Search

Yield mapping technologies can help to increase the quantity and quality of agricultural production. Current systems only focus on the quantification of the harvest, but the quality has equal or greater importance in some perennial crops and impacts directly on the financial profitability. Therefore, a system was developed to quantify and relate the quality obtained in the classification line with the plants of the orchard and for decision-making. The system is comprised of hardware, which obtains the location of the harvester bag during harvesting and unloading at the unloading site, and software that processes the collected data. The cloud of real-time data contributed from the different collectors (bins) allows the construction of yield maps, considering the multi-stage harvesting system. Further, the system enables the creation of a detailed map of the plants and fruits harvested. As the harvest focuses on quality, it takes place in stages, depending on the ripening of the fruits. In addition to the yield maps, the system allows identification of the efficiency of each worker undertaking the harvest by the number of performed discharges and by the time spent. The system was developed in partnership with the Federal Technological University of Paraná and Embrapa Uva & Vinho and was tested in apple orchards in southern Brazil. Although the system was evaluated with only data from apple cultivation, monitoring the quality and quantifying other orchard fruits can positively impact the fruit sector. [ABSTRACT FROM AUTHOR]

There is increasing interest in how big data will affect U.S. crop production, yet little is known about the field‐level effects of "small" (i.e., individual farm) data. We help to fill this void by studying the relationship between Midwest corn production and the information contained in yield and soil maps. Research on this relationship is lacking, perhaps because maps are information inputs that may not enter the production function in a way comparable to conventional inputs. Using detailed USDA survey data, we implement a stochastic frontier analysis to evaluate how mapping technologies influence field productivity. Controlling for farmers' endogenous choice of technologies, we find evidence of direct (frontier‐shifting) and indirect (efficiency‐enhancing) productivity effects. Depending on model, field output increases by 5.6% or 11.9% as a result of map adoption. Yield maps increase expected efficiency by 8.5%, and soil maps increase expected efficiency by 7.2%, on average. These effects differ by operator demographics, such as years of experience with the field, and structural characteristics, such as whether the field is insured and if it is owned by the operator. Given that yield and soil maps are not universally adopted, our results suggest there remain opportunities to increase productivity through field‐level information use. [ABSTRACT FROM AUTHOR]

Understanding subfield crop yields and temporal stability is critical to better manage crops. Several algorithms have proposed to study within-field temporal variability but they were mostly limited to few fields. In this study, a large dataset composed of 5520 yield maps from 768 fields provided by farmers was used to investigate the influence of subfield yield distribution skewness on temporal variability. The data are used to test two intuitive algorithms for mapping stability: one based on standard deviation and the second based on pixel ranking and percentiles. The analysis of yield monitor data indicates that yield distribution is asymmetric, and it tends to be negatively skewed (p < 0.05) for all of the four crops analyzed, meaning that low yielding areas are lower in frequency but cover a larger range of low values. The mean yield difference between the pixels classified as high-and-stable and the pixels classified as low-and-stable was 1.04 Mg ha−1 for maize, 0.39 Mg ha−1 for cotton, 0.34 Mg ha−1 for soybean, and 0.59 Mg ha−1 for wheat. The yield of the unstable zones was similar to the pixels classified as low-and-stable by the standard deviation algorithm, whereas the two-way outlier algorithm did not exhibit this bias. Furthermore, the increase in the number years of yield maps available induced a modest but significant increase in the certainty of stability classifications, and the proportion of unstable pixels increased with the precipitation heterogeneity between the years comprising the yield maps. [ABSTRACT FROM AUTHOR]

Abstract: The objective of this work was to evaluate the spatial and temporal variability of the dry matter yield of irrigated corn for silage, as well as its economic return. The study was conducted in an irrigated silage corn field of 18.9 ha in the municipality of São Carlos, in the state of São Paulo, Brazil. The spatial variability of the yield of three crop seasons, normalized yield indexes, production cost, profit, and soil electrical conductivity (EC) were modeled using semivariograms. Yield maps were obtained by kriging, and management zones were mapped based on average yield, normalized index, and EC. The results showed a structured spatial variability of corn yield, production cost, profit, and soil EC within the irrigated area. The adopted precision agriculture tools were useful to indicate zones of higher yield and economic return. The sequences of yield maps and the analysis of spatial and temporal variability allow the definition of management zones, and soil EC is positively related to corn yield.

