An empirical and a dynamic-empirical model for the estimation of maize seedbed temperature

Abstract

Soil temperature is the main factor in determining the germination of maize seeds and the emergence time of the crop. It controls the rate of phenological development while the meristem is underground, which is until the V6 (six leaf collar) stage of maize. The research performed by the authors aimed to model maize seedbed temperatures at sowing depth (soil temperature at 5 cm depth) during the sowing-emergence-early development period. The research is based on measurements in ploughed plots of the maize experiments at the Látókép Experiment Site of the University of Debrecen (Eastern Hungary) in two growing seasons of 2021–2022. Two types of empirical models were established, a multilinear regression model (M1) and a dynamic-empirical model (M2), where the daily increase and decrease of soil temperature are determined by multilinear regression. Candidates for input variables for both models were various, easily available daily meteorological parameters. M2 model performed better than M1 when applied to an independent database of 2022. This is particularly valid for the maximum and minimum soil temperatures. It was found that both M1 and M2 can be used to predict the soil temperature of the maize seedbed before shading by the plants. For daily mean temperature, M1 and M2 give a similarly good estimation, while the dynamic-empirical model has to be preferred for the maximum and minimum temperatures. M2, which is based on daily temperature, global radiation and wind speed data, predicts the daily mean (RMSE = 1.4 °C), maximum (RMSE = 2.2 °C), and minimum (RMSE = 1.6 °C) of seedbed temperature not worse than many earlier soil temperature models do, even hybrid or mechanistic ones with a large number of parameters.