**5.2 Remote sensing**

Remote sensing technology is used to collect image data from space- or airborne cameras and sensor platforms. Aerial remote sensing platforms such as sensors

#### **Figure 4.**

*Illustration of GIS data being used in precision Ag.Source: http://www.cavalieragrow.ca/ifarm, cited in Hammonds [78].*

**Figure 5.**

*Components of remote sensing technology used in precision agriculture. Sources of photos, from left to right [80–83].*

on-board satellites and aircraft, including UAVs, have recently seen an increase in use. These technologies can be used to estimate and quantify many soil properties by integrating geo-referenced field data (soil and crop) with spectral properties of soil acquired by sensors. Khanal et al. [79] reported that integration of remotely sensed data and machine learning algorithms offers a cost-and time-effective approach for spatial prediction of soil properties and corn yield compared with traditional methods [79]. Remotely sensed images can overcome such limitations and improve spatial and temporal coverage of data on soil and the yield of crops. Aerial and ground-based drones can be used for soil and field analysis, crop planting, applying pesticides, crop monitoring, irrigation, and health assessment [72]. Recently a startups company developed a UAV-based seeding system that reduces costs by 85% [72]. Sensors gather data on soil water availability, soil fertility, soil compaction, soil temperature, crop growth rate such leaf area index, leaf temperature, pest, and disease infestation. A typical example of remote sensing technology's components is shown in **Figure 5**.

#### **5.3 Site-specific soil and crop management**

According to Robert et al. [40], "site-specific crop management is an information and technology-based agricultural management system to identify, analyze, and manage site-soil spatial and temporal variability within fields for optimum profitability, sustainability, and protection of the environment." The most general approach to address such soil property spatial variability is the creation and management of SSMZs or sub-zones [84, 85]. These production level SSMZs or sub-regions are homogenous and have similar characteristics and yield-limiting factors with equal productivity potential [84, 86]. Khosla and Alley [86] optimized a soil sampling grid method using homogenous management zones in a large field. Fleming et al. [87] developed nutrient maps based on production management zones for VRT nutrient application. The delineation of soil property spatial variability can also be accomplished by identifying location spatially coherent areas within the field [88]. The delineation of management zones of a large field can facilitate managing variability

*Precision Agriculture for Sustainable Soil and Crop Management DOI: http://dx.doi.org/10.5772/intechopen.101759*

among the different zones [89]. A recent study suggested that a 50-m soil sampling interval can be considered an optimal interval for delineating production management zones in medium- and small-scale farmlands [90].

Wells et al. [91] reported that precision deep tillage varying depth compared with deep tillage at 400 mm in one site out of three locations enhanced crop yield and farmers' benefits. Adoption of control traffic farming technology [92–94], reducing tire inflation pressure systems [25, 30, 95–99], and site-specific deep tillage [25, 88, 91, 100] are becoming viable methods for reducing soil compaction and enhancing the productivity and sustainability benefits in many cropping systems. These benefits can be triggered and expedited by adopting PA technologies. With the advances in PA and application of remote sensing tools, soil compaction assessment and mapping, modeling, and possible management can also be accomplished [9, 25, 101]. A recent study in the USA showed the adoption rates of grid sampling practice linked with site-specific lime and fertilizer application have been adopted on 40% of cropland (two out of five crop acres) [67]. On the other hand, the adoption of GNSS-assisted yield monitors for site-specific lime and fertilizer application was 43 and 59%, respectively [67]. Studies in the U.K. showed that autonomous equipment for PA is technically and economically feasible; and, if adopted, offers the potential to minimize costs for many farms [12].

#### **5.4 Variable-rate technology (VRT)**

PA is often misinterpreted as a complex technological intervention to agriculture in the developed world [5]. It is, however, shown to be profitable in less-developed regions in the world. For example, micro-dosing of nutrients to nutrient-starved soils in Africa showed an increase in crop grain yields [102]. Several case studies demonstrated that managing in-field variability benefited farmers in China [103]. Geostatistical methods are used to evaluate the spatial distribution of soil properties of a farm [104]. A detailed understanding of the spatial distribution of soil properties facilitates the SSM for maintaining crop and soil productivity while minimizing costs and decreasing the environmental impact [105–107].

Sustainable soil nutrient management, that is, site-specific nutrient management with a complete understanding of in-field soil and spatial variability, performs well in avoiding soil degradation and improving crop productivity [25, 108]. It is well known that soil physical and chemical properties are spatially variable and can be affected by farming practices such as irrigation and fertilization [48, 88, 109], so VRT application is crucial in the management of in-field soil variability. N is the most mobile and dynamic nutrient [86] and plays a vital role in maximizing crop yields and returns to farmers. These soil properties and management practices can affect N dynamics and the mechanisms of its losses from the soil. Remote sensing and GIS tools allow identifying, measuring, and developing maps of these changes across the field landscape. It has shown that VRT N management can potentially improve the N use efficiency by better adjusting N rates to crop needs [17]. A recent study demonstrated that site-specific P and K management could optimize target crop yield and save 21 kg ha−1 and 30 kg ha−1 of P and K, respectively, compared with conventional farming [90]. Therefore, the application of the correct products in the correct place at the correct rate is recognized as one of the key benefits of PA, which is generally accomplished with the use of VRT [67].

#### **5.5 Yield monitoring and mapping**

Today, modern combine harvesters are sold with integrated yield monitors as standard equipment, presenting a powerful tool for grain production. It allows

farmers to assess and delineate the effects of weather, soil properties, and management on grain production [71]. Shearer et al. [71] reported that there are three key benefits of the yield monitors: i) an operator can quickly view crop performance during harvest, ii) yield data can be transferred to a computer and summarized on a field-by-field or total-farm basis, and iii) this information can be geographically referenced to generate yield maps for year-to-year comparisons of high- and lowyielding areas of a field. However, proper installation, calibration, and operation are necessary to ensure the accuracy of these devices. Soil sampling followed by laboratory analyses provides detailed soil health information while yield monitors help in understanding the spatial variability in crop yield [79]. An example of different components of the yield monitoring and mapping system and yield map is shown in **Figure 6a** and **b**.

PA promotes the better use of information to improve the management of infield soil and spatial variability on the farm [111]. The yield maps are central to the management of arable management [111]. Yield mapping and soil sampling tend to be the first stage in implementing precision farming [59]. The yield monitoring

#### **Figure 6.**

*Different components of yield mapping system (a) and yield map (b) [110]. Red color indicates low-yielding areas, and green indicates higher than average crop yield.*

#### **Figure 7.**

*Monitoring of crop in Illinois by an automated agricultural robot, 'TerraSentia' [118]. Source: thesiliconreview. com (07/09/2018).*

system allows the collection of geo-referenced yield data and generation of yield maps for visualizing crop performance variability. Several interpolation techniques such as the inverse of distance, inverse of square distance, and ordinary kriging are commonly used in developing yield maps [112].