• Biomass modelling with high spatial resolution. • Assimilation of remote sensing data into a mechanistic crop growth model. • Biomass forecast using historical meteorological data. • Development of a web-application to visualize biomass in space and time. • User-friendly interface with the possibility to analyse and download the data. Spatial information on crop status and development is required by agricultural managers for a site specific and adapted management. Here, a prototype of a web application is presented for the visualization of biomass production of maize (Zea mays). The web application displays past biomass development and future predictions for user-defined regions of interest along with summary statistics. Biomass is modelled using the crop growth model (CGM) APSIM (Agricultural Production Systems Simulator) using meteorological data from 2001 to 2014. Information on current crop status and subfield heterogeneity is assimilated into APSIM through high-resolution optical satellite imagery. The use of recent satellite data and regional, historical meteorological data increases the reliability of the biomass information provided. Through its unique combination of high-resolution satellite imagery together with mechanistic crop growth modelling, this web application can overcome the often sparse temporal or sparse spatial resolution of biomass information, which is based on remote sensing images or on crop growth modelling alone. The prototype presented, with its high resolution biomass maps, can be the basis for variable rate application as farmers can react site-specifically to plant development. [ABSTRACT FROM AUTHOR]

This research is an outcome of the R&D activities of Ecodevelopment S.A. (steadily supported by the Hellenic Agricultural Organization—Demeter) towards offering precision farming services to rice growers. Within this framework, a new methodology for topdressing nitrogen prediction was developed based on machine learning. Nitrogen is a key element in rice culture and its rational management can increase productivity, reduce costs, and prevent environmental impacts. A multi-source, multi-temporal, and multi-scale dataset was collected, including optical and radar imagery, soil data, and yield maps by monitoring a 110 ha pilot rice farm in Thessaloniki Plain, Greece, for four consecutive years. RapidEye imagery underwent image segmentation to delineate management zones (ancillary, visual interpretation of unmanned aerial system scenes was employed, too); Sentinel-1 (SAR) imagery was modelled with Computer Vision to detect inundated fields and (through this) indicate the exact growth stage of the crop; and Sentinel-2 image data were used to map leaf nitrogen concentration (LNC) exactly before topdressing applications. Several machine learning algorithms were configured to predict yield for various nitrogen levels, with the XGBoost model resulting in the highest accuracy. Finally, yield curves were used to select the nitrogen dose maximizing yield, which was thus recommended to the grower. Inundation mapping proved to be critical in the prediction process. Currently, Ecodevelopment S.A. is expanding the application of the new method in different study areas, with a view to further empower its generality and operationality.

Increasing the efficiency of agricultural operations is a necessity in the current environmental context. Because of this, the Azcuba Business Group has undertaken the task of introducing high technology machinery such as the Case IH A8000 sugarcane harvesters into its Sugar Companies. These are among the leading machineries of their type, providing high efficiency and availability for the application of Advanced Farming Systems. All of this is supported by a novel technological package, which has not been used in its entirety by the Cuban Sugar Company to date. The present work shows the potential of the software for the management of these machines Spatial Management System, from the generation of reports and yields maps. The generation of the maps is based on the information recorded by three Case IH A8000 sugarcane harvesters in a case study at the Villa Clara's Sugarcane Factory Hector Rodriguez. [ABSTRACT FROM AUTHOR]

Estimating maize ( Zea mays L.) yields at the field level is of great interest to farmers, service dealers, and policy-makers. The main objectives of this study were to: i) provide guidelines on data selection for building yield forecasting models using Sentinel-2 imagery; ii) compare different statistical techniques and vegetation indices (VIs) during model building; and iii) perform spatial and temporal validation to see if empirical models could be applied to other regions or when models' coefficients should be updated. Data analysis was divided into four steps: i) data acquisition and preparation; ii) selection of training data; iii) building of forecasting models; and iv) spatial and temporal validation. Analysis was performed using yield data collected from 19 maize fields located in Brazil (2016 and 2017) and in the United States (2016), and normalised vegetation indices (NDVI, green NDVI and red edge NDVI) derived from Sentinel-2. Main outcomes from this study were: i) data selection impacted yield forecast model and fields with narrow yield variability and/or with skewed data distribution should be avoided; ii) models considering spatial correlation of residuals outperformed Ordinary least squares (OLS) regression; iii) red edge NDVI was most frequently retained into the model compared with the other VIs; and iv) model prediction power was more sensitive to yield data frequency distribution than to the geographical distance or years. Thus, this study provided guidelines to build more accurate maize yield forecasting models, but also established limitations for up-scaling, from farm-level to county, district, and state-scales. Highlights • We forecast maize yield using Sentinel 2 imagery. • Data selection impacted yield forecast models. • Red-edge NDVI presented high performance to forecast within field yield variability. • Model performance was more affected by data distribution than field location or year. [ABSTRACT FROM AUTHOR]

