Landsat phenology and machine learning enables sub-field corn yield mapping at 30-meter resolution

SSRN Electronic Journal · 2026-01-01

Midwestern farmers' willingness to engage with precision agriculture technologies and on‐farm precision experimentation

Agronomy Journal · 2026-04-24

Abstract Precision agriculture technologies (PATs) have revolutionized the agriculture industry and provide many benefits to farmers. Among these benefits is the ability to conduct experiments in a process known as on‐farm precision experimentation (OFPE). By conducting these experiments and through collaboration with researchers, crop consultants, and extension agents, farmers can learn site‐specific management practices to better address challenges in their operations. However, adoption rates of these technologies have remained below 50% in the United States. Furthermore, very few studies have explored the factors that influence a farmer's decision to conduct experiments, or their willingness to collaborate with external stakeholders such as researchers, crop consultants, and extension agents. Therefore, the objectives of this study are to measure farmers' perceptions of PATs and OFPE, identify farmer characteristics associated with OFPE, and explore their willingness to engage in collaborative OFPE. Respondents perceive that PATs help them make better decisions for their operations (84.6%), and 96.3% of those who conduct on‐farm experiments use PATs to do so. Operators whose farms are 1,000 acres or larger, who operate on rented acreage, and who adopt PATs are more likely to engage in on‐farm experimentation. While most OFPE is conducted without external engagement, nearly half of respondents are interested in collaborative OFPE. These results may help guide researchers, crop consultants, and extension personnel as they look for opportunities to collaborate with farmers. This collaboration can help address questions of mutual interest and work toward solutions to the economic and environmental challenges facing agriculture.

Harvesting insights: interpretable machine learning to understand environmental drivers of U.S. maize and soybean yield

Scientific Reports · 2026-02-13 · 1 citations

Accurate crop yield prediction is crucial for enhancing food security and agricultural sustainability; however, existing models frequently struggle to capture the intricate relationships between environmental drivers and crop performance. Here we leveraged a large, spatially explicit yield monitor dataset of U.S. commercial maize (Zea mays) and soybean (Glycine max) fields (134 unique crop-site-years). Machine learning models were trained to predict yield with high accuracy (R2 > 0.87, RMSE < 1.13 Mg ha−1), and Shapley Additive Explanations were used to quantify how weather, soil, and terrain properties predict yield variability. Our results highlight the potential of machine learning to disentangle environmental constraints on crop production, thereby providing actionable insights for more resilient U.S. food systems. The results presented here represent a novel approach to identifying maize and soybean yield constraints that can inform the next generation of crop breeding and precision management strategies.

Quadratic-plateau geographically weighted regression model for estimating site-specific economically optimal input rates

Computers and Electronics in Agriculture · 2025-06-18 · 2 citations

• QP-GWR model addresses limitations of traditional GWR yield-input models. • Simulations demonstrate QP-GWR’s enhanced statistical and economic performance. • Model rankings show a gap between statistical accuracy and economic viability. • Dual metric evaluation is crucial for agricultural decision-making tools. • User-friendly QP-GWR Python implementation promotes on-farm precision strategies. In precision agriculture, site-specific input management is crucial for optimizing input rates according to unique field conditions allowing farmers to maximize their profits. However, traditional yield-input models like Geographically Weighted Regression (GWR) are hindered by the yield response functions misspecification problem, leading to biased input recommendations. This study develops a Quadratic-Plateau Geographically Weighted Regression (QP-GWR) model to estimate site-specific economically optimal input rates (EOIR), overcoming limitations in existing GWR models by adequately handling quadratic-plateau yield responses. Focused on corn nitrogen management, the study validates QP-GWR’s effectiveness through extensive simulations and on-farm trials, demonstrating its superior statistical and economic performance. Notably, compared to existing GWR models, QP-GWR achieves smaller bias and Root Mean Square Error (RMSE) in EOIR estimation, and enhances profits by $0.60 to $13.90 per ha, underscoring the critical role of this novel QP-GWR in modeling crop yield responses to management inputs. Additionally, the observed discrepancies between statistical and economic metrics underline the importance of evaluating both aspects in agricultural decision-making. Overall, these findings demonstrate our developed QP-GWR’s capability to enhance on-farm precision experimentation recommendations, particularly when dealing with plateauing yield responses. The user-friendly Python implementation of QP-GWR is available at: https://github.com/rssiuiuc/QP-GWR to facilitate broader adoption and application in precision agriculture.

Estimating Planting and Harvest Dates from Remotelysensed Data

SSRN Electronic Journal · 2025-01-01

Measuring the estimation bias of yield response to N using combined on‐farm experiment data

Journal of the Agricultural and Applied Economics Association · 2025-07-17

Abstract Accurately evaluating yield response to nitrogen is essential for increasing farm profitability. Data often come from randomized experiments, ensuring nitrogen is independent of other factors. However, when data from multiple experiments are combined, as many studies do, correlations between nitrogen and unobserved field heterogeneity can arise, potentially leading to biased results if the endogeneity problem is not addressed in regression analysis. This study examines the bias caused by this endogeneity using data from 41 large‐scale on‐farm precision experiments. We find that this bias can be both statistically and economically significant.

Optimizing precision conservation with on-farm precision experiment data: The role of crop insurance and spatially variable profit

The Science of The Total Environment · 2025-04-22

High Density Observation Network for Near Real Time Detection of Hail Events

Lecture notes in mobility · 2025-09-02

Abstract This paper outlines the problems associated with sudden onset hail events on the Transport Infrastructure Ireland road network which have led to a number of road traffic accidents in recent years. In an attempt to minimise these accidents a dense network of weather observing sensors is defined and explained which enable rapid identification of events as they happen in order to warn drivers approaching the event area. The paper sets out the problem in detail and also identifies the sensors to be used in the deployed network. This incudes the novel use of IoT based sensors as supplementary to a grid of reference grade present weather detectors. The methodology for the network design is explained and the primary network is shown with reference to the need to detect hail events that can be of relatively small footprint area, hence explaining the need for a dense observing network to detect these events in real time.

Bias in economic evaluation of variable rate application based on geographically weighted regression models with misspecified functional form

Journal of the Agricultural and Applied Economics Association · 2024-03-01 · 4 citations

Abstract Geographically weighted regression (GWR) has been presented as a valuable tool for estimating site‐specific yield response functions to derive recommendations of variable rate input. This study employs Monte Carlo simulations to illustrate that if GWR assumes a quadratic yield response functional form while the actual yield‐input relationship is quadratic‐plateau, it can significantly overestimate the economic value of variable rate application compared to its true value. Practitioners in precision agriculture should exercise caution when utilizing GWR for site‐specific input recommendations. Statistical community is also encouraged to develop tools in software packages providing GWR that allow more flexibility in functional form assumptions.

Quadratic-Plateau Geographically Weighted Regression Model for Estimating Site-Specific Economically Optimal Input Rates

SSRN Electronic Journal · 2024-01-01 · 1 citations