Predictive Tools for Developing Spatially Explicit Soil Properties for Applications at the Purdue Agronomy Center for Research and Education (ACRE)

Shams Rahmani, Purdue University
Soil illustration
Shutterstock

Abstract:  Soil organic matter content (OM) and cation exchange capacity (CEC) are important agronomic and edaphic soil properties. Accurate, high-resolution spatial information of OM and CEC are needed for high-resolution farm management and agronomic research. The objectives of this study were to: 1) create high resolution soil OM and CEC predictions in a low relief area using lidar elevation-based terrain attributes, and 2) compare the prediction accuracy of OM and CEC maps created by universal kriging (UK), Cubist, and random forest (RF). For this study, 174 soil samples were collected from 0 to 10 cm depth. The topographic wetness index (TWI), topographic position index (TPI), multi resolution valley bottom flatness (MrVBF), and multi resolution ridge top flatness indices (MrRTF) generated from the lidar data were used as covariates in model predictions. Based on an independent evaluation, no major differences were found in the prediction performance of all selected models. For OM, the predictive models provide results with R2 (0.44 – 0.45), RMSE (0.8 – 0.83 %), bias (0 – 0.22 %), and concordance (0.56 – 0.58). For CEC, the R2 ranged from 0.39 – 0.44, RMSE ranged from 3.62 – 3.74 cmolc kg-1, bias ranged from 0 – 0.17 cmolc kg-1, and concordance ranged from 0.55 – 0.57. We also compared the results to the USDA Soil Survey Geographic (SSURGO) data. For both OM and CEC, SSURGO was comparable with our predictive models, except for few map units where both OM and CEC were either higher or lower than predictive models. Although there were no statistically significant differences between predictive soil mapping and SSURGO, the soil-landscape relationship was maintained which can lead to further development of soil functional grouping which relate to crop phenotypic expression.

 

https://bluejeans.com/121646990/3043

Meeting ID: 121646990 / Participant passcode: 3043