Assessing the Accuracy of Soil and Water Quality Characterization Using Remote Sensing

Assessing the risks of agricultural management practices on agro-ecosystem sustainability has special relevance in Ohio, USA due to the states prominence in agricultural production. However, identifying detrimental management practices remains controversial, a situation that may explain the inability to halt the recurring harmful algal blooms in inland waters, or the build-up of nutrients in the agricultural soils. Thus, detailed and accurate information is required to identify soils and water susceptible to degradation, and to support counteractive remedial measures. In this study soil and water spectral reflectance data were acquired with an Analytical Spectral Device, and modeled with laboratory measured physical and chemical properties using the Analysis of Variance (ANOVA) and decision trees. Results reveal no site differences in pH for the water, but the differences in electrical conductivity (EC) were significant. Similarly, the pH for soils did not vary significantly with depth increments. However, the no till (NT) managed soils had significantly higher pH. EC varied with depth of the water, whereas the soil carbon: nitrogen (C/N) ratio varied with management in 4 out of 5 sites. Finally, this study shows that remotely sensed data can be utilized to effectively characterize agricultural management practices based on inherent soil and water properties, thus providing information critical for assessing the efficacy of Water Quality Trading initiatives.