|
Activity Number:
|
362
- SPEED: Food, Environment, Biomedical Imaging and Physical System Visualization/Learning, Part 2
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, July 30, 2019 : 11:35 AM to 12:20 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
|
Abstract #307797
|
|
|
Title:
|
Subfield Yield Analysis for Precision Agriculture
|
|
Author(s):
|
Jarad Niemi* and Luis Damiano
|
|
Companies:
|
Iowa State University and Iowa State University
|
|
Keywords:
|
spatial;
hierarchical model;
missing data;
spatio-temporal;
geo-statistics;
Bayesian analysis
|
|
Abstract:
|
Recently, subfield analysis based on decades-old soil maps have suggested high variation in yield within midwestern row crop agricultural fields including areas with expected losses. Modern precision agriculture provides geo-referenced locations with associated yield promising a more recent, relevant, and spatially resolved analysis of yield. But these data are complicated by duplicated or missing locations, inconsistent yield calculations, and default software-imputed values among other data nuances. We proposed a novel statistical methodology for dealing with these issues as well as the spatial locations actually representing polygons that may overlap due to harvester widths and time intervals between observations. The proposed methodology assumes a grid approximation to the field and accumulates yield within the grid cells. After smoothing the data, a hierarchical model is constructed to borrow information across seasons and allow for weather-dependent variation in yield with an ultimate goal of identifying subfield areas with expected losses.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2019 program
|
