COSMOS (COsmic-ray Soil Moisture Observing System) probes are a developing instrument in the field of remote sensing and are currently the only source of field scale soil moisture data. A COSMOS probe is essentially a hydrogen particle counter with a measurement depth ranging from 10-70 cm and a spatial resolution approaching 700 m in diameter. A majority of the hydrogen detected is in the form of water molecules. As vegetation contains both hydrogen in the vegetative dry matter and water, in addition to sitting above the ground, it may influence what is measured by the COSMOS probe as soil moisture. We hypothesized that COSMOS probes located in areas with rapid growth vegetation need to be checked for a vegetation effect on their soil moisture measurements.

In order to account for vegetation, however, a method of monitoring and modeling the amount of vegetation was needed. We assumed that an allometric relationship of stem diameter and canopy height would be able to model the moisture in a plant. In-situ soil samples were taken concurrently with the vegetation measurements and used to recalibrate the COSMOS probe's maximum counting rate parameter, N_0. N_0 was thought to be a site-specific constant that took background hydrogen into account. However, we found that the vegetation hydrogen still needs to be accounted for because as the vegetation grew, N_0 decreased and as the vegetation senesced, N_0 increased. This discovery led us to create a vegetation-corrected calibration of N_0 that could be used to remove the vegetation water column effect from what the COSMOS probe is observing over the crop season.

Aside from remote sensors, estimation of soil moisture over large areas requires either numerous, exhaustive point samples, or some method of upscaling a smaller number of measurements. Upscaling point samples to a larger area can be difficult. A new method for upscaling point scale measurements is the Feature Space Interpolation (FSI) method. FSI uses K-means clustering to separate a larger area into cluster groups. The location thought to best represent the cluster is identified for sampling purposes. Soil moisture samples can be taken at each of these points as a representation of the whole cluster. With our vegetation-corrected calibration of the COSMOS probe, we were able to validate the areal soil moisture within the COSMOS footprint found with the FSI method. We found that FSI can upscale soil moisture values to within 0.063 cm^3 cm^{-3} of what the COSMOS probe reports for soil moisture.