Summary
This peer-reviewed study demonstrates that Random Forest, an ensemble machine learning algorithm, can estimate root zone soil moisture with comparable or slightly superior accuracy to process-based hydrological models, particularly for interpolation tasks. A key advantage of the Random Forest approach is its independence from soil hydraulic parameters, making it potentially more practical for data-poor agricultural regions where such parameters are incomplete or unavailable. However, the model showed reduced accuracy during extreme moisture conditions, attributed to sparse training data and incomplete representation of subsurface processes in the selected covariates.
UK applicability
The methodology is directly applicable to UK agricultural and hydrological contexts, where machine learning approaches could support soil moisture monitoring and irrigation management without requiring extensive soil characterisation. UK research institutions have comparable soil and hydrological data infrastructure to the study site, though local validation would be needed to optimise model covariates for British soil and climate conditions.
Key measures
Root zone soil moisture predictions; interpolation and extrapolation accuracy; model performance in extreme wet and dry conditions
Outcomes reported
The study evaluated Random Forest machine learning algorithm performance for estimating root zone soil moisture (RZSM) on a daily timescale over a small agricultural catchment from 2016–2018, comparing predictions against a process-based model with data assimilation.
Topic tags
Dig deeper with Pulse AI.
Pulse AI has read the whole catalogue. Ask about this record, its theme, or how the findings apply to UK farming and policy — every answer cites the underlying studies.