The dynamics of SM at the land surface are governed by the set of components with diverse time and spatial scales. Variances in both weather and climate are therefore influenced by the SM state. Reynolds (1970) distinguished between static (e.g., soil texture, topography) and dynamic (e.
g., precipitation, vegetation) controlling factors. So the evaluation of SM patterns differ depending on related variables condition. Many of these factors are interrelated and vary spatially and/or temporally, making it challenging to recognize explicit cause and effect relationships between SM and its driving variables. Landscape factors, including topography, vegetation, and land use are important controlling factors of SM spatial and temporal variations. SM spatial variation was found to be significantly correlated to terrain attributes (e.g.
slope, elevation, and topographic wetness index). Therefore, terrain attributes have been used to predict SM variation via regression, geospatial, and hydrological modeling in several studies (Western et al. 1999, 2004; Lin et al. 2006; Takagi and Lin 2012). Influence of vegetation (e.
g. type, cover and distribution) on SM variation has also been reported in several studies, and spatial information on vegetation (usually interpreted from remote sensing image) has been used to simulate SM variation (Mohanty et al. 2000; Hupet and Vanclooster 2002).
Generally, long-term time series of SM quantities can reveal trends in the water cycle related to climate or hydrological condition in the area. Due to the fact that, at a large area basis the number of networks and gauging SM, in particular on a continuous basis, is still restricted (because of labor intensive, very slow, and may be very expensive) and furthermore, it is challenging to get reliable approximations at the large scale from point measurements because of the high variability and the low degree of observed autocorrelation, so for various applications the large number of satellite-based SM products show promise in assisting hydrologists to describe and measure the surface SM condition for large areas. Since the 1970’s a number of remote sensing techniques have been created to investigate and mapping SM by measuring different areas of the electromagnetic spectrum from the optical to microwave regions (Musick and Pelletier, 1988; Engman, 1991; Wang and Qu, 2009). As remote sensors do not measure SM content directly, mathematical models that describe the connection between the measured signals and SM content must be derived. Microwave remote sensing methods such as the Advanced Microwave Scanning Radiometer-Earth observing system (AMSR-E) on-board the Aqua satellite (since 2002), Soil moisture and ocean salinity satellite (SMOS since 2009), Multi-frequency Scanning Microwave Radiometer (MSMR since 1999) and Soil Moisture Active Passive (SMAP) (since January 2015) are presently operational, providing satellite data for the globe on a daily basis. Although these methods offer many procedures to achieve SM at large scale, they are almost low resolvable (typically ? 25km) and not appropriate in small catchment or field scale.
Optical/thermal infrared remote sensing data known as Surface Temperature/Vegetation Index Method provide finer resolution information (?1km). Recently Zhang and Zhou (2016) reviewed new technique that SM estimation can be extracted from optical / thermal remote sensing, which is mainly depends on the association between the SM and the surface reflectance and temperature or vegetation index. Such retrieval methods in this field like thermal inertia methods, emphasized on soil thermal characteristics, or triangular relationship method that shows a relationship among SM, Normalized Difference Vegetation Index (NDVI), and land surface temperature (LST) of a given region being used in practical applications, but the lack of considering the other factors such as topography or just being capable for bare or low density vegetation cover is their weakness.
The remotely sensed based vegetation indices to estimate soil moisture (e.g. NDVI, Normalize Difference Water Index (NDWI), and Normalized Multi-band Drought Index (NMDI)) are good option but the distribution of SM cannot be predicted by a single parameter and the determining parameter changes between different land surface factor intensities. Extensive efforts have been carried out over the past decades to estimate SM using satellite images through developing the relationship between remotely sensed LST and vegetation indexes has been reported by many researchers (e.g., Hosseini and Saradjian, 2011; Zhao and Li, 2013).
In addition to topography, vegetation and LST can now be mapped with high spatial resolution (from 30m to 1km) with the use of remotely sensed imagery. Therefor to predict SM conditions using the dependent landscape factors which is extracted from remotely sensed imagery instead of in-situ measurement, makes it possible to achieve the fast and real-time monitoring of SM conditions.