Copyright of Revista Brasileira de Engenharia Agricola e Ambiental - Agriambi is the property of Revista Brasileira de Engenharia Agricola e Ambiental and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The spatial and temporal variability of crop parameters are fundamental in precision agriculture. Remote sensing of crop canopy can provide important indications on the growth variability and help understand the complex factors influencing crop yield. Plant biomass is considered an important parameter for crop management and yield estimation, especially for grassland and cover crops. A recent approach introduced to model crop biomass consists in the use of RGB (red, green, blue) stereo images acquired from unmanned aerial vehicles (UAV) coupled with photogrammetric softwares to predict biomass through plant height (PHT) information. In this study, we generated prediction models for fresh (FBM) and dry biomass (DBM) of black oat crop based on multi-temporal UAV RGB imaging. Flight missions were carried during the growing season to obtain crop surface models (CSMs), with an additional flight before sowing to generate a digital terrain model (DTM). During each mission, 30 plots with a size of 0.25 m² were distributed across the field to carry ground measurements of PHT and biomass. Furthermore, estimation models were established based on PHT derived from CSMs and field measurements, which were later used to build prediction maps of FBM and DBM. The study demonstrates that UAV RGB imaging can precisely estimate canopy height (R2 = 0.68−0.92, RMSE = 0.019−0.037 m) during the growing period. FBM and DBM models using PHT derived from UAV imaging yielded R2 values between 0.69 and 0.94 when analyzing each mission individually, with best results during the flowering stage (R2 = 0.92−0.94). Robust models using datasets from different growth stages were built and tested using cross-validation, resulting in R2 values of 0.52 for FBM and 0.84 for DBM. Prediction maps of FBM and DBM yield were obtained using calibrated models applied to CSMs, resulting in a feasible way to illustrate the spatial and temporal variability of biomass. Altogether the results of the study demonstrate that UAV RGB imaging can be a useful tool to predict and explore the spatial and temporal variability of black oat biomass, with potential use in precision farming.

Many farmers sample their soil to measure the concentrations of plant nutrients, so as to decide how much fertilizer to apply. Now that fertilizer can be applied at variable rates farmers want to know whether maps of nutrient concentration made from grid samples or of field subdivisions (zones within their fields) are merited: do such maps lead to greater profit than would a single measurement on a bulked sample for each field when all costs are taken into account? We have examined the merits of grid-based and zone-based sampling strategies over single field-based averages using continuous spatial data on wheat yields at harvest in six fields in southern England and simulated concentrations of phosphorus (P) in the soil. We have taken into account current prices of wheat, P fertilizer and sampling and laboratory analysis. Variograms of yield provide guides for sampling. We show that where variograms have large variances and long effective ranges grid-sampling and mapping are feasible and have large probabilities of being cost-effective. Where effective ranges are short, sampling must be dense to reveal the spatial variation and be expensive, and variable-rate application of fertilizer is likely to be impracticable and almost certainly not cost-effective. We found zone-based sampling was less likely to be cost effective in a similar situation when the management zones were poorly correlated to P concentrations. [ABSTRACT FROM AUTHOR]

[Display omitted] • Precision agriculture and ecological niche model (ENM) tackle maize yield gap Yg. • Maxent-based NEM assesses spatial–temporal patterns of yield potential and Yg. • ENM-Maxent identifies and quantifies impact of topographic-edaphic variables on Yg. • Enhance agronomic strategies and decisions for sustainable Yg filling. • Integrated agronomical interpretation for improved precision agriculture. Precision agriculture (PA) and yield gap (Y g) analysis are promising strategies to achieve the desired sustainable intensification of agricultural production systems. Current crop Y g approaches do not consider the internal field yield variability caused by soil properties. Topographic and edaphic characteristics causing consistent high and low yield patterns in time and space can be interpreted as an ecological niche and used as proxies for potential yield (Y p) and Y g. Ecological niche models (ENMs) are statistical models originally developed to forecast a species' niche. However, its application to analyse crop yield spatio-temporal variability has never been made. This study aimed to fill this void by developing a novel approach: i) to quantify the magnitude and spatio-temporal distribution of Y p and Y g , ii) to identify the main factors that cause the Y g , and iii) to provide statistical and agronomical interpretation of the data to reduce the Y g. We performed this work using high-resolution maize yield maps from three seasons, with an ancillary dataset composed of soil electrical conductivity, soil properties and digital elevation models provided by "Quinta da Cholda", Portugal. The yield maps were averaged, resulting in a standardised multiyear yield map. The 90th and 10th yield percentiles were interpreted as proxies for Y p and Y g , and analysed by an ENM machine learning algorithm – maximum entropy (MaxEnt). The average Y g and Y p were quantified as 1.5 and 19.1 ton/ha. Y p was characterised by having silty, richer soils and lower elevations, with several nutritional factors above the critical limits to maintain higher yields. Y g had loam soils coupled with higher relative elevations and lower nutrition content. This innovative modelling approach can efficiently manage high-dimensional spatio-temporal data to support advanced PA solutions, allowing detailed support for narrowing the Y g. [ABSTRACT FROM AUTHOR]

