Latent heat and sensible heat flux simulation in maize using artificial neural networks

• Published in: Computers and electronics in agriculture Vol. 154; pp. 155 - 164
• Main Authors: Safa, Babak, Arkebauer, Timothy J., Zhu, Qiuming, Suyker, Andy, Irmak, Suat
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.11.2018; Elsevier BV
• Subjects: Air temperature; Artificial neural networks; Back propagation; Back propagation networks; Computer simulation; Corn; Covariance; Data processing; Deactivation; Earth surface; Eddy covariance; Empirical analysis; Enthalpy; Flux measurement; Heat flux; Heat transfer; Impact analysis; Latent heat; Latent heat and sensible heat exchange; Leaf area index; Multilayers; Neural networks; Rain; Sensitivity analysis; Training; Vapor pressure; Vortices; Wind speed; PAR; Flux measurement; CO2; H; VPD; ANNs; Ta; BP; MLP; P; R; S; V; DTime; Rn; WS; SWC; Ts; R2; Sensitivity analysis; Latent heat and sensible heat exchange; Artificial neural networks; RMSE; NCD; DOY; MAE; Eddy covariance; fPAR; LAI; LE
• ISSN: 0168-1699; 1872-7107
• DOI: 10.1016/j.compag.2018.08.038

•Simulation of the latent heat and sensible heat fluxes above two maize fields.•Development a multi-layer neural network trained by Back-Propagation (BP) algorithm.•Sensitivity of LE and S with respect to each input was analyzed over the growth stages.•The impact intensity of each input on LE and S was recognized by trained neural network. Latent Heat (LE) and Sensible Heat (H) flux are two major components of the energy balance at the earth’s surface, which play important roles in the water cycle and global warming. There are various methods for their estimation or measurement. Eddy covariance is a direct and accurate technique for their measurement. Some limitations lead to prevention of the extensive use of the eddy covariance technique. Therefore, simulation approaches can be utilized for their estimation. ANNs are the information processing systems, which can inspect the empirical data and investigate the relations (hidden rules) among them, and then make the network structure. In this study, multi-layer perceptron neural network trained by the steepest descent Back-Propagation (BP) algorithm was tested to simulate LE and H flux above two maize sites (rain-fed & irrigated) near Mead, Nebraska. Network training and testing was fulfilled using on two different subsets of the hourly data which were selected from days of year (DOY) 169 to 222 for 2001, 2003, 2005, 2007, and 2009. The results showed high correlation between actual and estimated data; the R2 values for LE flux in irrigated and rain-fed sites were 0.9576, and 0.9642; and for H flux 0.8001, and 0.8478, respectively. Furthermore, the RMSE values ranged from 0.0580 to 0.0721 W/m2 for LE flux and from 0.0824 to 0.0863 W/m2 for H flux. In addition, the sensitivity of the fluxes with respect to each input was analyzed over the growth stages so that the most powerful effects among the inputs for LE flux were identified as net radiation, leaf area index, vapor pressure deficit, and wind speed, and for H flux net radiation, wind speed, air temperature, leaf area index and vapor pressure deficit. Furthermore, to achieve the minimal set of input in order to speed up the analysis procedures, the impact intensity of each input on the fluxes was recognized by deactivation of each input vector in network training. This study reveals that artificial neural networks are not only a powerful technique for estimation of LE and H fluxes, but also can identify the effectiveness of each input on the fluxes.