Prediction of Soil Properties at Farm Scale Using a Model-Based Soil Sampling Scheme and Random Forest

Abstract

Digital soil mapping techniques can reduce the costs associated with improving soil information at the farm scale. Many techniques exist for generating a digital soil mapping. In this study, we tested the use of a random forest (RF) and two model-based soil-sampling scheme, conditioned with Latin hypercube (cLHS) and fuzzy c-means sampling (FCMS) to predict soil properties for a 68-ha agricultural field. The predictor set included the use of inexpensive auxiliary information such as digital elevation models, yield maps, surface reflectance, and apparent soil electrical conductivity measurements. Our results support the underlying assumption that both cLHS and FCMS capture adequately the full predictor distribution in the Pampas conditions. Despite the complex local variation of soils and the instability of normalized difference vegetation index and yield maps, the RF models predicted soil properties at farm scale. The RF model with cLHS and FCMS explained about 65% (R 2 = 0.648) and 57% (R 2 = 0.571) of the variability of soil organic matter and about 71% (R 2 = 0.714) and 68% (R 2 = 0.686) of the variability of clay, respectively. Our results provide a significant improvement of information related to soil organic matter and clay content with respect to the soil survey available. Extrapolation and applicability of this study to other areas remain to be tested.

How to compare sampling designs for mapping?

“…In our case studies, SRS and cLHS were equivalent in terms of map accuracy. Several studies (e.g., Castro‐Franco, Costa, Peralta, & Aparicio, 2015; Chu, Lin, Jang, & Chang, 2010; Contreras, Ballari, De Bruin, & Samaniego, 2019; Domenech, Castro‐Franco, Costa, & Amiotti, 2017; Schmidt et al, 2014) concluded that cLHS in combination with kriging or random forest for mapping gave the most accurate prediction. These studies promote the use of cLHS as an effective sampling design for mapping.…”

Industrial agriculture and agroecological transition systems: A comparative analysis of productivity results, organic matter and glyphosate in soil

“…ECa measurements were collected on September 9th, 2016 using a Veris® 3100 soil electrical conductivity sensor (Veris Technologies Inc., Salina, KS, USA). With this sensor, the system records ECa in mS m −1 by electrical resistivity at a shallow depth (0 to 30 cm, ECa 0-30 cm) and deep depth (0 to 90, ECa 0-90 cm) (Castro Franco et al, 2015). ECa measurements were made along parallel transects approximately 20 m apart on the surface of the experimental site.…”

Sample Size Optimization for Digital Soil Mapping: An Empirical Example

“…Techniques to estimate optimal sample size are critical to supporting soil management, crop productivity, and environmental conservation. Several applications include the prediction of crop yields [8,9], spatial variability of soil properties at the farm scale [10,11], and precision agriculture prescriptions [12,13] in agricultural systems. Sample size plays a critical role in predictive modeling of aboveground biomass and species distribution in forest systems [14,15] and in the calibration of models to predict soil properties from spectroscopy data [16,17] in soil science.…”