Spatially accurate and reliable estimates from fast-growing plantations are a key factor for planning energy supply. This study aimed to estimate the yield of biomass from short rotation willow plantations in northern Europe. The data were based on harvesting records from 1790 commercial plantations in Sweden, grouped into three ad hoc categories: low, middle and high performance. The predictors included climatic variables, allowing the spatial extrapolation to nearby countries. The modeling and spatialization of the estimates used boosted regression trees, a method based on machine learning. The average RMSE for the final models selected was 0.33, 0.39 and 1.91 (corresponding to R2 = 0.77, 0.88 and 0.45), for the low, medium and high performance categories, respectively. The models were then applied to obtain 1×1 km yield estimates in the rest of Sweden, as well as for Norway, Denmark, Finland, Estonia, Latvia, Lithuania and the Baltic coast of Germany and Poland. The results demonstrated a large regional variation. For the first rotation under high performance conditions, the country averages were as follows: >7 odt ha−1 yr−1 in the Baltic coast of Germany, >6 odt ha−1 yr−1 in Denmark, >5 odt ha−1 yr−1 in the Baltic coast of Poland and between 4-5 odt ha−1 yr−1 in the rest. The results of this approach indicate that they can provide faster and more accurate predictions than previous modeling approaches and can offer interesting possibilities in the field of yield modeling. [ABSTRACT FROM AUTHOR]

Agronomic input and management practices have traditionally been applied uniformly on agricultural fields despite the presence of spatial variability of soil properties and landscape position. When spatial variability is ignored, uniform agronomic management can be both economically and environmentally inefficient. The objectives of this study were to: i) identify optimal N fertilizer rates using an integrated spatio-temporal analysis of yield and site-specific N rate response; ii) test the sensitivity of site specific N management to nitrate leaching in response to different N rates; and iii) demonstrate the environmental benefits of variable rate N fertilizer in a Nitrate Vulnerable Zone. This study was carried out on a 13.6 ha field near the Venice Lagoon, northeast Italy over four years (2005–2008). We utilized a validated crop simulation model to evaluate crop response to different N rates at specific zones in the field based on localized soil and landscape properties under rainfed conditions. The simulated rates were: 50 kg N ha − 1 applied at sowing for the entire study area and increasing fractions, ranging from 150 to 350 kg N ha − 1 applied at V6 stage. Based on the analysis of yield maps from previous harvests and soil electrical resistivity data, three management zones were defined. Two N rates were applied in each of these zones, one suggested by our simulation analysis and the other with uniform N fertilization as normally applied by the producer. N leaching was lower and net revenue was higher in the zones where variable rates of N were applied when compared to uniform N fertilization. This demonstrates the efficacy of using crop models to determine variable rates of N fertilization within a field and the application of variable rate N fertilizer to achieve higher profit and reduce nitrate leaching. [ABSTRACT FROM AUTHOR]

On-farm experiments are used increasingly in agronomic research because they are commercially relevant, but they can carry greater risks of failure than traditional small plot experiments conducted by scientists. Experimental failures can result from farmer withdrawal, errors in treatment application or harvest, or non-provision of yield data by farmers. This paper describes the development and testing of approaches for on-farm experimentation and concludes which approaches should be adopted to maximise success. The programme of work included the largest on-farm research network in the UK, with farmers conducting around 50 on-farm experiments per year from 2017 to 2019 to compare fungicide programmes in winter wheat. The project developed management approaches to mitigate the risks of experimental failure such that in 2019, 96 % of experiments were completed and returned a yield result; a greater success rate than is commonly achieved in on-farm experiments. Statistical analysis of yield maps resulted in an average site SED (standard error of the difference between means) of 0.26 t/ha, which is comparable to that achieved in randomised, replicated small plot experiments. The large number of experimental sites enabled a greater level of precision in the cross-site analysis (SED 0.06 t/ha), showing the potential of on-farm experiments for detection of small yield effects if the appropriate yield analysis is undertaken. The project results received substantial publicity within the UK arable farming community, demonstrating the value of effective on-farm research for engaging and informing farmers. • On-farm experiments risk failure from treatment or harvesting errors, or non-provision of yield data by host farmers. • Risks can be mitigated by good farmer liaison, marking test tramlines, and providing instructions for yield data transfer. • Large series of on-farm experiments deliver precise comparisons of agronomic inputs in commercial situations. • The results of on-farm experiments are relevant and persuasive for the farming community. [ABSTRACT FROM AUTHOR]

