Parsimonious models of root zone temperature in soilless substrates through ensemble machine learning

Authors: CROSS, JAMES - The Ohio State University; Owen Jr, James; Shreckhise, Jacob; FIELDS, JEB - University Of Florida; NACKLEY, LLOYD - Oregon State University; Altland, James; DREWRY, DARREN - The Ohio State University

Submitted to: Smart Agricultural Technology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/4/2025
Publication Date: 8/22/2025
Citation: Cross, J.F., Owen Jr, J.S., Shreckhise, J.H., Fields, J.S., Nackley, L., Altland, J.E., Drewry, D.T. 2025. Parsimonious models of root zone temperature in soilless substrates through ensemble machine learning. Smart Agricultural Technology. https://doi.org/10.1016/j.atech.2025.101289.
DOI: https://doi.org/10.1016/j.atech.2025.101289

Interpretive Summary: Outdoor container-grown plants often suffer when their roots overheat, and nursery growers typically lack straightforward warnings of looming heat spikes. To bridge this gap, USDA-ARS scientists and university collaborators have created simple prediction models that draw on data from standard on-site weather stations. Field trials in Tennessee and Ohio demonstrated that just air temperature and sunlight readings can forecast damaging root-zone heat with over 80 percent accuracy up to five hours in advance. These models are the first step in developing user-friendly, low-cost tools that signal when to water before the growing mix gets too hot. Ultimately, this research will help growers protect plant health, conserve water, and reduce crop losses, especially during heat waves.

Technical Abstract: Containerized plant systems using soilless substrates are increasingly vital to global food and horticultural production, offering efficient resource use and expanded crop production in areas where native soils are not suitable. However, open-air production of containerized plants introduces risks for extreme root zone temperatures (RZTs) that impair root function, reduce growth and increase crop loss. While passive mitigation strategies exist, growers often rely on irrigation to cool substrates during high-temperature events. Without reliable real-time monitoring or predictive tools, RZT temperature management remains sub-optimal, ineffective and wasteful. To address this problem we evaluated the potential for predicting substrate temperature using accessible environmental observations and low-complexity modeling approaches. Environmental conditions and substrate temperatures were monitored from June to October 2024 at experimental sites in Tennessee and Ohio. Each site included trials comparing two ground cover types commonly found in open-air crop production, limestone gravel and black landscape fabric. Despite significant differences in cover temperature, only minor differences in substrate temperatures were observed. Ambient air temperature, downwelling shortwave radiation and vapor pressure deficit all demonstrated strong correlations with substrate temperature, with maximum correlations occurring between 2 to 5 hours prior to the substrate temperature observation. These strong lagged correlations motivated the development of relatively simple (one and two predictor) models of substrate temperature using both machine learning (shallow neural networks) and linear statistical models. Models incorporating air temperature and shortwave radiation achieved strong predictive performance for substrate temperatures (R2 = 0.84-0.93) with lead times up to five hours, demonstrating the utility of models developed using widely available weather observations.