Energy Assessment and Modelling for Rice-Wheat Cropping System using Artificial Neural Networks
Loading...
Date
2024-06-03
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Sher-e-Kashmir University of Agricultural Sciences & Technology of Jammu
Abstract
The study was conducted in the subtropics of Jammu provincewith an aim to assess the energy consumption in the rice-wheat cropping system. The data was collected from 384 farmers across three districts (Jammu, Samba, and Kathua) using an interview-based questionnaire. The multistage sampling technique was employed to select representative villagesand convenience sampling was used to select farmers from the purposively chosen districts. The study also aimed to develop an Artificial Neural Network (ANN) to predict the energy consumptionusing variable reduction techniques like correlation analysis and principal component analysis (PCA). By corelation analysis, 104-input variables were reduced to 17 variables for the paddy and 12 variables for the wheat crop. The six selected key variables using PCA for each crop were used to develop various models, including Multiple LinearRegression (MLR), Cobb-Douglas and ANNto predict energy consumption. The effectiveness of each model was evaluated using mean square error, root mean square error, coefficient of determination (R2)and mean absolute error.
The results indicated that the irrigation operation was a primary source of input energy in paddy crop with water being the one of major contributor in all three districts namely Jammu, Samba and Kathua with the total input energy of 31382.45, 28474.90 and 25244.89 MJ/ha respectively. In Jammu district, water (9999.23 MJ/ha) and nitrogen fertilizer (7829.00 MJ/ha) were the major sources of input energy and in Samba district highest contributors were electricity (7938.94 MJ/ha) and water (6739.13 MJ/ha). In Kathua district, the major contributors included diesel fuel (5950.27 MJ/ha), electricity (5420.95 MJ/ha)and water (5409.64 MJ/ha). In wheat crop, fertilizer application, particularly nitrogen-based, was the primary energy-consuming operation and total input energy was found as 19325.33, 17111.37, and 18350.89 MJ/hafor Jammu, Samba, and Kathua district respectively.The energy indices like grain yield, energy efficiency, specific energy, energy productivity, and net energy were also estimated for both crops in each district. The most suitable ANN model for predicting the total energy input for paddy crop production had a topology of 6-6-3-1, with R2 of 0.83, MSE of 11,509,599.06, RMSE of 3,392.58, and MAE of 2,563.34 MJ/ha. For wheat, the optimal ANN model had a topology of 6-6-4-1, with R2 of 0.79, MSE of 4,798,158.82, RMSE of 2,190.47, and MAE of 1,701.08 MJ/ha. The ANN model determined was foundbetter than MLR and Cobb Douglas model both in paddy and wheat crops. These findings contributed valuable insights into the energy consumption patterns and modelling in rice-wheat cropping system in the subtropics of Jammu province.
Description
Keywords
Citation
Zaffar, O.2024. Energy Assessment and Modelling for Rice-Wheat Cropping System Using Artificial Neural Networks