Quantitative relationship model between soil profile salinity and soil depth in cotton fields based on data...

Area of Science:

• Agricultural Science
• Soil Science
• Environmental Science

Background:

• Soil salinization hinders crop nutrient uptake, impacting cotton production in southern Xinjiang, which constitutes over 60% of China's total.
• Effective monitoring and management of soil salinity are crucial for sustainable agriculture in arid and semi-arid regions.

Purpose of the Study:

• To monitor dynamic changes in soil salinity profiles in drip-irrigated cotton fields.
• To develop and improve a model for predicting soil salinity and its impact on cotton yield.
• To assess the effectiveness of the Kalman filter algorithm in enhancing salinity prediction accuracy.

Main Methods:

• Utilized multivariate linear regression to model soil salinity and depth relationships.
• Applied the Kalman filter algorithm to calibrate and improve the regression model's accuracy.
• Collected soil salinity data from drip-irrigated cotton fields in the Alaer Reclamation Area.

Main Results:

• The Kalman filter algorithm improved model accuracy, with R² increasing by up to 0.26 in July.
• The calibrated model achieved a high fitting accuracy (R² = 0.79) with low RMSE (96.17 μS cm⁻¹).
• Predicted cotton yields ranged from 5,203-5,551 kg hm⁻², with estimated income of 4,953-7,441 RMB hm⁻².

Conclusions:

• Kalman filtering enhances the prediction accuracy of soil salinity models in cotton fields.
• The improved model provides a basis for understanding soil salt migration and its relationship with cotton yield.
• Findings support efficient prevention and control strategies for saline soils in irrigated agricultural areas.