**1. Overview and Aim**

Agriculture has been crucial in human life and, over the years, to meet the emerging challenges, production activities and rural landscapes have been gradually moulded and adapted to society demands [1]. For instance, diversifying crop production and optimizing its yield should satisfy the increasing food need due to the continuous global population growing and its changed eating habits [2]. Similarly, minimizing the environmental cost of new agricultural productive activities plays a key role in improving human well-being and preserving biodiversity and ecosystem health [3]. Indeed, the use of inorganic pesticides and chemical fertilizers over recent decades has been recognized as one of the main factors responsible for pollutants diffusion in agricultural soils [4,5], water [6,7], and air [8]. Moreover, the common agricultural policy (CAP) of the European Union, supplying grants for farming activities, requires a continued control of farmers' declarations, to ensure the legitimacy of grants. Satellite missions, such as Sentinel-1 and Sentinel-2, are going to support this control action, thus heavily entering into the new agriculture trend [9,10]. Moreover, insurance companies operating in the agricultural field are looking at the new satellite missions as potential tools for supporting their activities [11,12].

The above-mentioned issues are only a few examples of the problems caused by the agricultural sector which are currently being faced. Thus, a big effort is still required to detect best practices to operate through. Nevertheless, it is evident that a large amount of data concerning both temporal and spatial variations in crop conditions, rural landscapes and, more generally, in the overall agricultural systems, should be collected and integrated. Remote sensing relevance in supporting agricultural applications has been recognized since the 1970s [13]. Nevertheless, because of the reduced spatial and spectral resolutions of the first satellite sensors and their unsuitable revisiting times, remote sensing-based applications were limited for a long time. The new satellite missions, as the Sentinels of the European Space Agency (ESA), and the introduction of new tools, as Remotely Piloted Aircraft Systems (RPAS), has turned on the footlights on such techniques again, offering new opportunities to explore [14–17].

This Special Issue (SI) on "Remote Sensing in Agriculture: State-of-the-Art" moves within this framework and was aimed at gathering contributions useful to delineate the ongoing trends of remote sensing technology transfer to the agricultural sector. Reviewing of conventional methods, proposing of novel data-collecting tools and handling techniques were expected to populate this volume, included focuses about eventual limitations and challenges. Finally, 10 high-quality contributions were selected to be included in this SI, making it possible to significantly account for most of the above-mentioned expectations. Some highlights of the presented contributions are given in Section 2.

#### **2. Highlights of Research Articles**

The SI collected documents satisfyingly deal with the most of issues that the ongoing technology transfer process of remote sensing to agriculture is proposing. Contributions [12,18–20] present some applications based on Synthetic Aperture Radar (SAR) data. Specifically, Ajadi et al. [18] introduce a new method, based on Hidden Markov Random Field (HMRF), to identify crop lodging and to map its extension. Iowa and Illinois were the pilot sites selected to test the new approach. Research results will impact on future use of SAR-based information for operational crop lodging assessment. De Petris et al. [9] propose to map apple orchards damaged by a stormy event by adopting H-α-A polarimetric decomposition technique. Thus, a probability map of potentially storm-damaged orchards was produced. This result may support local funding restoration policies. Hoskera et al., [19] used Sentinel 1 data to estimate soil moisture by adopting both localized and generalized linear models. Particularly, the authors derive 39 localized linear models and 9 generalized linear models. Such models were validated using in situ data and all of them showed promising results. Lastly, Sun et al. [20] propose a novel approach to merge time series Sentinel-1 (S1) and Sentinel-2 (S2) data to map different crop kinds over oasis agricultural areas. A statistically homogeneous pixel (SHP) distributed scatterer interferometry (DSI) algorithm was applied to handle Sentinel 1 data while the random forest technique has been applied to exploit optical properties. The resultant map of five major crop types were generated by integrating the outcomes produced by both methods.

A single contribution proposes a review about the adoption of thermal data acquired by RPAS in supporting precision agriculture. After reporting their main applications and exploring their potentialities, this it offers a potential outlook of development [21].

Contributions [22–24], instead, deal with RPAS-based applications. Each paper refers to an acquisition experience operated with a different sensor, namely LiDAR, hyperspectral and multispectral sensors. LiDAR data were used to estimate fresh biomass and crop height for three different crops (potato, sugar beet, and winter wheat) [21]. Fresh biomass and crop height were assessed using 3DPI algorithm and the mean height of a variable number of points selected for each m2, respectively. The approach showed promising outcomes, albeit the authors outlined that results are strongly dependent on flight conditions. Ref. [23] presents a work where a hyperspectral imaging sensor was mounted on a ground-based vehicle and a RPAS to explore their potentialities in detecting and quantifying yellow rust in wheat at ground canopy and plot scale, respectively. It is the first time that such an experiment was conducted. The authors pinpointed limitations and challenges of such an approach. In [24], the authors adopted two different RPAS equipped with two multispectral mini sensors to analyse vegetated areas. Additionally, the authors explored the opportunity to integrate such sensors to detect vegetation changes.

Lastly, the studies proposed by [25,26] involve the application of satellite data to the agricultural sector. The former [25] derives nine vegetation indices from Moderate Resolution Imaging Spectroradiometer (MODIS) data to predict crop yield in Mongolia. The authors concluded that the Normalized Difference Water Index (NDWI) and Visible and Shortwave Infrared Drought Index (VSDI) are the optimal indicators to meet research purposes. Additionally, the end of June and the beginning of July have been recognised as the best timings to forecast the production yield. Conversely, [26] assess and compare the performance of MODIS, Landsat, and blended images in evaluating crop yield over a period of about 6 years (2009–2015). This contribution is aimed at detecting the best data to accurately monitor biophysical processes and yields by using freely available data.

**Conflicts of Interest:** The authors declare no conflict of interest.
