (c) ICN 50 cm, and (d) MAWN 25 cm. Each line represents one station during 1 year of measurement.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Interannual variability.
Data contained in the NASMD can also be used to examine the interannual variability of soil moisture across North America. Soil moisture variability, both seasonal and interannual, is tied to precipitation variability. Therefore, soil moisture interannual variability over a number of years can represent drought variability in that particular region. Stations from the ARM and ICN networks were used to analyze soil moisture interannual variability following the approach of Fan et al. (2011). Figure 8 shows mean monthly soil moisture anomalies (departures from the long-term mean) averaged over all sites within each network. Soil moisture anomalies represent relative soil wetness with respect to normal conditions for that location and time of year and they are useful for drought monitoring, especially in regions with heterogeneous climate and soil conditions. Monthly soil moisture anomalies, as shown in Fig. 8, can be used to examine the development and progression of soil moisture–drought. Figure 8 shows how drought conditions develop and tend to move down through the soil profile. It also depicts how soils recover from drought conditions and illustrates that soil moisture can vary significantly with depth. For example, Fig. 8a shows a sudden onset of below normal soil moisture in the latter half of 1999. These drier-than-normal conditions appear relatively consistent throughout the soil profile during the first month of the drought. Interestingly, the drought event in 1999/2000 appears most pronounced between 50 and 100 cm in the soil and the soil moisture anomalies near that surface and at depth are somewhat smaller. Conditions improve in the first half of 2000 and some wet anomalies are seen near the surface; however, this improvement is not felt at depth and the soil moisture conditions below 100 cm do not recover until 2002. The recovery in 2002 shows that although the near-surface soil moisture recovers right away, it takes a number of months of above normal rainfall (not shown) before the deeper layers of the soil recover. This can be seen by the slope of the red contours (e.g., the –4 contour) in early 2002. In contrast, the onset of the 2006 drought event was quite different (Fig. 8b). During the early part of the 2006 event there are large negative soil moisture anomalies in the upper 70 cm of the soil, but below that the soil is wetter than normal. As the dry conditions persist, the entire soil profile continues to dry out, particularly during the summer of 2006. Once these dry conditions reach 150 cm, they persist for almost a year. Again, the recovery of the deeper layers lags the near surface by several months.
Observed vertical profile of column soil moisture anomalies (mm of soil water/10) for ARM and ICN networks: (a) ARM soil moisture anomalies for 1998 to 2004, (b) ARM soil moisture anomalies for 2004 to 2010, (c) ICN soil moisture anomalies for 2004 to 2008, and (d) ICN soil moisture anomalies for 2008 to 2011. ARM soil moisture anomalies are calculated using 1998 to 2010 to calculate mean soil water content at each depth: 5, 15, 25, 35, 60, 85, 125, and 175 cm. ICN soil moisture anomalies are calculated using 2004 to 2011 to calculate mean soil water content at each depth: 5, 10, 50, 100, and 150 cm. Positive anomalies (wetter-than-normal conditions) are shown in blue and negative anomalies (drier-than-normal conditions) are shown in red.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Observed vertical profile of column soil moisture anomalies (mm of soil water/10) for ARM and ICN networks: (a) ARM soil moisture anomalies for 1998 to 2004, (b) ARM soil moisture anomalies for 2004 to 2010, (c) ICN soil moisture anomalies for 2004 to 2008, and (d) ICN soil moisture anomalies for 2008 to 2011. ARM soil moisture anomalies are calculated using 1998 to 2010 to calculate mean soil water content at each depth: 5, 15, 25, 35, 60, 85, 125, and 175 cm. ICN soil moisture anomalies are calculated using 2004 to 2011 to calculate mean soil water content at each depth: 5, 10, 50, 100, and 150 cm. Positive anomalies (wetter-than-normal conditions) are shown in blue and negative anomalies (drier-than-normal conditions) are shown in red.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Observed vertical profile of column soil moisture anomalies (mm of soil water/10) for ARM and ICN networks: (a) ARM soil moisture anomalies for 1998 to 2004, (b) ARM soil moisture anomalies for 2004 to 2010, (c) ICN soil moisture anomalies for 2004 to 2008, and (d) ICN soil moisture anomalies for 2008 to 2011. ARM soil moisture anomalies are calculated using 1998 to 2010 to calculate mean soil water content at each depth: 5, 15, 25, 35, 60, 85, 125, and 175 cm. ICN soil moisture anomalies are calculated using 2004 to 2011 to calculate mean soil water content at each depth: 5, 10, 50, 100, and 150 cm. Positive anomalies (wetter-than-normal conditions) are shown in blue and negative anomalies (drier-than-normal conditions) are shown in red.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Figures 8c and 8d show the monthly average soil moisture anomalies from 2004 to 2012 from the ICN stations. ICN record shows a prolonged period of anomalously dry soils between March 2005 and March 2008 interspersed with brief periods of normal soil wetness. Similar to ARM, visualizing soil moisture conditions in this way illustrates that moisture conditions in the soil profile are rarely homogeneous and that the response of the deeper soil layers significantly lags the surface.
