Precision Agriculture

Abstract Unmanned aerial vehicles (UAVs) are a recent innovation in precision agriculture technology. They are being used in a wide range of agricultural practices, whereby pesticide application is one of the most common uses of UAVs in China’s agriculture. However, the economic effects of UAVs in pesticide application have not been sufficiently investigated. To address the gap, this paper used propensity score matching to evaluate the economic effects of UAV adoption on outcome variables including revenue, pesticide costs, time spent on pesticide application, and pesticide application frequency based on a dataset covering over 2000 grain farmers across 11 provinces of China. Furthermore, generalized propensity score matching was used to evaluate the heterogeneity of outcome variables arising from differing UAV adoption intensities. The empirical results show that adoption of UAV increased revenue by approximately 434–488 dollars per hectare and reduced the time spent on pesticide application in the range of 14.4–15.8 h per hectare. Depending on the area with use of UAVs for pesticide spraying, UAV adoption has heterogeneous impacts on grain farmers’ revenue and the time spent on pesticide application. In terms of marginal revenue and marginal time spent on pesticide application, the optimal area with use of UAVs for pesticide spraying is estimated to be 20 hectares of arable land.

Abstract Hyperspectral imaging allows for rapid, non-destructive and objective assessments of crop health. Narrowband-hyperspectral data was used to select wavelength regions that can be exploited to identify wheat infected with soil-borne mosaic virus. First, leaf samples were scanned in the lab to investigate spectral differences between healthy and diseased leaves, including non-symptomatic and symptomatic areas within a diseased leaf. The potential of 84 commonly used vegetation indices to find infection was explored. A machine-learning approach was used to create a classification model to automatically separate pixels into symptomatic, non-symptomatic and healthy classes. The success rate of the model was 69.7% using the full spectrum. It was very encouraging that by using a subset of only four broad bands, sampled to simulate a data set from a much simpler and less costly multispectral camera, accuracy increased to 71.3%. Next, the classification models were validated on field data. Infection in the field was successfully identified using classifiers trained on the entire spectrum of the hyperspectral data acquired in a lab setting, with the best accuracy being 64.9%. Using a subset of wavelengths, simulating multispectral data, the accuracy dropped by only 3 percentage points to 61.9%. This research shows the potential of using lab scans to train classifiers to be successfully applied in the field, even when simultaneously reducing the hyperspectral data to multispectral data.

Abstract Hyperspectral remote sensing can quickly, nondestructively and accurately monitor crop water concentration and provide technical support for winter wheat growth monitoring, drought assessment, and variable irrigation. In this study, canopy spectral reflectance, leaf water concentration (LWC), leaf nitrogen concentration (LNC), leaf area index (LAI), and leaf dry matter (LDM) of four wheat cultivars were measured under different irrigation and nitrogen treatments, and the effects of nitrogen treatment and growth period on spectral reflectance and LWC were analyzed. Canopy spectral reflectance for different growth periods, irrigation, and nitrogen treatments showed significant changes, leading to the phenomena of “nitrogen treatment differentiation” and “growth period differentiation” for the normalized difference spectral index [NDSI (762, 1458, 2301)] and normalized difference infrared index (NDII) monitoring models. To reduce the influence of nitrogen treatment and growth period on the LWC estimation model, a modified normalized difference water index (mNDWI) was constructed by introducing the nitrogen factor (ratio of left and right peak area, RIDA) into the optimized combination of water-sensitive bands [ND (815, 1080), ND (1585, 1740), and ND (2030, 2260)]. Compared with NDSI (762, 1458, 2301), the R2 of mNDWI was improved by 36.2%–41.1% under different nitrogen levels and 18.6%–22.4% in different growth periods; this effectively reduced the impact of nitrogen status on LWC monitoring and realized the unified modeling and accurate inversion of LWC for the entire growth period. The new index mNDWI, especially mNDWI (815, 1080) and mNDWI (2030, 2260), can effectively monitor the LWC status of wheat under different cultivation conditions, which is important for the real-time diagnosis of plant moisture to guide precision field irrigation applications.

Abstract Spatio-temporal variability of soil fertility and cocoa pod borer (CPB) infestation rate provides strategic information about the soil nutrients and CPB population densities at different harvest intervals. This enables the transitioning of cocoa fields (cooca-gliricidia and cocoa-coconut) from conventional to modern precision management. Geostatistical methods were applied to interpolate the data collected from a systematic grid based on a cluster of six cocoa tree stands for both fields and produce maps representing the spatial variability of all soil variables and CPB attack. Cocoa fresh bean weight and CPB infestation data were collected at two week-intervals from cocoa-gliricidia and cocoa-coconut. All field data points were geo-referenced by a differential global positioning system. Data were processed for possible outliers, and analysed by variography and interpolation techniques for quantification of spatial variability. Results showed that both plots exhibited definable spatial structures and were described by exponential models. Precision cocoa management recorded an increase in crop yield by 52.8 and 37.5% at cocoa-gliricidia and cocoa-coconut, respectively. Site-specific nutrient management and integrated pest control in the critical zones showed improvement in cocoa yields, especially during the peak harvest season.