Napiergrass, which resembles sugarcane in stature and cultivation practices, is emerging as a candidate bioenergy crop. However, limited studies investigating harvesting and yield sensing of napiergrass are available. This study investigated stem-bending force, and the hydraulic pressures of basecutter, chopper and elevator drives in a John Deere 3522 sugarcane billet harvester as indicators of napiergrass yield. The coefficients of determination ( R 2 ) between napiergrass yield and hydraulic pressures were 0.73, 0.88 and 0.81, respectively for the basecutter, chopper and elevator drives. The highest correlation ( R 2 = 0.92) was found between stem-bending force and napiergrass yield. The yield prediction errors were 4.9% and 8.6% for the calibration and validation plots with the stem-bending force yield sensor. Cross-validation, in which each harvested row was treated as a data point, showed that the average yield prediction errors were 10.9% and 11.8% for the calibration and validation data sets. Yield maps were also generated employing the stem-bending yield sensor. In addition, it was expected that the stem-bending yield sensor could be utilized to control harvester operation such as travel speed. Further studies would be needed to extend the stem-bending concept to other thick stemmed crops. [ABSTRACT FROM AUTHOR]

Site-specific management of N (NSSM) is an attractive and intuitive approach to increasing N fertilizer use efficiency (NUE) of agricultural systems by adjusting fertilizer rates to the soil characteristics. The objective of this study is to assess: whether delineating of management zone (MZ) within fields improves NUE in wheat ( Triticum aestivum L.). This research was carried out at 5 commercial fields (between 26 and 84 ha), located in the south-eastern portion of the Province of Buenos Aires, Argentina. The MZ were delineated by using georeferenced measurements of apparent soil electrical conductivity, terrain elevation and soil depth. Spatially referenced wheat yields were recorded with a yield monitor equipped with DGPS. The interaction effect was significant ( p < 0.05) in most fields, thus indicating that the response to N fertilization is different among MZ. Also, NUE was significantly different ( p < 0.05) among MZ. The detection of soil spatial variability and the delineation of MZ are now possible on a commercial scale. The delineation of MZ affords the opportunity of variable rate application of N fertilizers on Typic Argiudolls and Petrocalcic Paleudolls, and the minimization of pollution risk due to an excessive application of resources. [ABSTRACT FROM AUTHOR]

Precision agriculture is a methodology to identify and exploit variability within an agricultural field. In India, a yield is measured during marketing of harvested crop and as a gross yield of the land owned by the farmer. The yield monitor mounted over grain combine harvester measures and records information such as grain flow, grain moisture, area covered and location. An automated yield monitoring system consisting of a yield sensor, global positioning system (GPS), field computer with custom software was mounted on a self propelled indigenous grain combine harvester for real-time crop yield mapping along with moisture data. By means of optical sensors (light emitter and detector), the height of the grain on the elevator paddles was measured and converted in the grain mass flow. Three rice fields were harvested to evaluate the performance of the yield monitor for grain yield and moisture mapping of harvested grains. The actual yield maps were generated by using Arc GIS software from the data collected for three different rice fields. The total area harvested by using indigenous combine harvester fitted with yield monitor and moisture sensor was 0.84 ha comprising three fields (1, 2 and 3) having areas 0.3, 0.22 and 0.32 ha respectively. On an average, yields for all the three fields (1, 2 and 3) were 4,325.91, 5,093.14 and 4,287.66 kg ha-1 with moisture content of 21.42%, 22.78% and 20.42% (wet basis) recorded respectively. On an average, yield variations recorded for all the three fields (1, 2 and 3) were having 30.59%, 40.80% and 40.39% coefficient of variation respectively. The minimum and maximum yields recorded within all the fields harvested were 577.08 and 7,661.48 kg ha-1 with an average yield of 4,287.66 kg ha-1 having 37.26% coefficient of variation (CV) in all the three fields. [ABSTRACT FROM AUTHOR]

This paper proposes an integrated framework for software that provides yield data cleaning and yield opportunity index ( Y) calculation for site-specific crop management (SSCM). The artifacts in many yield data sets, which inevitably occur, can pose a significant effect on the validity of Yi. Automated and standardised yield correction procedures were designed to improve the data quality by removing: (1) unreasonable outliers; (2) distribution outliers (globally and locally); and (3) position errors. The calculation of Yi uses two aspects of crop yield assessment, the magnitude of yield variation and the spatial structure of the variation. The cleaning algorithms were applied to four yield data sets with known integrity issues to demonstrate effectiveness. Approximately 13-20 % of the original yield data were removed, and this resulted in an increased mean yield of 0.13 t/ha (average). The semivariograms of cleaned data were shown to possess smaller nugget values compared with the original data. The opportunity index calculation algorithm was demonstrated on a field with nine seasons of yield data. The results demonstrated that using a ranking of Yi provides a rational, agronomic assessment of the opportunity for SSCM based on the quantity and pattern of production variability displayed in yield data sets. This provides farm managers with a rapid way to assess whether the observed variability deserves further investigation and eventual investment in SSCM operations. [ABSTRACT FROM AUTHOR]