Utility of soil moisture for drought monitoring.
Hundreds of drought indices and monitoring and forecasting systems exist, each producing slightly different interpretations of drought and its impacts. Soil moisture is an excellent proxy for agricultural drought as anomalously dry/wet soils influence the amount and rate of crop root uptake and surface runoff. Several previous studies have used soil moisture for agricultural drought monitoring (Quiring and Papakryiakou 2003; Narasimhan and Srinivasan 2005; Bolten et al. 2010). We use soil moisture data provided to the NASMD by the Oklahoma Mesonet to examine how well drought can be characterized with soil moisture. The Oklahoma Mesonet operates over 120 stations across the state of Oklahoma, measuring several meteorological variables at subhourly resolution (Illston et al. 2004).
We directly compare Oklahoma drought conditions to those reported over the same period by the U.S. Drought Monitor (http://droughtmonitor.unl.edu/). The U.S. Drought Monitor is a drought monitoring framework produced jointly by the National Oceanic and Atmospheric Administration, the U.S. Department of Agriculture, and National Drought Mitigation Center, and the University of Nebraska–Lincoln. The Drought Monitor product is a blend of several drought indicators including precipitation and temperature anomalies, satellite-derived vegetation productivity, and modeled soil moisture. The Drought Monitor represents drought intensity as one of five classes, from least to most intense: abnormally dry (D0), moderate drought (D1), severe drought (D2), extreme drought (D3), and exceptional drought (D4). Drought is reported weekly as a percent of a geographic region (e.g., state) in each drought category. To compare soil moisture from the Oklahoma Mesonet to Drought Monitor reports of Oklahoma drought, we used the measurements of daily volumetric water content at 5 cm and then converted these measurements to percentiles based on 2000 to 2013 data for 100 stations in Oklahoma. Percentiles were averaged for each 7-day period between 2000 and 2013 to match the resolution of the Drought Monitor, and drought classes were determined based on the weekly percentiles. Finally, the percent of the state of Oklahoma in each drought category for every week was calculated as the number of Oklahoma Mesonet stations in each category divided by the total number of stations (n = 100). This allowed us to directly compare drought onset, intensity, and duration represented by soil moisture anomalies with that reported by the U.S. Drought Monitor. Although the Oklahoma Mesonet measures soil moisture at 5, 25, 60, and 75 cm, the soil moisture percentiles that are used to generate Figs. 9 and 10 are based on the 5-cm measurements. In previous studies we have evaluated using measurements from other depths (or the entire soil column) and found that using measurements from deeper in the soil column does not improve the results (e.g., Ford and Quiring 2013, 2014a). Although these findings seem somewhat counterintuitive, they have been evaluated multiple times and we believe they are robust. Of course, only using the 5-cm measurements means that there is significantly more variability (noise) in the data.
Area plots of the percent of Oklahoma in drought between 2000 and 2013. Drought categories are color coded by severity. (top) The U.S. Drought Monitor and (bottom) based on soil moisture measured at 100 Oklahoma Mesonet stations.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Area plots of the percent of Oklahoma in drought between 2000 and 2013. Drought categories are color coded by severity. (top) The U.S. Drought Monitor and (bottom) based on soil moisture measured at 100 Oklahoma Mesonet stations.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Area plots of the percent of Oklahoma in drought between 2000 and 2013. Drought categories are color coded by severity. (top) The U.S. Drought Monitor and (bottom) based on soil moisture measured at 100 Oklahoma Mesonet stations.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Area plots showing the percent of Oklahoma in drought between (a) 2000 and 2002, (b) 2005 and 2007, and (c) 2010 and 2012. Areas are color coded by drought severity. The top plots in each panel show the U.S. Drought Monitor and the bottom plots are from soil moisture measured at Oklahoma Mesonet stations.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Area plots showing the percent of Oklahoma in drought between (a) 2000 and 2002, (b) 2005 and 2007, and (c) 2010 and 2012. Areas are color coded by drought severity. The top plots in each panel show the U.S. Drought Monitor and the bottom plots are from soil moisture measured at Oklahoma Mesonet stations.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Area plots showing the percent of Oklahoma in drought between (a) 2000 and 2002, (b) 2005 and 2007, and (c) 2010 and 2012. Areas are color coded by drought severity. The top plots in each panel show the U.S. Drought Monitor and the bottom plots are from soil moisture measured at Oklahoma Mesonet stations.
