106093 Yield Monitor Data Cleaning Effect on Corn Grain and Silage Yield Determination.
Poster Number 1134
Wednesday, October 25, 2017
Tampa Convention Center, East Exhibit Hall
Yield monitor data for corn (Zea mays L.) are increasingly available. Accurate yield data can aid in assessment of whole farm nutrient use efficiency, field balances, corn yield potentials, and yield stability over time. However, comparable data across fields, farms, and years require standardization in processing of raw yield monitor data for both grain and silage. Our objective was to evaluate the impact of data cleaning protocols on final yield data at field level (with and without headlands) and at soil type level within fields. Corn silage data from 145 fields (3 farms) and grain data from 88 fields (3 farms) were processed. Comparisons were done to evaluate yields among three levels of cleaning: (1) none; (2) automated cleaning (“auto”) with filter settings derived for 10 fields per farm; and (3) automated cleaning with manual inspection for unrepresentative patterns, after the auto cleaning step was completed (“auto+”). The auto+ cleaning process was done by two people to evaluate people to people differences. Spatial Soil Management (SMS) was used to read raw data and transfer to Ag Leader format, while Yield Editor was used to clean the data (auto and auto+). Result showed the necessity of data cleaning, especially for corn silage. However, the manual evaluation of auto+ cleaning was not needed for any of the three spatial levels (less than 5% deviation between methods), as long as (1) each field or subfield included at least 100 harvester measurement points, and (2) a moisture filter was applied for corn silage data.