Abstract: Precision agriculture was applied in handpicked fruit orchards in Greece in the last few years. The results of applications in apples and vines are presented in this paper. Yield, quality and NDVI maps were created for three years. The results for both crops showed that there was high variability in the measured parameters. Comparison of the maps indicated that fruit quality was inferior in areas where yield was high. NDVI maps in apples at the beginning of June were correlated to the final yield variability. This result can be used as an early warning of yield variability and could be used to manage the predicted variability. Similar results were indicated for vines. [Copyright &y& Elsevier]

Abstract: The non-destructive assessment of forage mass in legume–grass mixtures as a tool for yield mapping in precision farming applications has been investigated in two field experiments. An ultrasonic sensor was used to determine sward heights. Forage mass–height relationships were evaluated by carrying out static measurements on binary legume–grass mixtures of white clover (Trifolium repens L.), red clover (Trifolium pratense L.), and lucerne (Medicago sativa L.) with perennial rye grass (Lolium perenne L.) across a wide range of sward heights (5.0–104.2cm) and forage mass (0.15–11.25tha−1). Mobile measurements, hereafter referred to as “on-the-go” were conducted by mounting the ultrasonic sensor in combination with a high-precision Differential Global Positioning System (DGPS) on a vehicle. Data were recorded along experimental plots consisting of perennial rye grass and grass–clover mixtures similar to the mixtures that were used for the static experiment. The static experiment revealed a relationship between ultrasonic sward height and forage mass explaining 74.8% of the variance with a standard error (SE) of 1.05tha−1 in a common dataset. The type of legume species, weed proportion, and growth period had a significant impact on the above mentioned relationship. Legume-specific regression functions had higher R 2-values of up to 0.855 (white clover mixture). Datasets including legume-specific mixtures and pure swards of both components reached comparable R 2 values between 0.799 and 0.818 but exhibited higher SE values. The abundance of weeds resulted in increased ultrasonic sward heights for the same levels of forage mass. On-the-go measurements across experimental field plots yielded a sward height range of 1.4–70.4cm. Abrupt forage mass changes at the transition from treatment plots to cut interspaces resulted in a significant deviation from stubble height within a distance of 50cm to plot borders. When legume-specific equations derived from static measurements were applied to sward heights, forage mass was overestimated by 21.4% on average. Mean residuals from predicted forage mass ranged between 0.893 (pure grass) and 1.672 (red clover mixture) and increased significantly if the point sampling distance along the track was increased to more than 0.82m on average across all plots. The prediction accuracy of forage mass from ultrasonic height measurements is promising; however, further modifications to the technique are necessary. One such improvement can be the use of spectral reflectance signatures in combination with the ultrasonic sensor. [Copyright &y& Elsevier]

In-season prediction of crop yield is a topic of research studied by several scientists using different methods. Seasonal forecasts provide critical insights to different stakeholders who use the information for strategic and tactical decisions. In this study, we propose a novel scalable method to forecast in season subfield crop yield through a machine learning model based on remotely sensed imagery and data from a process-based crop model on a cumulative crop drought index (CDI) designed to capture the impact of in-season crop water deficit on crops. To evaluate the performance of our proposed model, we used 352 growers' fields of different sizes across the states of Michigan, Indiana, Iowa, and Illinois, with 2520 respective yield maps generated by combine harvesters equipped with precise high-resolution yield monitor sensor, over multiple years (from 2006 up to 2019). We obtained high resolution digital elevation model, climate, and soil data to execute the SALUS model, a process-based crop model, to calculate the CDI for each field used in the study. We used Landsat Analysis Ready Dataset (ARD) products generated by USGS as image source to calculate the green chlorophyll vegetation index (GCVI). We found that the inclusion of the CDI in remote sensing-based random forest models substantially improved in-season subfield corn yield prediction. The addition of the CDI in the yield prediction model showed that the greatest improvements in predictions were observed in the driest year (2012) in our case study. The proposed approach also showed that the subfield spatial variations of corn yield are better captured with the inclusion of CDI for most fields. The earliest prediction in the growing season with GCVI and CDI together outperformed the latest prediction with GCVI alone, highlighting the potential of CDI for predicting spatial variability of maize yield around grain filling period, which is on average close to two months before typical crop harvest in the US Midwest. • A novel approach was developed to better predict in-season subfield corn yields. • Simulated plant water deficit was added in the model based on vegetation index. • Subfield predictions were conducted in 352 fields across US corn belt states with different environmental conditions. • In-season subfield maize yield prediction improved when crop water deficit was added remote sensing imagery-based model. [ABSTRACT FROM AUTHOR]

