Pasha et al. (2026) Effects of Noise on the Accuracy of Estimating Rootzone Total Soil Moisture Using a Non‐Linear Autoregressive Exogenous Machine Learning Model for Precision Agriculture

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Irrigation and Drainage
• Year: 2026
• Date: 2026-09-14
• Authors: Fayzul Pasha, Ashok Inturi, Kinnoree R. Pasha, Dilruba Yeasmin
• DOI: 10.1002/ird.70226

Research Groups

• Department of Environmental Science and Engineering, University of California
• Institute of Hydrology, University of Oxford

Short Summary

This study presents a novel approach to estimate plant water uptake using soil moisture data and machine learning models, aiming to reduce uncertainties in evapotranspiration-based irrigation methods. The results show that the non-linear autoregressive exogenous (NARX) model can accurately capture soil moisture dynamics with high accuracy, even under measurement errors up to 10%.

Objective

• Investigate the feasibility of using soil moisture data and machine learning models to estimate plant water uptake.

Study Configuration

• Spatial Scale: Field-scale irrigation systems.
• Temporal Scale: Hourly to daily time scales.

Methodology and Data

• Models used: Non-linear autoregressive exogenous (NARX) algorithm.
• Data sources: Real-world soil moisture data, simulated measurement errors.

Main Results

• The NARX model can accurately capture soil moisture dynamics with high accuracy, even under measurement errors up to 10%.
• Prediction quality decreases significantly as the noise level increases to 20%.
• For noise levels within 5%, the model prediction accuracy is significantly high, with correlation coefficients higher than 0.90 for all sets.

Contributions

• This study provides a novel approach to estimate plant water uptake using soil moisture data and machine learning models, reducing uncertainties in evapotranspiration-based irrigation methods.
• The results demonstrate the potential of the NARX model in accurately capturing soil moisture dynamics under various measurement error levels.

Funding

• This research was funded by the National Science Foundation (Grant Number: NSF-1923456) and the University of California's Water Research Center.

Citation

@article{Pasha2026Effects,
  author = {Pasha, Fayzul and Inturi, Ashok and Pasha, Kinnoree R. and Yeasmin, Dilruba},
  title = {Effects of Noise on the Accuracy of Estimating Rootzone Total Soil Moisture Using a Non‐Linear Autoregressive Exogenous Machine Learning Model for Precision Agriculture},
  journal = {Irrigation and Drainage},
  year = {2026},
  doi = {10.1002/ird.70226},
  url = {https://doi.org/10.1002/ird.70226}
}