Citation: Bulletin of the American Meteorological Society 97, 8; 10.1175/BAMS-D-13-00263.1
Figure 9 shows the evolution of drought, as represented by the percent of Oklahoma in each drought category, events between 2000 and 2013. The top plot shows the U.S. Drought Monitor and the bottom plot shows drought from Oklahoma Mesonet soil moisture anomalies. Oklahoma has experienced a number of severe drought events over the last 12 years, including particularly extensive events from 2002 to 2003, 2005 to 2007, and 2010 to 2012. These events are well represented in both datasets; however, the U.S. Drought Monitor shows much more spatially extensive and longer duration drought events than the soil moisture. This could be attributed to the large influence precipitation events have on near-surface soil moisture. This is one drawback to using 5-cm soil moisture measurements. During the very extreme drought events (2006, 2011), a single precipitation event does not provide enough relief to alleviate the drought indices that are used to develop the U.S. Drought Monitor. However, one precipitation event is enough to increase near-surface soil moisture enough that it reduces the severity of drought conditions (i.e., soil moisture percentile) for a short period of time.
Three major drought events in Oklahoma between 2000 and 2013 are examined in greater detail using both records (Fig. 10). The drought events occur between (Fig. 10a) 2000 and 2002, (Fig. 10b) 2005 and 2007, and (Fig. 10c) 2010 and 2012. Events from 2000 to 2002 and 2005 to 2007 are captured by both records. Drought onset is months earlier in the soil moisture record than the U.S. Drought Monitor. However, drought demise roughly occurs at the same time in both datasets. The spatial extent of the 2000–02 drought across Oklahoma as represented by the Drought Monitor is consistently larger than the soil moisture record, although the peak drought severity (drought class) is higher in the soil moisture record (Fig. 10a). Similarly, the 2005–07 drought onset is earlier in the soil moisture dataset, but it is not as spatially extensive as represented by the Drought Monitor (Fig. 10b).
The large drought event occurring between mid-2010 and early 2012 is considered one of the most severe drought events to influence the southern Great Plains in the last century (Hoerling et al. 2013). The onset and demise of the event (Fig. 10c) are similar for both datasets; however, the spatial extent and overall severity is much larger in the Drought Monitor representation than the soil moisture dataset. Throughout 2011, the percent of Oklahoma in each drought event peaks and troughs in the soil moisture dataset, attributable to short time periods of rainfall. Soil moisture can be used to represent drought conditions; however, the duration and spatial extent of soil moisture drought may not be reflective of general agricultural or hydrologic conditions, particularly during severe, long-lasting drought. That being said, the near-surface soil moisture measurements from the Oklahoma Mesonet responded to drier-than-normal conditions immediately preceding drought events much more quickly than the U.S. Drought Monitor. This suggests that a spatially extensive network of in situ soil moisture observations can be used to provide timely drought early warning in the United States.
DISCUSSION.
Web-based application.
The NASMD website is available at http://soilmoisture.tamu.edu/. Our online system allows the user to query the entire database to find data that meet their research needs. Users may query the database not only to find soil moisture in a specific geographic region or time period, but queries based on metadata are also possible. For example, a user may select only soil moisture observations that were taken at the 5-cm depth in soils with less than 20% sand content or select all soil moisture observations from networks using heat dissipation sensors. The website will also allow the user to download all of the metadata along with the soil moisture observations, including sensor change/recalibration dates. Data and metadata from all sites can be downloaded directly from the NASMD online platform. All orders are queued and data are delivered via a zipped folder to the user’s desired e-mail address.
Scientific benefits of the NASMD.
The NASMD provides a wealth of in situ soil moisture observations for research use. One benefit of the NASMD is the utility of soil moisture observations for validating land surface models and satellite-derived soil moisture estimates. Numerous studies employ land surface models (LSM) to investigate land–atmosphere interactions as well as interannual and interdecadal climate variability. However, significant model-to-model variations in LSM-estimated soil moisture combined with a lack of observational soil moisture data make it difficult to adequately assess the accuracy of LSM-estimated soil moisture (Robock et al. 2003; Schaake et al. 2004). Oleson et al. (2004) found that the NCAR Community Land Model (CLM) version 3.5 has a wet soil moisture bias and systematically underestimates soil moisture variability. Other studies have found that previous versions of CLM and similar LSMs also have issues and biases related to vegetation and soil parameterizations (Oleson et al. 2008; Jiang et al. 2009). The NASMD and other soil moisture data sources provide accurate, in situ observations by which these models can be calibrated and validated. For example, Ford et al. (2014b) used data from the NASMD to compare the model-derived and observed soil moisture–energy flux relationships. They found that the relationships between soil moisture and evaporation/latent heat are highly variable in time and space. This is only one example of how observed soil moisture can be used to evaluate and improve our understanding of model-simulated land–atmosphere interactions. Of course it should be noted that many land–atmosphere studies, including Ford et al. (2014b), also utilize other measurements of energy and water fluxes and so locations with flux tower are particularly valuable. There are a total of 96 stations in the NASMD that are collocated with flux towers [AmeriFlux (58 stations), ARM (17 stations), and Fluxnet Canada (21 stations)].