Abstract: Sensing, interpreting and acting upon within-field spatial variation in crop performance through precision agriculture (PA) techniques stands to benefit farmers economically and environmentally. The increases in crop gross margin required to offset the cost of purchasing and operating PA technology can be calculated to help growers make PA investment decisions. Economic modelling shows potential benefits of <$5/ha to $40/ha for variable rate management. This is supported by on-farm trials showing benefits of $29–63/ha for zone management in the northern sandplain of Western Australia. The full benefits of zone management can only be realised by developing methods for defining management zones that are consistent in performance, and accounting for crop nutrient requirements within zones by allowing for seasonal effects on yield potential. Various methods can be used to define zones of consistent performance in fields that can be targeted for variable rate fertiliser inputs. In many situations yield variation can be related to variation in soil plant-available water capacity. Predictive systems using geophysical information will enable inexpensive extrapolation of valuable point-based soil characterisation. Constraints to adoption by farmers include lack of training and technical support, equipment incompatibility, perceived riskiness of economic returns, and barriers to use of “hi tech” elements. Future research, development and extension should target a wider farmer audience who are aware of spatial variability but are not currently using PA technologies. [Copyright &y& Elsevier]

Abstract: The main aim of this research was to develop and evaluate sensor system to create forage yield maps on a three-point linkage type of rotary mower-conditioner. The method was based on the mowing machine conditioner''s power requirement measured by a torque sensor or on material change in momentum measured by a curved impact plate. Laboratory measurements were taken to determine the dependence of conditioner power input and signals from the impact plate on material mass flow. A mixture of grass and alfalfa was used. There was a very good linear relationship between the conditioner''s power, impact force from an impact plate, and material feed rate through the mower. The calculated coefficients of determination (R 2) were about 0.95. It was possible to differentiate a material feed rate difference 0.5kgs−1 using either method. The effects of material changes and mower parameters on the accuracy of feed rate measurement were then measured. It was observed that the results form torque sensor was influenced by crop variety, maturity and intensity of conditioning. The same influence was not observed for impact plate. A field of 0.54ha was then harvested. A comparison of data from the torque sensor and impact plate with data from hand measurement were made by means of statistical and geostatistical analysis. Conditioner power input measurement and crop impact force were used to create grass yield maps. [Copyright &y& Elsevier]

Abstract: Yield maps contain a wealth of information and can be an important tool for making informed decisions on paddock management. However, yield datasets obtained from combine harvesters often have many errors arising from a variety of sources. It is therefore important to attempt to rectify as many of these errors as possible so that the yield map represents the true yield as accurately as possible, rather than some systematic or operator error. This research defines the most significant errors associated with raw yield datasets and presents and applies a methodology for dealing with the unknown crop width, the time lag of grain, inappropriate GPS recordings, yield surges, and other outlying values. In total, 16.6% of the original yield dataset acquired over a 96 ha paddock located in the wheat belt of Western Australia was removed because they exhibited one or more of the aforementioned errors. The amount of uncertainty in the filtered dataset was substantially reduced. The accuracy of three spatial interpolation techniques was assessed over the entire paddock using the root mean squared error (RMSE). Finally, it is discussed that the map used to assist paddock management should be chosen depending on how much smoothing and data aggregation is desirable and allowable, not simply on the RMSE statistic. [Copyright &y& Elsevier]

Data and yield maps were obtained from farmers using their own combine harvester equipped with commercial systems for yield measurements and satellite positioning. Three quarters of the datasets collected were abandoned due to being incomplete or problems occurring when overlaying maps from different years. Intra-field yield variation was investigated in the remaining 82 fields. Some fields had data present for 4 years. The predominant crops were winter wheat and winter barley. The coefficient of variation (cv) ranged from 0.05 to 0.22, which was comparable to the residual cv usually found in agricultural experiments. In spring and winter rape the average cv was 0.15, cereals varied from 0.05 to 0.22. The distribution of yields within fields was generally non-normal; it was negatively skewed (left tailed) and broader than a normal distribution. Although the highest yielding crops and fields had the lowest intra-field variation, the distributions of yield were mostly negatively skewed. Regression analyses of intra-field yield pattern between years gave R2 values from 0 to 0.81, with an average of 0.27. Similar crops and crops with small average yield differences gave the highest R2 values, even when they were separated by more than one year in time. Prediction of intra-field variation, trend maps of temporal stability, and maps of classified management, could therefore be improved by developing them for individual crops based on the most appropriate previous crops. It appears unlikely that trend maps will account for much more than 1/3 of the annual intra-field variation on average. The possibilities for developing maps of classified management based on similar quality yield maps as obtained from farmers in the present investigation, appears limited because of the high frequency of erroneous data sets, systematic errors in the recorded data and their restricted yield predictive ability. [Copyright &y& Elsevier]

