The application of Artificial Neural Network (ANN) model to the simulation of denitrification rates in mesocosm-scale wetlands

Abstract Denitrification and its regulating factors are of great importance to aquatic ecosystems, as denitrification is a critical process to nitrogen removal. Additionally, a by-product of denitrification, nitrous oxide, is a much more potent greenhouse gas than carbon dioxide. However, the estimation of denitrification rates is usually clouded with uncertainty, mainly due to high spatial and temporal variations, as well as complex regulating factors within wetlands. This hampers the development of general mechanistic models for denitrification as well, as most previously developed models were empirical or exhibited low predictability with numerous assumptions. In this study, we tested Artificial Neural Network (ANN) as an alternative to classic empirical models for simulating denitrification rates in wetlands. ANN, multiple linear regression (MLR) with two different methods, and simplified mechanistic models were applied to estimate the denitrification rates of 2-year observations in a mesocosm-scale constructed wetland system. MLR and simplified mechanistic models resulted in lower prediction power and higher residuals compared to ANN. Although the stepwise linear regression model estimated similar average values of denitrification rates, it could not capture the fluctuation patterns accurately. In contrast, ANN model achieved a fairly high predictability, with an R 2 of 0.78 for model validation, 0.93 for model calibration (training), and a low root mean square error (RMSE) together with low bias, indicating a high capacity to simulate the dynamics of denitrification. According to a sensitivity analysis of the ANN, non-linear relationships between input variables and denitrification rates were well explained. In addition, we found that water temperature, denitrifying enzyme activity (DEA), and DO accounted for 70% of denitrification rates. Our results suggest that the ANN developed in this study has a greater performance in simulating variations in denitrification rates than multivariate linear regressions or simplified nonlinear mechanistic model.

[1]  E. Prepas,et al.  The application of artificial neural networks to flow and phosphorus dynamics in small streams on the Boreal Plain, with emphasis on the role of wetlands , 2006 .

[2]  Paul G. Risser,et al.  Scales and Global Change. , 1989 .

[3]  Klaus Butterbach-Bahl,et al.  Modeling denitrification in terrestrial and aquatic ecosystems at regional scales. , 2006, Ecological applications : a publication of the Ecological Society of America.

[4]  J. Arah,et al.  Diffusion-Reaction Models of Denitrification in Soil Microsites , 1990 .

[5]  Sovan Lek,et al.  Improved estimation, using neural networks, of the food consumption of fish populations , 1995 .

[6]  Holger Johnsson,et al.  Simulation of field scale denitrification losses from soils under grass ley and barley , 1991, Plant and Soil.

[7]  M. Gevrey,et al.  Review and comparison of methods to study the contribution of variables in artificial neural network models , 2003 .

[8]  Robinson W. Fulweiler,et al.  Modeling denitrification in aquatic sediments , 2009 .

[9]  I. Dimopoulos,et al.  Role of some environmental variables in trout abundance models using neural networks , 1996 .

[10]  I. Dimopoulos,et al.  Application of neural networks to modelling nonlinear relationships in ecology , 1996 .

[11]  Sovan Lek,et al.  Predicting fish distribution in a mesotrophic lake by hydroacoustic survey and artificial neural networks , 1999 .

[12]  Chris Freeman,et al.  Nitrous Oxide Emissions and the Use of Wetlands for Water Quality Amelioration , 1997 .

[13]  Jay F. Martin,et al.  Interaction and spatial distribution of wetland nitrogen processes , 1997 .

[14]  Christoph Müller,et al.  The use of artificial neural networks (ANNs) to simulate N2O emissions from a temperate grassland ecosystem , 2004 .

[15]  Costas Papaconstantinou,et al.  Predicting demersal fish species distributions in the Mediterranean Sea using artificial neural networks , 2003 .

[16]  Bart Muys,et al.  Regionalisation of the parameters of a hydrological model: Comparison of linear regression models with artificial neural nets , 2006 .

[17]  Lars Leonardson,et al.  Seasonal dynamics of denitrification activity in two water meadows , 2004, Hydrobiologia.

[18]  O. Folorunso,et al.  Spatial Variability of Field‐Measured Denitrification Gas Fluxes , 1984 .

[19]  R. Hanley,et al.  Artificial neural network application for multi-ecosystem carbon flux simulation , 2005 .