Observed soil moisture also has utility for forecasting the occurrence of extreme heat. Ford and Quiring (2014) examined the statistical relationship between monthly extreme temperatures and observed soil moisture in Oklahoma using quantile regression. They found that soil moisture most strongly impacts extreme heat events at the high end of the conditional distribution, suggesting that soil moisture anomalies have the largest impact on the most extreme heat events. They also assessed the potential of soil moisture for predicting the probability of extreme heat events using soil moisture from the previous month and demonstrated that the skill of these forecasts is comparable to the NOAA Climate Prediction Center monthly mean temperature forecasts (2-week lead time).
Several recent studies have employed satellite-estimated soil moisture to investigate land–atmosphere interactions and possible soil moisture trends. Taylor et al. (2011) employ land surface temperature data as a proxy for soil moisture variability to examine spatial patterns and frequency of convective storm initiation in the Sahel. They found that over a third of all storm initiations occurred over areas with the steepest land surface temperature (soil moisture) gradients. Dorigo et al. (2012) used soil moisture estimated in a merged microwave-based dataset to examine global soil moisture trends over the last two decades. They were able to identify various regions of the globe with strong wetting and drying trends. Ford et al. (2014a) showed that mean absolute differences between standardized soil moisture data from the Oklahoma Mesonet and remote sensing retrievals from the Soil Moisture and Ocean Salinity (SMOS) platform varied from 0.14 to 0.43 (cm3 cm–3). Satellite soil moisture products can provide an accurate depiction of large-scale soil moisture variability; however, extensive validation is necessary to estimate the accuracy of the satellite-derived soil moisture products. Observed soil moisture from the NASMD will aid in the validation of satellite-based soil moisture estimates employed by these and similar studies as well as the calibration of satellite platforms estimating soil moisture such as the Soil Moisture and Ocean Salinity (SMOS) and Soil Moisture Active Passive (SMAP) missions.
SUMMARY.
The North American Soil Moisture Database provides in situ soil moisture observations from nearly 2,000 sites across the United States, Canada, and Mexico. Data contributed from observation networks are integrated into the NASMD, quality controlled using the QAQC algorithm, and are provided to users via a centralized web-based portal. Several data products are available including the “raw” volumetric soil water content product from the observation networks, data that are quality assured by the NASMD, data that are gap filled using the DAR infilling process, and standardized gridded soil wetness data.
The NASMD is a retrospective database for soil moisture data and thus it has primarily been designed to support research applications. Typically, soil moisture data are updated every 6 months. Obviously this is not suitable for applications such as drought monitoring. Therefore, a parallel effort is underway to provide real-time soil moisture data. The President’s Climate Action Plan calls for the development of a National Drought Resilience Partnership that will manage drought-related risks by linking information with drought preparedness and longer-term resilience strategies. One component of the National Drought Resilience Partnership is developing a coordinated national soil moisture network. A workshop was organized by the National Integrated Drought Information System (NIDIS) to articulate a plan of action for development of a coordinated National Soil Moisture Network. The aim of the workshop, which was held in Kansas City, Missouri (November 2013), was to take stock of federal and state in situ monitoring networks, satellite remote sensing missions, numerical modeling capabilities, and how to merge these capacities into a network of networks providing a comprehensive suite of national real-time soil moisture products. As a result of this workshop, NIDIS funded the development of pilot soil moisture monitoring system. This project, which is jointly being led by the USGS Center for Integrated Data Analytics and Texas A&M University, will develop a common, robust infrastructure to integrate and serve disparate soil moisture data from distributed systems and support a suite of value-added soil moisture tools and visualizations. Other recent events, such as the launch of NASA SMAP mission on 31 January 2015 and the upcoming release of NLDAS-3 (which will provide modeled soil moisture for the North American domain on a 1/32° grid) highlight that we are entering a new era that will provide unprecedented access to new soil moisture datasets that can be used to improve drought monitoring and forecasting, calibration/validation satellites and land surface models, and documenting how soil moisture influences the climate system on seasonal to interannual time scales.
ACKNOWLEDGMENTS
We thank the data providers from all of the networks. Without their efforts and support, the NASMD would not be possible. This work was funded by the National Science Foundation (Award AGS-1056796).
REFERENCES
Al Bitar, A., D. Leroux, Y. H. Kerr, O. Merlin, P. Richaume, A. Sahoo, and E. F. Wood, 2012: Evaluation of SMOS soil moisture products over continental U.S. using the SCAN/SNOTEL network. IEEE Trans. Geosci. Remote Sens., 50, 1572–1586, doi:10.1109/TGRS.2012.2186581.
