Combined application of Life Cycle Assessment and Adaptive Neuro-Fuzzy Inference System for modeling energy and environmental emissions of oilseed production

In this study energy and economic analyses and environmental Life Cycle Assessment (LCA) of canola production in Mazandaran province of Iran were conducted and then an intelligent system of three level Adaptive Neuro-Fuzzy Inference System (ANFIS) was implemented to predict three mentioned indices based on energy consumption from different inputs. The functional unit was considered to be one hectare of canola production. Energy use efficiency and energy productivity were found to be 3.73 and 0.14 kg MJ−1, respectively. The LCA results indicated that total emissions of canola production was 2488.72 pPt ha−1, from which off-farm emissions and on-farm emissions contributed as 1780.43 and 708.29 pPt ha−1, respectively. Emissions due to production and application of chemical fertilizers, especially nitrogen, had the pivotal role on environmental burdens. Coefficients of determination for predicting output energy, benefit to cost ratio and environmental emissions final score (EEFS) were estimated to be 0.90, 0.87 and 0.92, respectively. It is concluded that chemical fertilizer is one of the main energy consuming inputs and emission sources, in particular, for impact categories of global warming, acidification and eutrophication. Optimization of fertilizer application in canola production in the region is generally beneficial from energy, economic and environmental points of view. It is proposed that implementation of multi-level ANFIS is a useful tool in helping to predict the energy, economic and environmental indices of agricultural production systems.