# Evaluate Accuracy of Fertilizer Application ## **Context** In modern agriculture, most agronomic operations are executed based on pre-prepared Variable Rate Application (VRA) maps. These maps typically contain multiple fertilizer rates tailored to different field zones, addressing the inherent heterogeneity of soil and crop conditions. By following such VRA maps, farmers aim to optimize agricultural input usage and maximize yields. In this case, we will focus on evaluating the accuracy of a Nitrogen fertilizer application by comparing the intended application rates (as indicated in the VRA map) to the actual application rates recorded by equipment during the field operation. ## **Used Machinery-Logged Data** After the fertilizer application is completed, the machinery’s onboard systems record various parameters related to the operation. Of particular interest here are two key metrics: * **Target Rate****:** The intended amount of fertilizer to be applied at any given location, as specified in the application map (VRA or Flat Rate Application). * **Applied Rate****:** The actual amount of fertilizer applied at that same location. By examining these two attributes in a geospatial context, we can evaluate how closely the actual application followed the prescribed plan. The comparison involves visualizing both **Target Rate** and **Applied Rate** maps, analyzing their statistical distributions, and confirming that the required amount of fertilizer was indeed delivered accurately.

Machinery-Logged Data: Target Rate vs Applied Rate for Flat Rate Application

Machinery-Logged Data: Target Rate vs Applied Rate for Variable Rate Application (VRA)

## **Assess Application Accuracy** To assess the accuracy of the fertilizer application, we will utilize a pre-saved Equation named Spatial Correlation Analysis (Data Layers Similarity) that measures the similarity between Applied Rate and Target Rate values on a spatial basis.

Search the existing Data Layers Similarity Equation

The similarity values range from 0 to 1, where 0 indicates no match and 1 signifies 100% value-spatial match. In other words, the closer the similarity score is to 1, the more accurately the fertilizer application was executed. Achieving values consistently near 1 indicates precise adherence to the application plan (VRA or FlatRate) and ensures that the field receives the intended amount of fertilizer for optimal crop performance.