Location: Cropping Systems and Water Quality ResearchTitle: Improving canopy sensor algorithms with soil and weather information
|Sudduth, Kenneth - Ken|
|BEAN, G - University Of Missouri|
Submitted to: International Nitrogen Conference
Publication Type: Abstract Only
Publication Acceptance Date: 8/3/2016
Publication Date: 9/9/2016
Citation: Kitchen, N.R., Sudduth, K.A., Bean, G.M., Drummond, S.T., Yost, M.A. 2016. Improving canopy sensor algorithms with soil and weather information [abstract]. International Nitrogen Conference. Presentation.
Technical Abstract: Nitrogen (N) need to support corn (Zea mays L.) production can be highly variable within fields. Canopy reflectance sensing for assessing crop N health has been implemented on many farmers’ fields to side-dress or top-dress variable-rate N application, but at times farmers report the performance of this approach unsatisfying. Another study has shown that the performance of canopy sensing algorithms for rate N fertilization can be improved by including soil and weather factors. The objective of this analysis was to validate the performance of weather and soil modified corn algorithms using an independent dataset. The validation dataset was a 16-field investigation conducted over four growing seasons (2004-2007) on three major soil areas of Missouri: alluvium, deep loess, and claypan. Multiple blocks of randomized N rate response plots were arranged end-to-end so that blocks traversed the length of each field (400 to 800 m in length). Each block consisted of eight N treatments from 0 to 235 kg N/ha on 34 kg N/ha increments, side-dressed sometime between vegetative growth stages V7 and V11. Canopy sensing was done at the time of side-dress application. From these, the economic optimal N rate (EONR) was calculated and compared to the un-adjusted, weather-adjusted, and weather+soil-adjusted algorithm N recommendation rates. Generally, N rate recommendations were only slightly improved by the adjusted algorithms, mostly for loess and alluvium soil fields. This was true when examined by individual blocks or when EONR was calculated at the field level (average over all blocks). While on average, recommendations did not improve with the adjusted algorithms, the relationship between EONR and algorithm N recommendation did improve on claypan soils (r^2 values of 0.40, 0.82, and 0.89 for unadjusted, weather, and weather+soil algorithms, respectively). These results hint that soil and weather information may help improve canopy sensing N applications in some cases, but additional algorithm development and validation is still needed.