The objective of this study was to determine the usefulness of on-farm precipitation measurement, through determining spatial and temporal precipitation variability and its effect on corn yield. CERES-Maize (DSSAT version 3.5) was used with three precipitation data sources, for an Indiana farm—an on-farm National Weather Service (NWS) station, the nearest non-urban NWS station with electronic reporting (27 km from the farm), and a weighted mean of the three nearest such stations (27–35 km away)—to simulate 31 years of crop yield on 1-ha grid cells. Described as a percentage of the mean, spatial precipitation variability among the three data sources by corn phenological phase was 21–104%, while temporal (year-to-year) variability was 20–49%. The difference in simulated yield based on spatial precipitation variability was 15.8%, while year-to-year yield variability was 21.5%. The apparent yield difference based on spatial precipitation variability was of the same order as year-to-year variability, which suggests having on-farm precipitation data may be necessary for accurate yield modeling. [Copyright &y& Elsevier]

Many yield maps exhibit systematic errors that attenuate the underlying yield variation. Two errors are dealt with in detail in this paper: those that occur when the harvester has a narrow finish to a land and those that occur when the harvester is filling up at the start of a harvest run. The authors propose methods to correct or remove erroneous data by the use of an expert filter, or alternatively use of an interpolation technique called potential mapping. [ABSTRACT FROM AUTHOR]

• A ten years real farm case study is reported and studied. • Digital agriculture solutions improved corn yield on average by 31% on a ten years basis. • Applied nitrogen reduced by around 23% while the total yield was improving. • Precision and digital agriculture approaches evolve with changing conditions and scenarios. Farmer's management decisions and environmental factors are the main drivers for field spatial and temporal yield variability. In this study, a 22 ha field cultivated with corn for more than ten years using different prescription maps of nitrogen application rates was investigated. Prescription maps were developed based on archived yield maps, soil analysis and recently integrated with Sentinel 2 satellite images. In addition, farmer experience and availability of variable rate application (VRA) requirements had an influence on the development of the homogeneous management zones. The initial approach with VRA was quite simple, based on a simple partitioning of the field into three rectangular zones (defined mainly based on previous yield maps and farmer experience). The partitioning changed with time and knowledge, evolving to the final five irregularly shaped zones (defined based on Farm works decision support software). Furthermore, since 2010 the farmer began using soil moisture sensor for irrigation decisions. Results of the present study highlight an improvement in corn yield and a reduction in total applied nitrogen. Corn yield improved on average by 31% on a ten years basis to reach more than 14 ton/ha dm. in 2018. At the beginning of VRA, yield maps showed a high spatial variation between field zones compared to reduced variation in the following seasons. In addition, the nitrogen applied reduced by around 23% while the total yield was improving. These results showed an increase in the partial factor productivity from less than 54 to around 87 kg of corn grain per kg of nitrogen applied. This promising result shows that farmer management decisions can improve every season by continuous monitoring of crop performance, understanding field variability and taking advantage of recently developed decision support software tools. [ABSTRACT FROM AUTHOR]

This research is an outcome of the R&D activities of Ecodevelopment S.A. (steadily supported by the Hellenic Agricultural Organization—Demeter) towards offering precision farming services to rice growers. Within this framework, a new methodology for topdressing nitrogen prediction was developed based on machine learning. Nitrogen is a key element in rice culture and its rational management can increase productivity, reduce costs, and prevent environmental impacts. A multi-source, multi-temporal, and multi-scale dataset was collected, including optical and radar imagery, soil data, and yield maps by monitoring a 110 ha pilot rice farm in Thessaloniki Plain, Greece, for four consecutive years. RapidEye imagery underwent image segmentation to delineate management zones (ancillary, visual interpretation of unmanned aerial system scenes was employed, too); Sentinel-1 (SAR) imagery was modelled with Computer Vision to detect inundated fields and (through this) indicate the exact growth stage of the crop; and Sentinel-2 image data were used to map leaf nitrogen concentration (LNC) exactly before topdressing applications. Several machine learning algorithms were configured to predict yield for various nitrogen levels, with the XGBoost model resulting in the highest accuracy. Finally, yield curves were used to select the nitrogen dose maximizing yield, which was thus recommended to the grower. Inundation mapping proved to be critical in the prediction process. Currently, Ecodevelopment S.A. is expanding the application of the new method in different study areas, with a view to further empower its generality and operationality. [ABSTRACT FROM AUTHOR]