Spatiotemporal Assessment of Soil Organic Carbon Change Using Machine-Learning in Arid Regions

Hassan Fathizad,Mohammad Ali Hakimzadeh Ardakani,Brandon Heung,Thomas Scholten,Mojtaba Zeraatpisheh,Ruhollah Taghizadeh-Mehrjardi

doi:10.3390/agronomy12030628

Hassan Fathizad, Mohammad Ali Hakimzadeh Ardakani + Show 4 more

Open Access

https://doi.org/10.3390/agronomy12030628

Copy DOI

Abstract

Soil organic carbon (SOC) is an essential property of soil, and understanding its spatial patterns is critical to understanding vegetation management, soil degradation, and environmental issues. This study applies a framework using remote sensing data and digital soil mapping techniques to examine the spatiotemporal dynamics of SOC for the Yazd-Ardakan Plain, Iran, from 1986 to 2016. Here, a conditioned Latin hypercube sampling method was used to select 201 sampling sites. A set of 37 environmental predictors were obtained from Landsat imagery taken in 1986, 1999, 2010 and 2016. Here, SOC was modeled for 2016 using the Random Forest (RF), support vector regression (SVR), and artificial neural networks (ANN) machine-learners by correlating environmental predictors with soil data. The results showed that RF yielded the highest accuracy (R2 = 0.53), compared to the other two learners. By performing a variable importance analysis of the RF model, normalized difference vegetation index, modified vegetation index, and ground-adjusted vegetation index were determined to be the most important environmental predictors. By applying the model calibrated from 2016 data to 1986, 1999 and 2010, the results showed a substantial decrease in SOC; these decreases in SOC were mainly attributed to land use changes and agricultural activities.

Highlights

Soil organic carbon (SOC) has a significant impact on many soil functions, such as the production of food and other biomass; and the provisioning of biological habitats and genetic resources
The results show that from 1986 to 2016, the classes with SOC content > 0.6% and 0.3–0.6% decreased by 25,888 ha (5.26%) and 138,272 ha (28.63%), respectively
This study demonstrated the effectiveness of the Random Forest (RF) model for predicting the spatio‐ temporal patterns of SOC content of the oasis and arid‐agroecosystem area which the ap‐ proach may be utilized in other arid conditions

Summary

Introduction

Soil organic carbon (SOC) has a significant impact on many soil functions, such as the production of food and other biomass; and the provisioning of biological habitats and genetic resources It is an important indicator for assessing and managing soil fertility, soil quality, and soil degradation [1,2]; accurate information on the spatiotemporal patterns of SOC are required to support sustainable land use and management. DSM approaches involve the creation and operation of terrestrial spatial information systems obtained from field and laboratory observations, combined with spatial and non-spatial inference systems to generate raster-based map prediction and their respective uncertainty estimates [10–12] These approaches apply statistical tools to quantify the relationships between soil properties and environmental predictors [8]. The Random Forest (RF) learner has become increasingly popular in DSM research [13,19,20]

Objectives

Methods

Findings

Conclusion