early detection of drought onset."/>
Figure 1.5 Standardized Relative Humidity Index (SRHI) for (a) August 2010, (b) probability of drought detection, and (c) missed drought ratio, which indicates that relative humidity can be used in conjunction with other drought indices for early detection of drought onset
(Farahmand et al., 2015).
Measurements of relative humidity via remote sensing are often undertaken with IR‐based observing platforms (e.g., the AIRS20) (Fetzer et al., 2006; B. Tian et al., 2004). However, clouds tend to bias the IR observations, which is a major limiting factor since no observation of wet conditions will be available after a strict cloud screening (John et al., 2011). Another major issue is the variation of relative humidity due to changes in saturated vapor pressure, as it is significantly influenced by air temperature. Therefore, even with a fixed water vapor content, changes in air temperature will result in variations in relative humidity (Moradi et al., 2016). On the other hand, microwave sounder retrievals can produce large errors owing to modeling errors of Earth’s limb radiances (e.g., Microwave Limb Sounder) (Lambert et al., 2007). In general, too much uncertainty arises from observations of water vapor in diurnal and spatial distribution of the troposphere (Boyle & Klein, 2010), and having a course resolution of 2–3 km in both IR and microwave sounders, these instruments are unable to portray a detailed vertical structure of water vapor.
Vergados et al. (2015) used the Global Positioning System Radio Occultation (GPSRO) observations from the Constellation Observing System for Meteorology, Ionosphere, and Climate (COSMIC) mission that resolved the challenges associated with the presence of cloud. The authors demonstrated that the GPSRO‐derived relative humidity data possess high quality. Sondeur Atmosphérique du Profil d’Humidité Intertropicale par Radiométrie (SAPHIR) from the Megha Tropiques satellite provides relative humidity data with temporal resolution of several observations per day and has six channels specifically for the water vapor absorption line at 183 GHz with a spatial resolution of 10 km at nadir for all the channels. Using the measurements of SAPHIR, Moradi et al. (2016) found larger diurnal amplitude over land compared to the ocean; larger oceanic amplitude over convective regions compared to subsidence regions; and showed that in tropical regions, relative humidity of the troposphere showed large inhomogeneity in diurnal variation. Brogniez et al. (2016) further improved the relative humidity estimates from the SAPHIR Sounder by producing uncertainty estimates of the relative humidity through a Bayesian framework. Studies also suggested the appropriateness of using algorithms based on data from satellites such as the Advanced Very High Resolution Radiometer (AVHRR), National Oceanic and Atmospheric Administration (NOAA) Geostationary Operational Environmental Satellites (GOES), and MODIS to derive estimates of the surface level relative humidity (Han et al., 2005; Ramírez‐Beltrán et al., 2019).
The frequency of unusually dry and hot conditions has increased in various parts of the world (Griffin & Anchukaitis, 2014; Seager & Hoerling, 2014). Some studies reported that the ever‐increasing anthropogenic radiative forcing is responsible for the recent changes in Earth’s hydrological cycle (Chikamoto et al., 2017; Littell et al., 2016; Williams et al., 2015). Chikamoto et al. (2017) demonstrated that droughts enhance wildfire probabilities in forested systems that take a huge toll on the economy, environment, and local communities in the countryside. Wildfire smoke tremendously increases the level of air pollution and therefore proliferates mortality, and respiratory and cardiovascular morbidity. Accurate measurement of relative humidity is essential for retrieving Aerosol Optical Thickness (AOT) and quantifying particulate matter (PM). Aerosol optical thickness can be derived from the MODIS on board NASA’s Terra and Aqua satellites. The humid air surrounding hygroscopic aerosols causes swelling and this will substantially increase the scattering efficiency of the particles (Hess et al., 1998; Twohy et al., 2009). Gupta et al. (2006) found that a relative humidity ranging from 50% to 80% would increase AOT less than 5%, whereas a relative humidity range of 98–99% results in a more pronounced increase (more than 25%). These results indicate that relative humidity data can be used to enhance the measurements of PM and devise mitigation strategies (Bowman & Johnston, 2005) to reduce the adverse impacts of the hazard (i.e., drought‐associated events such as wildfires).