Anderson, W. B., B. F. Zaitchik, C. R. Hain, M. C. Anderson, M. T. Yilmaz, J. Mecikalski, and L. Schultz, 2012: Towards an integrated soil moisture drought monitor for East Africa. Hydrol. Earth Syst. Sci., 16, 2893–2913, doi:10.5194/hess-16-2893-2012.
Bolten, J. D., W. T. Crow, Z. Xiwu, T. J. Jackson, and C. A. Reynolds, 2010: Evaluating the utility of remotely sensed soil moisture retrievals for operational agricultural drought monitoring. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 3, 57–66, doi:10.1109/JSTARS.2009.2037163.
Cai, W., T. Cowan, P. Briggs, and M. Raupach, 2009: Rising temperature depletes soil moisture and exacerbates severe drought conditions across southeast Australia. Geophys. Res. Lett., 36, L21709, doi:10.1029/2009GL040334.
Collow, T. W., A. Robock, J. B. Basara, and B. G. Illston, 2012: Evaluation of SMOS retrievals of soil moisture over the central United States with currently available in situ observations. J. Geophys. Res., 117, D09113, doi:10.1029/2011JD017095.
Dirmeyer, P. A., F. J. Zeng, A. Ducharne, J. C. Morrill, and R. D. Koster, 2000: The sensitivity of surface fluxes to soil water content in three land surface schemes. J. Hydrometeor., 1, 121–134, doi:10.1175/1525-7541(2000)001<0121:TSOSFT>2.0.CO;2.
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, doi:10.5194/hess-15-1675-2011.
Dorigo, W. A., R. de Jeu, D. Chung, R. Parinussa, Y. Liu, W. Wagner, and D. Fernández-Prieto, 2012: Evaluating global trends (1988–2010) in harmonized multi-satellite surface soil moisture. Geophys. Res. Lett., 39, L18405, doi:10.1029/2012GL052988.
Dorigo, W. A., and Coauthors, 2013: Global automated quality control of in situ soil moisture data from the International Soil Moisture Network. Vadose Zone J., 12, doi:10.2136/vzj2012.0097.
Durre, I., M. J. Menne, B. E. Gleason, T. G. Houston, and R. S. Vose, 2010: Comprehensive automated quality assurance of daily surface observations. J. Appl. Meteor. Climatol., 49, 1615–1633, doi:10.1175/2010JAMC2375.1.
Fan, Y., H. M. van den Dool, and W. Wu, 2011: Verification and intercomparison of multimodel simulated land surface hydrological datasets over the United States. J. Hydrometeor., 12, 531–555, doi:10.1175/2011JHM1317.1.
Fiebrich, C. A., C. R. Morgan, A. G. McCombs, P. K. Hall Jr., and R. A. McPherson, 2010: Quality assurance procedures for mesoscale meteorological data. J. Atmos. Oceanic Technol., 27, 1565–1582, doi:10.1175/2010JTECHA1433.1.
Findell, K. L., and E. A. B. Eltahir, 1997: An analysis of the soil moisture-rainfall feedback, based on direct observations from Illinois. Water Resour. Res., 33, 725–735, doi:10.1029/96WR03756.
Findell, K., P. Gentine, B. R. Lintner, and C. Kerr, 2011: Probability of afternoon precipitation in eastern United States and Mexico enhanced by high evaporation. Nature, 4, 434–439, doi:10.1038/ngeo1174.
Ford, T. W., and S. M. Quiring, 2013: Influence of MODIS-derived dynamic vegetation on VIC-simulated soil moisture in Oklahoma. J. Hydrometeor., 14, 1910–1921, doi:10.1175/JHM-D-13-037.1.
Ford, T. W., and S. M. Quiring, 2014a: In-situ soil moisture coupled with extreme temperatures: A study based on the Oklahoma Mesonet. Geophys. Res. Lett., 41, 4727–4734, doi:10.1002/2014GL060949.
Ford, T. W., and S. M. Quiring, 2014b: Comparison and application of multiple methods for the interpolation of soil moisture observations. Int. J. Climatol., 34, 2604–2621, doi:10.1002/joc.3862.
Ford, T. W., E. K. Harris, and S. M. Quiring, 2014a: Estimating root zone soil moisture using near-surface observations from SMOS. Hydrol. Earth Syst. Sci., 18, 139–154, doi:10.5194/hess-18-139-2014.
Ford, T. W., C. O. Wulff, and S. M. Quiring, 2014b: Assessment of observed and model-derived soil moisture-evaporative fraction relationships over the United States Southern Great Plains. J. Geophys. Res. Atmos., 119, 6279–6291, doi:10.1002/2014JD021490.
Ford, T. W., A. D. Rapp, and S. M. Quiring, 2015a: Does afternoon precipitation occur preferentially over dry or wet soils in Oklahoma? J. Hydrometeor., 16, 874–888, doi:10.1175/JHM-D-14-0005.1.
