Tuesday, November 6, 2007
193-2

Mobile Soil pH Sensor: Initial Assessment in Central Kentucky.

Adam Pike, Thomas Mueller, Frank Sikora, and Blazan Mijatovic. University of Kentucky, University of Kentucky, Pssb-n-122 Ag Science Center North, Lexington, KY 40546

Capturing the spatial variability of soil pH across a given landscape can be hard to obtain when trying to produce a quality variable rate liming application map. Grid sampling has proven to be the most economically friendly approach at understanding large scale variation but still lacks in explaining the small scale variability. On the go soil pH mapping will be evaluated to test the validity of the measurements. A Veris Mobile Sensor Platform (MSP) equipped with both electrical conductivity (EC) capability and a two soil pH electrodes will be used to take the measurements. At each point of measurement, a soil sample will be taken from adjoining sites and from the instrument itself to verify the on the go measurement. The Veris can take up to 20 pH samples per acre (depends on speed) as compared to one composite sample done on a grid basis representing either 1 or 2.5 acres. This means more samples to represent the entire field. Samples were sent to Regulatory Services for pH analysis. Given the small scale variability being captured by the Veris instrument, measured versus predicted regression functions have been shown to be nearly identical with a RMS of 0.2188 and a SE of 0.04084. A zig zag pattern was driven throughout the field to derive a validation data set along with random samples. The zig zag pattern seemed to over predict whereas the random samples over predicted but not as much. Additional results will come as the data set is analyzed further.