Author(s)

Babak Safa^1*, Timothy J. Arkebauer², Qiuming Zhu³, Andy Suyker⁴, Suat Irmak⁵

¹Daugherty Water for Food Global Institute at the University of Nebraska, Lincoln, USA.
²Department of Agronomy & Horticulture, University of Nebraska-Lincoln, Lincoln, USA.
³Department of Computer Science, University of Nebraska-Omaha, Omaha, USA.
⁴School of Natural Resources, University of Nebraska-Lincoln, Lincoln, USA.
⁵Biological Systems Engineering Department, University of Nebraska-Lincoln, Lincoln, USA.

The eddy covariance technique is an accurate and direct tool to measure the Net Ecosystem Exchange (NEE) of carbon dioxide. However, sometimes conditions are not amenable to measurements using this technique. Thus, different methods have been developed to allow gap-filling and quality assessment of eddy covariance data sets. In this study first, two different Artificial Neural Networks (ANNs) approaches, the Multi-layer Perceptron (MLP) trained by the Back-Propagation (BP) algorithm, and the Radial Basis Function (RBF), were used to fill missing NEE data measured above rain-fed maize at the University of Nebraska-Lincoln Agricultural Research and Development Center near Mead, Nebraska. The gap-filled data were then compared by different statistical indices to gap-filled data obtained with the technique suggested by Suyker and Verma in 2005 [S&V method], and the ANN approach presented by Papale in 2003. The results showed that the RBF network was able to find better fits for missing values compared to the MLP (BP) network and S&V method. In addition, unlike the S&V method, which depends on different gap-filling procedures over the year; the structure of RBF and MLP (BP) networks was constant. However, data analysis indicated Papale’s approach gave better fits than the RBF and MLP (BP) methods. Thus, based on this work, Papale’s approach is the best method to estimate the missing data; though the applied statistical indices, which were used for model evaluation, show little difference between Papale’s approach and the RBF and MLP (BP).

Gap Filling, Net Ecosystem Exchange of Carbon Dioxide, Artificial Neural Networks, Eddy Covariance System

Safa, B. , Arkebauer, T. , Zhu, Q. , Suyker, A. and Irmak, S. (2021) Gap Filling of Net Ecosystem CO₂ Exchange (NEE) above Rain-Fed Maize Using Artificial Neural Networks (ANNs). Journal of Software Engineering and Applications, 14, 150-171. doi: 10.4236/jsea.2021.145010.