Ford, T. W., A. D. Rapp, S. M. Quiring, and J. Blake, 2015b: Soil moisture–precipitation coupling: Observations from the Oklahoma Mesonet and underlying physical mechanisms. Hydrol. Earth Syst. Sci., 19, 3617–3631, doi:10.5194/hess-19-3617-2015.
Gentine, P., A. A. M. Holtslag, F. D’Andrea, and M. Ek, 2013: Surface and atmospheric controls on the onset of moist convection over land. J. Hydrometeor., 14, 1443–1462, doi:10.1175/JHM-D-12-0137.1.
Georgakakos, K. P., and D. Bae, 1994: Climatic variability of soil water in the American Midwest: Part 2. Spatio-temporal analysis. J. Hydrol., 162, 379–390, doi:10.1016/0022-1694(94)90237-2.
Guo, Z., and P. A. Dirmeyer, 2013: Interannual variability of land–atmosphere coupling strength. J. Hydrometeor., 14, 1636–1646, doi:10.1175/JHM-D-12-0171.1.
Harrison, K. W., S. V. Kumar, C. D. Peters-Lidard, and J. A. Santanello, 2012: Quantifying the change in soil moisture modeling uncertainty from remote sensing observations using Bayesian inference techniques. Water Resour. Res., 48, W11514, doi:10.1029/2012WR012337.
Hirschi, M., and Coauthors, 2011: Observational evidence for soil-moisture impact on hot extremes in southeastern Europe. Nature, 4, 17–21, doi:10.1038/ngeo1032.
Hoerling, M., and Coauthors, 2013: Anatomy of an extreme event. J. Climate, 26, 2811–2832, doi:10.1175/JCLI-D-12-00270.1.
Hollinger, S. E., and S. A. Isard, 1994: A soil moisture climatology of Illinois. J. Climate, 7, 822–833, doi:10.1175/1520-0442(1994)007<0822:ASMCOI>2.0.CO;2.
Hubbard, K., S. Goddard, W. Sorensen, N. Wells, and T. Osugi, 2005: Performance of quality assurance procedures for an applied climate information system. J. Atmos. Oceanic Technol., 22, 105–112, doi:10.1175/JTECH-1657.1.
Illston, B. G., J. B. Basara, and K. C. Crawford, 2004: Seasonal to interannual variations in soil moisture measured in Oklahoma. Int. J. Climatol., 24, 1883–1896, doi:10.1002/joc.1077.
Illston, B. G., J. B. Basara, C. A. Fiebrich, K. C. Crawford, E. Hunt, D. K. Fisher, R. Elliott, and K. Humes, 2008: Mesoscale monitoring of soil moisture across a statewide network. J. Atmos. Oceanic Technol., 25, 167–182, doi:10.1175/2007JTECHA993.1.
Ingleby, B., and M. Huddleston, 2007: Quality control of ocean temperature and salinity profiles—Historical and real-time data. J. Mar. Syst., 65, 158–175, doi:10.1016/j.jmarsys.2005.11.019.
Jackson, T. J., and Coauthors, 2012: Validation of Soil Moisture and Ocean Salinity (SMOS) soil moisture over watershed networks in the U.S. IEEE Trans. Geosci. Remote Sens., 50, 1530–1543, doi:10.1109/TGRS.2011.2168533.
Jiang, X., G.-Y. Niu, and Z.-L. Yang, 2009: Impacts of vegetation and groundwater dynamics on warm season precipitation over the Central United States. J. Geophys. Res., 114, D06109, doi:10.1029/2008JD010756.
Journée, M., and C. Bertrand, 2011: Quality control of solar radiation data within the RMIB solar measurements network. Sol. Energy, 85, 72–86, doi:10.1016/j.solener.2010.10.021.
Khong, A., J. K. Wang, S. M. Quiring, and T. W. Ford, 2015: Soil moisture variability in Iowa. Int. J. Climatol., 35, 2837–2848, doi:10.1002/joc.4176.
Koster, R. D., and Coauthors, 2004: Regions of strong coupling between soil moisture and precipitation. Science, 305, 1138–1140, doi:10.1126/science.1100217.
Koster, R. D., and Coauthors, 2006: GLACE: The Global Land–Atmosphere Coupling Experiment. Part I: Overview. J. Hydrometeor., 7, 590–610, doi:10.1175/JHM510.1.
Kumar, S. V., R. H. Reichle, K. W. Harrison, C. D. Peters-Lidard, S. Yatheendradas, and J. A. Santanello, 2012: A comparison of methods for a priori bias correction in soil moisture data assimilation. Water Resour. Res., 48, W03515, doi:10.1029/2010WR010261.
Legates, D. R., R. Mahmood, D. F. Levia Jr., T. DeLiberty, S. M. Quiring, C. Houser, and F. E. Nelson, 2011: Soil moisture: A central and unifying theme in physical geography. Prog. Phys. Geogr., 35, 65–86, doi:10.1177/0309133310386514.
