Prediction of soil fertility properties in Southern Brazil via proximal sensing
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.Portable X-ray fluorescence (pXRF) combined with machine learning (ML) accurately predicts key soil fertility properties. This approach optimizes soil analysis for improved agricultural management in Brazil.
Area Of Science
- Soil Science
- Analytical Chemistry
- Agricultural Science
Background
- Proximal sensing and machine learning (ML) show promise for soil characterization.
- Effectiveness of these methods across varied soil conditions requires further investigation.
Purpose Of The Study
- To evaluate the efficiency of portable X-ray fluorescence (pXRF) for predicting 17 soil fertility properties in Rio Grande do Sul, Brazil.
- To compare the performance of six ML algorithms for pXRF-based soil analysis.
Main Methods
- Analyzed 468 soil samples using pXRF and conventional methods.
- Employed six ML algorithms: Projection Pursuit Regression, Partial Least Squares, Random Forest, Support Vector Machine, Extreme Gradient Boosting, and Cubist.
- Assessed prediction accuracy using R², RMSE, nRMSE, RPD, and RPIQ.
Main Results
- Cubist and Random Forest algorithms demonstrated superior performance.
- High prediction accuracies (R²) were achieved for available/exchangeable Al, Mg, Mn, Cu, K, P-rem, H+Al, and total N.
- Further research is needed for predicting organic carbon and available B, Fe, Na, Zn.
Conclusions
- pXRF coupled with ML algorithms offers an efficient method for soil fertility assessment.
- This approach can accelerate decision-making for agricultural management in Brazil.
- Optimized soil analysis leads to improved crop management strategies.
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