1.2.4. Evapotranspiration
Evapotranspiration (ET) is an important variable in agriculture, accurate estimation of which is essential for modeling agricultural drought. Evapotranspiration directly affects socioeconomic systems and agriculture, as irrigation water demand and crop yield are determined by this variable. Ecosystem and agriculture responses to drought are depicted by the ratio between actual ET (AET) and potential ET (PET) (Thornthwaite, 1948). Accordingly, several drought indices have been proposed that incorporate ET into their calculation including the PDSI, Crop Water Stress Index (CWSI; Jackson et al., 1981), Supply–Demand Drought Index (SDDI; Rind et al., 1990), Water Deficit Index (WDI; Moran et al., 1994), Reconnaissance Drought Index (RDI; Tsakiris & Vangelis, 2005), Evaporative Drought Index (EDI; Yao et al., 2010), Standardized Precipitation Evapotranspiration Index (SPEI; Vicente‐Serrano et al., 2010), Evaporative Stress Index (ESI; M. C. Anderson et al., 2016), Drought Severity Index (DSI; Mu et al., 2013), Green Water Scarcity Index (GWSI; Núñez et al., 2013), Green Water Stress Index (GrWSI; Wada, 2013), Standardized Palmer Drought Index (SPDI; Ma et al., 2014), Multivariate Drought Index (MDI; Rajsekhar et al., 2015), effective Reconnaissance Drought Index (eRDI; Tigkas et al., 2017), Normalized Ecosystem Drought Index (NEDI; Chang et al., 2018), and Aggregate Drought Index (ADI; S. Wang et al., 2018).
Both RDI and SPEI are widely used water‐balance‐system agricultural drought indices that utilize precipitation and PET as their input (Tsakiris et al., 2007; Vicente‐Serrano et al., 2010). While SPEI uses the Penman–Monteith method to derive PET (Figure 1.6), RDI utilizes temperature‐based methods to estimate PET and can use satellite‐retrieved air temperature data (Dalezios et al., 2012). Recently, Tigkas et al. (2017) modified the RDI index by substituting precipitation by effective precipitation (the amount of water that contributes to crop development and is absorbed by the root system), which can more effectively describe plant water consumption. The modified index (eRDI) has the advantage of considering different stages of crop development and has shown higher correlation to reduction of crop yield in the location studied (Tigkas et al., 2017). Despite the advantages of utilizing the temperature‐based method of PET, it suffers from several shortcomings as other factors such as net radiation, wind speed, and relative humidity that have strong influence on PET are being neglected in the process (Donohue et al., 2010; McVicar et al., 2012). Ma et al. (2014) outlined some issues regarding the climatic water balance system used by SPEI and suggested that SPEI would be more realistic if soil‐moisture‐related hydrometeorological processes are considered. They redefined the procedure of PDSI calculation on the basis of the mathematical framework of SPEI and proposed a new multiscalar drought index. While indices such as SPI or PDSI can be used as early warning systems to detect potential drought imposed on an ecosystem, NEDI offers an actual drought stress response to limited water availability.
The remotely sensed methods of ET estimation can be categorized into four groups, including water balance systems (R. G. Allen et al., 1998; Senay, 2008), surface energy balance systems (R. G. Allen et al., 2007; Anderson & Kustas, 2008), vegetation indices (Glenn et al., 2011), and hybrid approaches that incorporate vegetation indices and surface temperature measurements (Kalma et al., 2008; Yang & Shang, 2013). MODIS data have been frequently used worldwide to obtain land surface temperature data and derive ET for purposes of drought monitoring such as estimation of Evaporative Stress Index (ESI; Figure 1.7; M. C. Anderson et al., 2007) and DSI. Some other satellites capable of measuring land surface temperature include Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) on board Terra, Landsat, AVHRR on board polar orbiting platforms of NOAA, and visible and infrared imager (MVIRI) on board Meteosat satellites. The ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS)