Leib, B. G., J. D. Jabro, and G. R. Matthews, 2003: Field evaluation and performance comparison of soil moisture sensors. Soil Sci., 168, 396–408, doi:10.1097/01.ss.0000075285.87447.86.
Markewitz, D., S. Devine, E. A. Davidson, P. Brando, and D. C. Nepstad, 2010: Soil moisture depletion under simulated drought in the Amazon: Impacts on deep root uptake. New Phytol., 187, 592–607, doi:10.1111/j.1469-8137.2010.03391.x.
McPherson, R. A., and Coauthors, 2007: Statewide monitoring of the mesoscale environment: A technical update on the Oklahoma Mesonet. J. Atmos. Oceanic Technol., 24, 301–321, doi:10.1175/JTECH1976.1.
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 (CAM3). J. Geophys. Res., 115, D21118, doi:10.1029/2010JD014449.
Merchant, C. J., P. Le Borgne, A. Marsouin, and H. Roquet, 2008: Optimal estimation of sea surface temperature from split-window observations. Remote Sens. Environ., 112, 2469–2484, doi:10.1016/j.rse.2007.11.011.
Mozny, M., M. Trnka, Z. Zalud, P. Hlavinka, J. Nekovar, V. Potop, and M. Virag, 2012: Use of a soil moisture network for drought monitoring in the Czech Republic. Theor. Appl. Climatol., 107, 99–111, doi:10.1007/s00704-011-0460-6.
Myoung, B., and J. W. Nielsen-Gammon, 2010: The convective instability pathway to warm season drought in Texas. Part I: The role of convective inhibition and its modulation by soil moisture. J. Climate, 23, 4461–4473, doi:10.1175/2010JCLI2946.1.
Namias, J., 1991: Spring and summer 1988 drought over the contiguous United States—Causes and prediction. J. Climate, 4, 54–65, doi:10.1175/1520-0442(1991)004<0054:SASDOT>2.0.CO;2.
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, doi:10.1016/j.agrformet.2005.07.012.
Oglesby, R. J., and D. J. Erickson III, 1989: Soil moisture and the persistence of North American drought. J. Climate, 2, 1362–1380, doi:10.1175/1520-0442(1989)002<1362:SMATPO>2.0.CO;2.
Oleson, K. W., and Coauthors, 2004: Technical description of the Community Land Model (CLM). NCAR Tech. Note NCAR/TN-461+STR, 174 pp., doi:10.5065/D6N877R0.
Oleson, K. W., and Coauthors, 2008: Improvements to the Community Land Model and their impact on the hydrological cycle. J. Geophys. Res., 113, G01021, doi:10.1029/2007JG000563.
Pal, J. S., and E. A. B. Eltahir, 2001: Pathways relating soil moisture conditions to future summer rainfall within a model of the land–atmosphere system. J. Climate, 14, 1227–1242, doi:10.1175/1520-0442(2001)014<1227:PRSMCT>2.0.CO;2.
Quiring, S. M., and T. N. Papakryiakou, 2003: An evaluation of agricultural drought indices for the Canadian prairies. Agric. For. Meteor., 118, 49–62, doi:10.1016/S0168-1923(03)00072-8.
Reybold, W. U., and G. W. TeSelle, 1989: Soil geographic data bases. J. Soil Water Conserv., 44, 28–29.
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, doi:10.1175/1520-0477(2000)081<1281:TGSMDB>2.3.CO;2.
Robock, A., and Coauthors, 2003: Evaluation of the North American Land Data Assimilation System over the southern Great Plains during the warm season. J. Geophys. Res., 108, 8846, doi:10.1029/2002JD003245.
Rowlandson, T. L., B. K. Hornbuckle, L. M. Bramer, J. C. Patton, and S. D. Logsdon, 2012: Comparisons of evening and morning SMOS passes over the Midwest United States. IEEE Trans. Geosci. Remote Sens., 50, 1544–1555, doi:10.1109/TGRS.2011.2178158.
Schaake, J. C., and Coauthors, 2004: An intercomparison of soil moisture fields in the North American Land Data Assimilation System (NLDAS). J. Geophys. Res., 109, D01S90, doi:10.1029/2002JD003309.
Schroeder, J. L., W. S. Burgett, K. B. Haynie, I. Sonmez, G. D. Skwira, A. L. Doggett, and J. W. Lipe, 2005: The West Texas Mesonet: A technical overview. J. Atmos. Oceanic Technol., 22, 211–222, doi:10.1175/JTECH-1690.1.
Schubert, S. D., M. J. Suarez, P. J. Pegion, R. D. Koster, and J. T. Bacmeister, 2004: Causes of long-term drought in the U.S. Great Plains. J. Climate, 17, 485–503, doi:10.1175/1520-0442(2004)017<0485:COLDIT>2.0.CO;2.
