Spatially and Temporally localized Production Efficient Modelling

Keywords: Yield, Corn, Production Efficient Modelling, PEM, NPP, GPP
Session Type: Virtual Paper Abstract
Day: Tuesday
Session Start / End Time: 3/1/2022 09:40 AM (Eastern Time (US & Canada)) - 3/1/2022 11:00 AM (Eastern Time (US & Canada))
Room: Virtual 56

Abstract

Earth Observation applied to yield indices allows for the democratization of yield calculation and the extrapolation algorithms at larger scales, and finer temporal ranges. Modern production efficiency models apply a semi-empirical process that includes Earth Observation products (Net Primary production and Gross Primary Production) to create a yield index for a defined area. Production efficiency models also include actual yield, a fixed Harvest Index, moisture content, and root to shoot ratio. Here a new production efficiency model is introduced that includes a spatially/temporally dynamic Harvest index, efficiency models. Former models use a fixed normalized harvest index, Root to shoot, and moisture content for a singular crop type over a period of years. Fixed approaches in past papers have led to a need for post-processing optimization (correction) and a lack of hardiness in overall results to climatically abnormal years. This approach is extracted across Nebraska’s rainfed cornfields over a selective 16-year time period. Net Primary production and Gross Primary Production were calculated dynamically as well based on the maturity of the Corn before harvest. These dynamic calculations, allow for the removal of optimizations found in production efficiency models as a form of post-processing correction. Further changes in resolution via a change to Landsat 8 from MODIS results in higher spatial dimensionality in a singular field. Dynamic changes result in a yearly producers accuracy across a single field of <20%, in a 16 year climatically variable period.