Predictions of Cu, Zn, and Cd Concentrations in Soil Using Portable X-Ray Fluorescence Measurements
Portable X-ray fluorescence (PXRF) measurements on 1520 soil samples were used to create national prediction models for copper (Cu), zinc (Zn), and cadmium (Cd) concentrations in agricultural soil. The models were validated at both national and farm scales. Multiple linear regression (MLR), random f...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-01-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/20/2/474 |
| _version_ | 1811187021191839744 |
|---|---|
| author | Karl Adler Kristin Piikki Mats Söderström Jan Eriksson Omran Alshihabi |
| author_facet | Karl Adler Kristin Piikki Mats Söderström Jan Eriksson Omran Alshihabi |
| author_sort | Karl Adler |
| collection | DOAJ |
| description | Portable X-ray fluorescence (PXRF) measurements on 1520 soil samples were used to create national prediction models for copper (Cu), zinc (Zn), and cadmium (Cd) concentrations in agricultural soil. The models were validated at both national and farm scales. Multiple linear regression (MLR), random forest (RF), and multivariate adaptive regression spline (MARS) models were created and compared. National scale cross-validation of the models gave the following R<sup>2</sup> values for predictions of Cu (R<sup>2</sup> = 0.63), Zn (R<sup>2</sup> = 0.92), and Cd (R<sup>2</sup> = 0.70) concentrations. Independent validation at the farm scale revealed that Zn predictions were relatively successful regardless of the model used (R<sup>2</sup> > 0.90), showing that a simple MLR model can be sufficient for certain predictions. However, predictions at the farm scale revealed that the non-linear models, especially MARS, were more accurate than MLR for Cu (R<sup>2</sup> = 0.94) and Cd (R<sup>2</sup> = 0.80). These results show that multivariate modelling can compensate for some of the shortcomings of the PXRF device (e.g., high limits of detection for certain elements and some elements not being directly measurable), making PXRF sensors capable of predicting elemental concentrations in soil at comparable levels of accuracy to conventional laboratory analyses. |
| first_indexed | 2024-04-11T13:55:12Z |
| format | Article |
| id | doaj.art-a7ae4431f01046b0a3ebe284b4dd84c3 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-11T13:55:12Z |
| publishDate | 2020-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-a7ae4431f01046b0a3ebe284b4dd84c32022-12-22T04:20:22ZengMDPI AGSensors1424-82202020-01-0120247410.3390/s20020474s20020474Predictions of Cu, Zn, and Cd Concentrations in Soil Using Portable X-Ray Fluorescence MeasurementsKarl Adler0Kristin Piikki1Mats Söderström2Jan Eriksson3Omran Alshihabi4Department of Soil and Environment, Swedish University of Agricultural Sciences, SE-75007 Uppsala/SE-53223 Skara, SwedenDepartment of Soil and Environment, Swedish University of Agricultural Sciences, SE-75007 Uppsala/SE-53223 Skara, SwedenDepartment of Soil and Environment, Swedish University of Agricultural Sciences, SE-75007 Uppsala/SE-53223 Skara, SwedenDepartment of Soil and Environment, Swedish University of Agricultural Sciences, SE-75007 Uppsala/SE-53223 Skara, SwedenDepartment of Soil and Environment, Swedish University of Agricultural Sciences, SE-75007 Uppsala/SE-53223 Skara, SwedenPortable X-ray fluorescence (PXRF) measurements on 1520 soil samples were used to create national prediction models for copper (Cu), zinc (Zn), and cadmium (Cd) concentrations in agricultural soil. The models were validated at both national and farm scales. Multiple linear regression (MLR), random forest (RF), and multivariate adaptive regression spline (MARS) models were created and compared. National scale cross-validation of the models gave the following R<sup>2</sup> values for predictions of Cu (R<sup>2</sup> = 0.63), Zn (R<sup>2</sup> = 0.92), and Cd (R<sup>2</sup> = 0.70) concentrations. Independent validation at the farm scale revealed that Zn predictions were relatively successful regardless of the model used (R<sup>2</sup> > 0.90), showing that a simple MLR model can be sufficient for certain predictions. However, predictions at the farm scale revealed that the non-linear models, especially MARS, were more accurate than MLR for Cu (R<sup>2</sup> = 0.94) and Cd (R<sup>2</sup> = 0.80). These results show that multivariate modelling can compensate for some of the shortcomings of the PXRF device (e.g., high limits of detection for certain elements and some elements not being directly measurable), making PXRF sensors capable of predicting elemental concentrations in soil at comparable levels of accuracy to conventional laboratory analyses.https://www.mdpi.com/1424-8220/20/2/474pxrfsoilcopperzinccadmiummachine learningprecision agriculture |
| spellingShingle | Karl Adler Kristin Piikki Mats Söderström Jan Eriksson Omran Alshihabi Predictions of Cu, Zn, and Cd Concentrations in Soil Using Portable X-Ray Fluorescence Measurements Sensors pxrf soil copper zinc cadmium machine learning precision agriculture |
| title | Predictions of Cu, Zn, and Cd Concentrations in Soil Using Portable X-Ray Fluorescence Measurements |
| title_full | Predictions of Cu, Zn, and Cd Concentrations in Soil Using Portable X-Ray Fluorescence Measurements |
| title_fullStr | Predictions of Cu, Zn, and Cd Concentrations in Soil Using Portable X-Ray Fluorescence Measurements |
| title_full_unstemmed | Predictions of Cu, Zn, and Cd Concentrations in Soil Using Portable X-Ray Fluorescence Measurements |
| title_short | Predictions of Cu, Zn, and Cd Concentrations in Soil Using Portable X-Ray Fluorescence Measurements |
| title_sort | predictions of cu zn and cd concentrations in soil using portable x ray fluorescence measurements |
| topic | pxrf soil copper zinc cadmium machine learning precision agriculture |
| url | https://www.mdpi.com/1424-8220/20/2/474 |
| work_keys_str_mv | AT karladler predictionsofcuznandcdconcentrationsinsoilusingportablexrayfluorescencemeasurements AT kristinpiikki predictionsofcuznandcdconcentrationsinsoilusingportablexrayfluorescencemeasurements AT matssoderstrom predictionsofcuznandcdconcentrationsinsoilusingportablexrayfluorescencemeasurements AT janeriksson predictionsofcuznandcdconcentrationsinsoilusingportablexrayfluorescencemeasurements AT omranalshihabi predictionsofcuznandcdconcentrationsinsoilusingportablexrayfluorescencemeasurements |