The future projection of population growth is expected to increase the food demand, with implications of a massive expansion of global agriculture areas. Srinivasan et al. ^[1] predicted that the present rate of increasing agricultural yield would not satisfy the projected rising food demand in 2050 and beyond; thus, new techniques and innovations are necessary to overcome this challenge. Furthermore, since food production is highly linked to soil, without proper agricultural management, the environment, and natural resources’ sustainability will be threatened as well.

Nowadays, the total land area is estimated at around 130,575,894 km². However, only around 12% is suitable for crop production without many limitations, while only 3% of the total land area is considered highly productive [2,3]^[2][3]. This is because not all agricultural soils are fertile and productive. Moreover, and regarding human interventions, not all soils are used efficiently. With few exceptions, the times from antiquity to the present day are characterized by the absence of a strategic demarcation of soil use-zones according to soil’s physical and chemical properties and in combination with local climatic conditions. This led to soil over-exploitation without a plan and without considering its future sustainability. Therefore, employing tired and degraded soil to continue providing food for the growing population of the planet will be challenging.

Soil quality protection is part of sustainable soil management and becoming a necessity globally. In the framework of agriculture, the meaning of soil quality is defined as the ability of soil to perform its function of sustainable agricultural production and enable it to respond to sustainable land management ^[4]. Sustainable soil management also includes precision agriculture, which refers to the successful implementation of the identification and understanding of important parameters in order to design the appropriate management plans successfully. One of these parameters is the detailed understanding of soil physicochemical properties that are important for agricultural practice.

The significance of precision agriculture has gained the widespread attention of scientific communities as it has become a critical issue amid the increasing worldwide food demand. Soil properties such as soil water content, organic matter, soil nutrient, soil texture, and soil structure are some typical agronomic parameters regularly monitored by farmers. Traditionally, these properties are assessed by in situ measurement, sampling, and followed by laboratory analysis. These methods are invasive, laborious, time-consuming, and cannot represent a larger area. In this context, spatial and timely observations are crucial to capture the variability of soil properties. The development of soil sensors, ranging from electrical, electromagnetic, optical, and radiometric to mechanical sensors, offers opportunities to improve the effectiveness of soil monitoring ^[5].

Currently, non-invasive techniques such as remote sensing, geophysics, and soil modeling have been successfully employed in agricultural studies by a number of researchers. The aforementioned techniques provide valuable means for the characterization of soil properties. Physicochemical information required by farmers can be retrieved through remote sensing and geophysics techniques, while modeling can increase peourple's understanding of complex soil processes.

Maintaining and improving soil quality is essential to secure food production and can be achieved by the regular monitoring of soil properties. Therefore, accurate measurements of soil properties are considered as a preliminary step for the successful implementation of precision agriculture as their variabilities could affect a variation in crop yield. In addition, understanding their variabilities in space and time will be essential and crucial for the decision-making process in agriculture management. The agronomic soil properties required to be observed regularly include supporting factors such as SM, SOM, soil nutrients, soil texture, soil structure, and degrading factors such as soil compaction and salinity. Conventional soil sampling and laboratory analysis are accurate but time-consuming and labor-intensive. Adequate information on the spatial distribution of soil properties can support the implementation of precision agriculture strategies. However, their acquisition could be restricted by the cost related to laboratory analysis. Remote sensing observation, geophysical surveys, and soil modeling offer alternative solutions for soil property assessment in a non-invasive and time- and cost-effective way.

The use of remote sensing observations for modern agricultural management has progressed tremendously due to its advantages in observing many parameters required for precision farming. Moreover, their spatial and temporal coverages of current and future missions have progressed significantly. The utilization of optical remote sensing in the wavelength domain of visible (VIS), near-infrared (NIR), and short-wave infrared (SWIR) in soil studies has been highlighted by numerous studies. Several physical and chemical soil attributes, including SM, SOM, soil nutrient, soil texture, and soil structure, can be successfully retrieved using these spectral regions.

Despite its advantages, the inability of the optical satellite to penetrate cloud and vegetation cover has been recognized as a major limitation. On the other hand, microwave-based remote sensing can measure soil properties under a variety of topographic and vegetation cover without being constrained by weather conditions. The most popular application of microwave sensing in soil studies is SM monitoring due to its superiority in SM retrieval. However, the ability of microwave sensing is often limited by its temporal and spatial resolutions; thus, it cannot be well performed at a farm scale. Among the developed remote sensing technologies, thermal sensing applications are still limited. Commonly, thermal sensing is integrated with optical sensing to assess SM variabilities through the triangle method. The use of an optical and thermal camera mounted on a UAV might overcome the limitation of SM measurement through microwave sensing. Other soil properties can be also addressed without worrying about the impact of cloud cover due to the UAV’s low altitude.

Proximal sensing (geophysical acquisition) offers the opportunity to bridge the gap between small-scale point-based measurements and large-scale remote sensing observations. Geophysical techniques can help to understand the underlying processes regulating the soil–plant–atmosphere continuum. Their techniques in precision agriculture include ER, EMI, GPR, and seismic, which differ in the type of measured physical properties. Similar to remote sensing, various soil properties have been successfully monitored using geophysics, such as SM, SOM, soil structure, soil texture, soil compaction, and soil salinity. Among them, the application of seismic methods in agricultural studies is still limited to soil compaction monitoring. Geophysical techniques have different sensitivity to soil properties. For instance, GPR and seismic methods respond primarily to the soil interface, while ER and EMI respond to bulk properties [128]^[6]. Therefore, the combination with geophysical measurement might become a future trend in agricultural applications to better understand subsoil characteristics.

Numerical soil modeling generally exploits the Richards equation and the convection–dispersion equation to simulate hydrological fluxes and dissolved substance transport in the vadose zone, respectively. Their implementations would be beneficial in understanding the physicochemical processes, particularly in the complex agricultural system. Various numerical packages have been developed to solve both the Richards and convection–dispersion equations, including HYDRUS, CATHY, FEFLOW, TOUGH, and VS2DI. Among them, HYDRUS is the most widely used numerical software to simulate underlying processes such as infiltration, root water uptake, soil contaminants, and the salinization process. Some limitations appear since Richards’ equation is basically a single-phase flow equation where the contribution of air flow in the soil is considered not significant [173]^[7].

Other challenges in soil modeling could come from the complexity of physical and biochemical processes in the unsaturated zone, the location of the targeted simulation area at the interface between different spheres, and computational difficulties in dealing with highly nonlinear and coupled processes [176]^[8]. Therefore, future work should incorporate various physical, chemical, and biological parameters into models. Since vadose zone processes are part of a large environment, coupling a small-scale with into a large-scale model is necessary for a better understanding of complex natural processes—for instance, coupling between HYDRUS and groundwater models such as MODFLOW. Lastly, developing advanced numerical and visualization code sets with a friendly user interface will be a future challenge that should be addressed.

Overall, remote sensing techniques offer the ability to image the Earth’s surface at a local, regional, or even global scale through different platforms, from airborne to satellites. They have been proven as reliable tools to acquire soil properties on the soil surface and below the soil surface at up to a few centimeters. Meanwhile, near-surface geophysics are usually deployed to characterize detailed subsoil properties due to their abilities to perform deeper acquisition. In particular, information derived from geophysical surveys can be used to validate remote sensing observations and soil modeling. Soil modeling is commonly utilized to predict the complex natural processes occurring in the vadose zone. Input for soil modeling can be from remote sensing and near-surface geophysics. The comparison of these three different techniques is shown in Table 4Table 1.