An improved model to simulate soil water and heat: A case study for drip-irrigated tomato grown in a greenhouse

Soil physical, chemical, and biological processes are influenced by soil water and heat conditions, which greatly depend on the crop root systems in different soil layers. Hence, understanding root distribution, soil water and heat dynamics is essential to improve water and nutrient use efficiency....

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Agricultural water management 2023-03, Vol.277, p.108121, Article 108121
Hauptverfasser: Gong, Xuewen, Li, Xiaoming, Li, Yu, Bo, Guokui, Qiu, Rangjian, Huang, Zongdong, Gao, Shikai, Wang, Shunsheng
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Soil physical, chemical, and biological processes are influenced by soil water and heat conditions, which greatly depend on the crop root systems in different soil layers. Hence, understanding root distribution, soil water and heat dynamics is essential to improve water and nutrient use efficiency. A case study of tomato grown in a greenhouse with drip irrigation was conducted to test an improved two-dimensional (2D) model, namely UZflow-2D, for simulating the soil water and heat dynamics. The performance of UZflow-2D model was also compared with the well-known Hydrus-2D model. A 2D root length density (RLD) function, as a main subsidiary model of UZflow-2D, was proposed by investigating the roots distribution from lateral and radial latitudes at the four main growth stages (seedling, flowering, fruit-setting, and picking). Experimental data in 2015 was used to calibrate the parameters of UZflow-2D and Hydrus-2D models, and validated using the data of 2016. Results showed that more than 70% of the total RLD was concentrated in soil layers of 0–20 cm during the seedling, 0–40 cm during the flowering, and 0–60 cm during the fruit-setting and picking stages. The 2D RLD function perform well for locations within and between two rows, with determination coefficients higher than 0.77. Both UZflow-2D and Hydrus-2D models perform well in simulating soil water and soil temperature dynamics, while UZflow-2D model produced improved accuracy in modeling the soil water and heat dynamics within and between two rows of drip-irrigated tomato plants. The overall root mean square error (RMSE) was ∼0.008 cm3 cm−3 for soil water contents, and ∼0.434 °C for soil temperature for UZflow-2D model. However, the overall RMSE was ∼0.012 cm3 cm−3 for soil water contents, and ∼0.504 °C for soil temperature for Hydrus-2D model. Hence, the UZflow-2D model can be served as an alternatively useful tool to simulate water and heat dynamics. •An improved model, UZflow-2D, was developed for simulating soil water and heat dynamics.•An exponential function can well describe 2D root length density distribution for varying growth stages.•The UZflow-2D model can well capture 2D soil water and heat dynamics.•The UZflow-2D model improved the accuracy compared to the Hydrus-2D model.
ISSN:0378-3774
1873-2283
DOI:10.1016/j.agwat.2022.108121