Moran eigenvector filtering of multi-year yield data with application to zone development

Authors: Long, Daniel; GRIFFITH, DANIEL - University Of Texas; KVIEN, CRAIG - University Of Georgia; CLAY, DAVID - South Dakota State University

Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/31/2020
Publication Date: 3/10/2021
Publication URL: https://handle.nal.usda.gov/10113/7709351
Citation: Long, D.S., Griffith, D.A., Kvien, C.V., Clay, D.E. 2021. Moran eigenvector filtering of multi-year yield data with application to zone development. Agronomy Journal. 4(1). Article e201404. https://doi.org/10.1002/agg2.20140.
DOI: https://doi.org/10.1002/agg2.20140

Interpretive Summary: Crop yield mapping in the same field after a series of years may provide information for identifying consistently high or low yielding zones within a field. But unstable areas remain that are difficult to interpret and manage. The objective of this study was to examine if an analytical data filtering technique, termed: Moran eigenvector spatial filtering (MESF), could provide useful information for describing the variation in multi-year crop yield data that is shared across years and for constructing static MZ for precision agriculture. This study found that multi-year yield data consist of spatially structured and unstructured random effects (i.e., error). Eigenvector spatial filtering captured structured random effects shared across years. This structured variation possesses less noise than original, unfiltered data. Distinctive map patterns were revealed that proved useful for construction of static management zones for precision agriculture.

Technical Abstract: A time-series of yield monitor data may be used to identify field areas of consistently low or high yield to serve as management zones (MZ) for variable-rate application. However, transient factors that affect yield in one year, but not every year, detract from this approach. The objective of this study was to examine if Moran eigenvector spatial filtering (MESF) could provide useful information for describing the variation in multi-year crop yield data that is shared across years and for constructing static MZ for precision agriculture. The MESF technique was applied to 7-yr of corn and soybean yield rotation data from a dryland field in east-central South Dakota and 8-yr of corn, soybean, and peanut rotation data from an irrigated field in southwest Georgia. A random effects (RE) model was estimated that utilized eigenfunctions of a geographic connectivity matrix to account for spatially structured (SSRE) and unstructured RE in standardized Z scores of multi-year crop yield. The MESF method was evaluated with conventional averaging of unfiltered yield data as a reference for comparison. In South Dakota, the SSRE accounted for 26% of the yield variance shared across years. Distinct patterns appeared to be related to changes in soil type and landscape position. The Georgia field yielded similar results. MESF is effective for revealing structured variation in a time-series of yield monitor data and may be useful for defining static MZ within fields.