[20]  Jie Zhang,et al.  A comparison of different methods for combining multiple neural networks models , 2002, Proceedings of the 2002 International Joint Conference on Neural Networks. IJCNN'02 (Cat. No.02CH37290).

[21]  G. David Garson,et al.  Interpreting neural-network connection weights , 1991 .

[22]  Stefan Zerbe,et al.  Productivity of reed (Phragmites australis Trin. ex Steud.) in continental-arid NW China in relation to soil, groundwater, and land-use , 2007 .

[23]  Marco Tomassini,et al.  Soft computing - integrating evolutionary, neural, and fuzzy systems , 2001 .

[24]  Giovanni Coco,et al.  The use of machine learning algorithms to design a generalized simplified denitrification model , 2010 .

[25]  S. Manel,et al.  Comparing discriminant analysis, neural networks and logistic regression for predicting species distributions: a case study with a Himalayan river bird , 1999 .

[26]  Paul W. Leadley,et al.  Global change, nitrification, and denitrification: A review , 2005 .

[27]  Christopher M. Bishop,et al.  Neural networks for pattern recognition , 1995 .

[28]  W. Parton,et al.  Generalized model for N2 and N2O production from nitrification and denitrification , 1996 .

[29]  A. T. C. Goh,et al.  Back-propagation neural networks for modeling complex systems , 1995, Artif. Intell. Eng..

[30]  C. Craft,et al.  Dynamics of nitrogen and phosphorus retention during wetland ecosystem succession , 1996, Wetlands Ecology and Management.

[31]  Aloyce W. Mayo,et al.  Nitrogen transformation in horizontal subsurface flow constructed wetlands I: Model development , 2005 .

[32]  H. Brix.  Do macrophytes play a role in constructed treatment wetlands , 1997 .

[33]  Elizabeth Pattey,et al.  Mathematical modeling of nitrous oxide emissions from an agricultural field during spring thaw , 1999 .

[34]  George Tchobanoglous,et al.  Wastewater Engineering Treatment Disposal Reuse , 1972 .

[35]  Sovan Lek,et al.  Artificial neural networks as a tool in ecological modelling, an introduction , 1999 .

[36]  William J. Mitsch,et al.  Wetlands and Lakes as Nitrogen Traps: Kessler, E. and M. Jansson, eds. 1994. Special Issue of Ambio 23:319–386. Royal Swedish Academy of Sciences, Stockholm. , 1995 .

[37]  J. Meisinger,et al.  Influence of Sample Size on Measurement of Soil Denitrification1 , 1987 .

[38]  David S. Schimel,et al.  Scale and the measurement of nitrogen-gas fluxes from terrestrial ecosystems. , 1988 .

[39]  Arthur J. Gold,et al.  Nitrate dynamics in riparian forests : microbial studies , 1992 .

[40]  Keunyea Song,et al.  Denitrification rates and community structure of denitrifying bacteria in newly constructed wetland , 2011 .

[41]  William J. Jewell,et al.  Predicting nitrogen and phosphorus removal in wetlands due to detritus accumulation: A simple mechanistic model , 2006 .

[42]  W. Schlesinger.  On the fate of anthropogenic nitrogen , 2009, Proceedings of the National Academy of Sciences.

[43]  A. Shamseldin.  Application of a neural network technique to rainfall-runoff modelling , 1997 .

[44]  S. Lek,et al.  Applications of artificial neural networks for patterning and predicting aquatic insect species richness in running waters , 2003 .

[45]  K. A. Smith,et al.  A MODEL OF THE EXTENT OF ANAEROBIC ZONES IN AGGREGATED SOILS, AND ITS POTENTIAL APPLICATION TO ESTIMATES OF DENITRIFICATION1 , 1980 .

[46]  Chang Shu,et al.  Artificial neural network ensembles and their application in pooled flood frequency analysis , 2004 .

[47]  M. R. Carter,et al.  Soil Sampling and Methods of Analysis , 1993 .

[48]  Véronique Beaujouan,et al.  A nitrogen model for European catchments: INCA, new model structure and equations , 2002 .

[49]  Aloyce W. Mayo,et al.  Modelling nitrogen removal in a coupled HRP and unplanted horizontal flow subsurface gravel bed constructed wetland , 2005 .

[50]  G. Pinay,et al.  Spatial and Temporal Patterns of Denitrification in a Riparian Forest , 1993 .

[51]  B. J. Bailey,et al.  Neural Network Models of the Greenhouse Climate , 1994 .