Shafer, M. A., C. A. Fiebrich, D. S. Arndt, S. E. Fredrickson, and T. W. Hughes, 2000: Quality assurance procedures in the Oklahoma Mesonetwork. J. Atmos. Oceanic Technol., 17, 474–494, doi:10.1175/1520-0426(2000)017<0474:QAPITO>2.0.CO;2.
Tang, C., and T. C. Piechota, 2009: Spatial and temporal soil moisture and drought variability in the Upper Colorado River Basin. J. Hydrol., 379, 122–135, doi:10.1016/j.jhydrol.2009.09.052.
Tang, Q., E. R. Vivoni, F. Muñoz-Arriola, and D. P. Lettenmaier, 2012: Predictability of evapotranspiration patterns using remotely sensed vegetation dynamics during the North American Monsoon. J. Hydrometeor., 13, 103–121, doi:10.1175/JHM-D-11-032.1.
Taylor, C. M., A. Gounou, F. Guichard, P. P. Harris, R. J. Ellis, F. Couvreux, and M. De Kauwe, 2011: Frequency of Sahelian storm initiation enhanced over mesoscale soil-moisture patterns. Nat. Geosci., 4, 430–433, doi:10.1038/ngeo1173.
Taylor, C. M., R. A. de Jeu, F. Guichard, P. P. Harris, and W. A. Dorigo, 2012: Afternoon rain more likely over drier soils. Nature, 489, 423–426, doi:10.1038/nature11377.
Tei, S., A. Sugimoto, H. Yonenobu, T. Yamazaki, and T. C. Maximov, 2013: Reconstruction of soil moisture for the past 100 years in eastern Siberia by using d13C of larch tree rings. J. Geophys. Res. Biogeosci., 118, 1256–1265, doi:10.1002/jgrg.20110.
Teuling, A. J., and Coauthors, 2010: Contrasting response of European forest and grassland energy exchange to heatwaves. Nat. Geosci., 3, 722–727, doi:10.1038/ngeo950.
Wang, A., D. P. Lettenmaier, and J. Sheffield, 2011: Soil moisture drought in China, 1950–2006. J. Climate, 24, 3257–3271, doi:10.1175/2011JCLI3733.1.
Wu, W., and R. E. Dickinson, 2004: Time scales of layered soil moisture memory in the context of land–atmosphere interaction. J. Climate, 17, 2752–2764, doi:10.1175/1520-0442(2004)017<2752:TSOLSM>2.0.CO;2.
Wu, W., M. A. Geller, and R. E. Dickinson, 2002: The response of soil moisture to long-term variability of precipitation. J. Hydrometeor., 3, 604–613, doi:10.1175/1525-7541(2002)003<0604:TROSMT>2.0.CO;2.
Xia, Y., M. B. Ek, Y. Wu, T. W. Ford, and S. M. Quiring, 2015a: Comparison of NLDAS-2 simulated and NASMD observed daily soil moisture. Part I: Comparison and analysis. J. Hydrometeor., 16, 1962–1980, doi:10.1175/JHM-D-14-0096.1.
Xia, Y., M. B. Ek, T. W. Ford, and S. M. Quiring, 2015b: Comparison of NLDAS-2 simulated and NASMD observed daily soil moisture. Part II: Impact of soil texture and vegetation type mismatch. J. Hydrometeor., 16, 1981–2000, doi:10.1175/JHM-D-14-0097.1.
Xia, Y., T. W. Ford, Y. Wu, S. M. Quiring, and M. B. Ek, 2015c: Automated quality control of in situ soil moisture from the North American Soil Moisture Database using NLDAS-2 products. J. Appl. Meteor. Climatol., 54, 1267–1282, doi:10.1175/JAMC-D-14-0275.1.
You, J., K. G. Hubbard, R. Mahmood, V. Sridhar, and D. Todey, 2010: Quality control of soil water data in Applied Climate Information System—Case study in Nebraska. J. Hydrol. Eng., 15, 200–209, doi:10.1061/(ASCE)HE.1943-5584.0000174.
Yuan, X., E. F. Wood, N. W. Chaney, J. Sheffield, J. Kam, M. Liang, and K. Guan, 2013: Probabilistic seasonal forecasting of African drought by dynamical models. J. Hydrometeor., 14, 1706–1720, doi:10.1175/JHM-D-13-054.1.
Zampieri, M., F. D’Andrea, R. Vautard, P. Ciais, N. de Noblet-Ducoudré, and P. Yiou, 2009: Hot European summers and the role of soil moisture in the propagation of Mediterranean drought. J. Climate, 22, 4747–4758, doi:10.1175/2009JCLI2568.1.
Zhang, Y., and M. Wegehenkel, 2006: Integration of MODIS data into a simple model for the spatial distributed simulation of soil water content and evapotranspiration. Remote Sens. Environ., 104, 393–408, doi:10.1016/j.rse.2006.05.011.