Ground Observations and Environmental Covariates Integration for Mapping of Soil Salinity: A Machine Learning-Based Approach

Abstract

Soil salinization is a severe danger to agricultural activity in arid and semi-arid areas, reducing crop production and contributing to land destruction. This investigation aimed to utilize machine learning algorithms to predict spatial soil salinity (dS m−1) by combining environmental covariates derived from remotely sensed (RS) data, a digital elevation model (DEM), and proximal sensing (PS). The study is located in an arid region, southern Iran (52°51′–53°02′E; 28°16′–28°29′N), in which we collected 300 surface soil samples and acquired the spectral data with RS (Sentinel-2) and PS (electromagnetic induction instrument (EMI) and portable X-ray fluorescence (pXRF)). Afterward, we analyzed the data using five machine learning methods as follows: random forest—RF, k-nearest neighbors—kNN, support vector machines—SVM, partial least squares regression—PLSR, artificial neural networks—ANN, and the ensemble of individual models. To estimate the electrical conductivity of the saturated paste extract (ECe), we built three scenarios, including Scenario (1): Synthetic Soil Image (SySI) bands and salinity indices derived from it; Scenario (2): RS data, PS data, topographic attributes, and geology and geomorphology maps; and Scenario (3): the combination of Scenarios (1) and (2). The best prediction accuracy was obtained for the RF model in Scenario (3) (R2 = 0.48 and RMSE = 2.49), followed by Scenario (2) (RF model, R2 = 0.47 and RMSE = 2.50) and Scenario (1) for the SVM model (R2 = 0.26 and RMSE = 2.97). According to ensemble modeling, a combined strategy with the five models exceeded the performance of all the single ones and predicted soil salinity in all scenarios. The results revealed that the ensemble modeling method had higher reliability and more accurate predictive soil salinity than the individual approach. Relative improvement (RI%) showed that the R2 index in the ensemble model improved compared to the most precise prediction for the Scenarios (1), (2), and (3) with 120.95%, 56.82%, and 66.71%, respectively. We applied the best model in each scenario for mapping the soil salinity in the selected area, which indicated that ECe tended to increase from the northwestern to south and southeastern regions. The area with high ECe was located in the regions that mainly had low elevations and playa. The areas with low ECe were located in the higher elevations with steeper slopes and alluvial fans, and thus, relief had great importance. This study provides a precise, cost-effective, and scientific base prediction for decision-making purposes to map soil salinity in arid regions.

Quantitative Estimation of Saline-Soil Amelioration Using Remote-Sensing Indices in Arid Land for Better Management

“…Cotton and sugar beet are the most important salt-tolerant crops, while sugar cane, fava beans, and peas are the least tolerant of salinity. Most vegetable crops are moderately resistant to salinity, while most fruit crops, especially deciduous, are sensitive to salinity [114][115][116][117][118][119][120][121][122][123][124].…”

Towards the Improvement of Soil Salinity Mapping in a Data-Scarce Context Using Sentinel-2 Images in Machine-Learning Models

“…Since satellite optical images have a global coverage and a periodic revisit time, many authors investigated their ability to monitor soil salinity. Generally, soil salinity and/or vegetation indexes are derived from the combination of different spectral bands to train machine-learning models to retrieve soil salinity estimates [10][11][12][13][14][15][16][17][18][19]. With a higher spatial resolution than MODIS (500 m) and Landsat (30 m), and a revisit time of 5 days, Sentinel-2 (10 m) is particularly suitable for that purpose.…”

“…In northern Iran, Classification and Regression Trees (CART), RF, and SVM models were used on Sentinel-2 images and RF provided the most reliable estimates [25]. In southern Iran, after use of ANN, RF, SVM, PLSR and k-Nearest Neighbor (kNN) models [11], RF and SVM achieved the best soil salinity prediction accuracy. According to the above-mentioned studies, RF and SVM appear as the most efficient models to estimate soil salinity from Sentinel-2 images.…”

Multifrequency electromagnetic induction soil moisture characterization under different land uses in western Newfoundland

“…The extensively sampled variable (covariate) is usually measured more cheaply and quickly than the target variable. EC a measured using EMI sensors can be used as a covariate for mapping of soil moisture (Tarr et al 2005;Naimi et al 2021). Spatial variability in soil moisture can be predicted by ordinary kriging from the limited soil samples and by cokriging when a strong statistically significant relationship exists between soil moisture and an easily measured auxiliary variable.…”