are associated with rapidly developing drought conditions that are often known as “flash droughts” (M. C. Anderson et al., 2013; Otkin et al., 2016). Ford et al. (2015) demonstrated that measurements of soil moisture in situ would drastically enhance the identification of flash droughts. Therefore, identification and quantification of drought at different timescales with high‐resolution satellite imagery is crucial for decision making and developing drought mitigation strategies (D’Odorico et al., 2010). Several drought indices have been proposed to address deficiency in soil moisture, including the Crop Moisture Index (CMI; Palmer, 1965), Keetch–Byram Drought Index (KBDI; Keetch & Byram, 1968), Soil Moisture Percentile (Sheffield et al., 2004), Soil Moisture Deficit Index (SMDI; Narasimhan & Srinivasan, 2005), Scaled Drought Condition Index (SDCI) that uses multisensor data (Rhee et al., 2010), Microwave Integrated Drought Index (MIDI) that integrates precipitation, soil moisture, and land surface temperature derived from microwave sensors such as TRMM and AMSR‐E (Zhang & Jia, 2013), Soil Moisture Drought Index (SODI; Sohrabi et al., 2015), and Standardized Soil Moisture Index (SSI; Hao & Aghakouchak, 2013; Figure 1.3).
Figure 1.3 Near real‐time drought monitoring and prediction system by the Global Integrated Drought Monitoring and Prediction System (GIDMaPS) using the Standardized Soil Moisture Index (SSI) for February 2016 based on the Modern‐Era Retrospective analysis for Research and Applications (MERRA) data set. D0 indicates abnormally dry; D1 moderate drought; D2 severe drought; D3 extreme drought; D4 exceptional drought; and the same applies to wetness (W) scale.
What is required for agricultural drought and land surface models is the water content of the plant root zone in soil. This requires observatories in situ that are able to measure soil‐water content at deeper layers of soil and provide more accurate estimations of soil moisture for purposes of drought monitoring as well as validation of satellite estimations of soil moisture. The cosmic‐ray soil moisture observing system (COSMOS; Zreda et al., 2012) and the German terrestrial environmental observatories (TERENO; Zacharias et al., 2011) are two examples of such in situ measurement networks. Moreover, the International Soil Moisture Network (ISMN) (http://www.ipf.tuwien.ac.at/insitu) provides a long record of global in situ soil moisture data, however, these measurements are typically available at point scales and contain significant spatial and temporal gaps. While point‐based measurements are time consuming and costly, passive and active microwave sensor data retrieved from satellites readily provide spatiotemporally consistent observations of soil moisture from the top 5 cm of soil (Entekhabi et al., 2010; L. Wang & Qu, 2009). Given that agricultural drought monitoring requires information about soil moisture content of the entire soil column (i.e., surface and root zone), remotely sensed soil moisture data alone are not adequate for drought monitoring and complementary information about root zone soil moisture needs to be provided using modeling and data assimilation (e.g., Mladenova et al. 2019). Surface soil moisture data are derived mainly from passive or active microwave satellites (De Jeu et al., 2008; Njoku et al., 2003; Takada et al., 2009; Wagner et al., 1999). Currently, the Soil Moisture Active Passive (SMAP; Figure 1.4; Entekhabi et al., 2010) and the Soil Moisture Ocean Salinity (SMOS; Kerr et al., 2010) missions are the main sources of the remote‐sensing‐based soil moisture estimates. These data sets have been used extensively for drought monitoring (e.g., Mishra et al., 2017; Sadri et al., 2018; Sánchez et al., 2016). Soil moisture also can be inferred from other microwave sensors (Entekhabi et al., 2010; Martínez‐Fernández et al., 2016; Moradkhani, 2008; Scaini et al., 2015) such as: the Scanning Multichannel Microwave Radiometer (SMMR), the SSM/I, the European Remote Sensing (ERS) scatterometer, the TRMM microwave imager, the Advanced Scatterometer (ASCAT), and Advanced Microwave Scanning Radiometer2 (AMSR2). Long‐term soil moisture data appropriate for monitoring drought can be obtained through certain databases such as the Water Cycle Multimission Observation Strategy (WACMOS), which is derived from multiple satellites (Ambaw, 2013). Similarly, the European Space Agency's Climate Change Initiative (ESA CCI) offers a soil‐moisture data set with a record of over 30 years that is particularly suitable for monitoring agricultural drought. The ESA CCI merges soil moisture retrievals of a number of different satellites and provides three types of product: active microwave, passive microwave, and combined active–passive microwave (Gruber et al., 2019). The ESA CCI soil‐moisture data set, however, has large gaps over densely vegetated areas. Martínez‐Fernández et al. (2016) show the reliability of the CCI soil‐moisture data set for purposes of modeling agricultural drought.
Figure 1.4 Soil moisture observation by NASA’s Soil Moisture Active Passive (SMAP) satellite. (a) Soil moisture observation of the United States. (b) Global view.
(Courtesy of NASA’s earth observatory: https://earthobservatory.nasa.gov/images).
Monitoring agricultural drought requires high‐resolution data to reveal detailed variations of soil moisture. To improve the spatial resolution of soil moisture data, several downscaling methods have been used, such as machine learning frameworks (Im et al., 2016; Park et al., 2017), DISaggregation based on Physical And Theoretical scale CHange (DISPATCH) which uses shortwave and thermal data from Moderate‐Resolution Imaging Spectroradiometer (MODIS) to downscale SMOS data (Merlin et al., 2015), and Smoothing Filter‐based Intensity Modulation (SFIM) which integrates microwave data from SMAP, Sentinel‐1, and AMSR2 to downscale soil moisture data to an enhanced resolution of 0.1° × 0.1° (Santi et al. 2018).
1.2.3. Relative Humidity
Water vapor has a significant influence on Earth’s climate and energy distribution as it displaces nearly half the trapped heat in an upward and poleward direction and is considered a natural greenhouse gas (Sherwood et al., 2010). Advances in remote sensing have made it possible to monitor water vapor and relative humidity through satellite sensors. Relative humidity is defined as the amount of water available in air with respect to the required water vapor for saturation at a specific temperature. Remotely sensed relative humidity data can be used as an early detection variable to monitor drought (Farahmand et al., 2015). A recently proposed Standardized Relative Humidity Index (SRHI) offers potential information about early drought detection and can be used in conjunction with other indices such as SPI or the Palmer Drought Severity Index (PDSI; Palmer, 1965) for drought monitoring and early warning systems (Farahmand et al., 2015; Figure 1.5). Studies also show that a combination of near‐surface air temperature, vapor pressure deficit, and relative humidity can enhance the detection of drought onset (Behrangi et al., 2016). To detect onset of drought, Farahmand et al. (2015) used the Atmospheric Infrared Sounder (AIRS20) satellite’s relative humidity data and developed a SRHI. The AIRS mission provides relative humidity data with a spatial resolution of 1° and covers a period ranging from 2002 to present. The authors suggested that due to the limited period of recorded data from AIRS for the purposes of drought analysis (< 30 years), the Gravity Recovery and Climate Experiment (GRACE) observations, Evaporative Stress Index data, and a combination of AIRS and reanalysis data sets could be used to extend the observation records.
