Statistics in Precision Agriculture
Organizer: E. Barry Moser
(bmoser@lsu.edu)
Louisiana State University

Description:
Precision agriculture deals with the precise control of agricultural operations, such as soil fertilization, herbicide and pesticide applications, and irrigation, in such a way as to reduce costs (including costs of materials and costs due to environmental impact) while maintaining or increasing yields. These techniques may involve geostatistical maps and models, spatial models for plant diseases and arthropod population dynamics, incorporation of remotely sensed data, data collected on different spatial scales and with different measurement precision, and may use specialized sampling schemes taking advantage of or accounting for spatial correlations.

This session will focus on statistical issues and methodologies that arise in precision agriculture settings. Precision agriculture data are collected on such small spatial scales that enormous data sets are generated. How are these data to be analyzed to extract as much information as possible? How should remotely sensed data, such as satellite imagery, be combined with crop yields measured at 3 meter intervals and with soil samples taken every 25 meters, each on different grids? Given that yields, disease incidence, and arthropod abundance are spatially correlated, how should one sample to determine critical levels and treatment responses? Many statistical questions need to be addressed in order to provide important information useful in performing agricultural management on precise scales.

Format:
The format will be 3 (30 minute) papers with a discussant. Presenters will focus on issues in precision agriculture that each is dealing with in conjunction with their Agricultural Experiment Stations.

Participants:
Robert G. Downer (presentation, Hypothesis Tests in the Presence of Spatial Correlation)
Bob Downer has been an assistant professor of experimental statistics at Louisiana State University for 2 years. His research interests include spatial statistics, smoothing, disease mapping and biostatistical applications. He has worked as a statistical consultant, biostatistician and survey methodologist. In an experimental design context, Bob will speak on the adjustment of hypothesis tests in the presence of spatial correlation. This work was motivated through collaboration with precision agriculture researchers at the Dean Lee Experiment Station of the Louisiana Agricultural Center.

Patrick D. Gerard, David Evans, and Michael Cox (presentation, Statistical Issues in the Analysis of Remotely Sensed Data as Pertains to Precision Agriculture)
Pat Gerard is Associate Professor of Experimental Statistics at Mississippi State University. His research interests include nonparametric regression and density estimation and line transect sampling. His talk will focus on some of the basic statistical issues in analysis of remotely sensed data as it pertains to precision agriculture. Many of these issues have arisen from preliminary analyses of data collected in conjunction with a grant from NASA supporting an investigation into the use of remote sensing data in agriculture.

Raúl E. Macchiavelli and Rocío del P. Rodríguez (presentation, Use of Spatial Statistics to Design Sampling Plans for Monitoring Rust in Coffee Trees)
Raul Macchiavelli is an associate professor of Biometry at the University of Puerto Rico. He has ample experience as a statistical consultant in Argentina, the US and the Caribbean. His current research interest include repeated measures models and spatial statistics. Raul will present an application of spatial models for designing sampling plans for plant disease control. This work originated in a collaboration with plant pathologists at the Agricultural Experiment Station of the University of Puerto Rico.

James Beaver and Raúl E. Macchiavelli (poster, Effect of Number of Seed Bulked when Using the Multiple-Seed Procedure for Self-Pollinated Crops)
University of Puerto Rico

Bradley Tiffee and Robert G. Downer (poster, Geostatistical Analysis of Spatial Nutrient Data in a Precision Agriculture Experiment)
Louisiana State University

Discussant:
Barry Moser