Satellite- and Ground-Based Measurement of WDVI to Reduce Herbicide Use in Potato Haulm Killing.

When using variable-rate technology (VRT), tractor-mounted sensors are typically used to measure crop status. Crop status can also be measured with a satellite. The first objective of our work was to establish a relationship between the Weighted Difference Vegetation Index (WDVI) in potato as measured with a Cropscan reflectance meter on the one hand and WDVI as measured with the Aster and Worldview satellites on the other hand. The scale at which VRT is applied is an important determinant of the reduction in input use. The second objective was to estimate the potential to reduce herbicide use for potato haulm killing as a function of the size of decision units, using the above-determined relationship between Cropscan-WDVI and satellite-WDVI, satellite imagery of 13 producers’ fields and a previously developed decision rule for herbicide application. We found that ground-based WDVI and satellite-based WDVI are strongly and linearly related, thus making it feasible to calculate herbicide application rate for potato haulm killing on the basis of satellite-based measurements. We found that when the size of the decision unit was 15 x 15 m² (the size of an Aster pixel), a  reduction in herbicide use of at least 50% would be achieved in one out of every two of the fields, and a reduction of at least 33% would be achieved in all fields. When the size of the decision unit was 30 x 30 m², a reduction of at least 33% would be achieved in one out of every two of the fields. In conclusion, satellite-measurements can be used in place of ground-based crop reflectance measurements for determining herbicide application rate for potato haulm killing. Our work confirms previous estimates that a 50% reduction in herbicide use for potato haulm killing can be achieved with VRT.