Pre-harvest forecast of rice yield for Bhagalpur District in Bihar
Loading...
Files
Date
2017-07
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Department of Statistics, Mathematics and Computer Application, BAU, Sabour
Abstract
In Bihar, rice is the most leading food crop which is grown in 3.151 mha with production of 6.601 mt and productivity 20.95 q/ha during the season 2014-15 [(Directorate of Economics and Statistics, G.O.B. 2015-16)].Reliable and timely forecast of crop production are required for various policy decisions relating to storage, distribution, pricing, marketing, import-export etc. Pre-harvest forecast of rice yield has a great importance in Bihar as there is much more area and production of paddy in this state [Kumar et al. (2013)]. Therefore, present investigation was carried out on “Pre-harvest forecast of rice yield for Bhagalpur district in Bihar” with the objectives, (i) To establish the relationship between crop yield (Y) and biometrical characters as well as farmer’s appraisal (X's), (ii) To test the validity of forecasting model through suitable statistical tools and (iii) Pre-harvest forecasting of rice yield based on biometrical characters as well as farmers' appraisal for Bhagalpur district in Bihar for the year 2016-17.
In this study 50 samples were collected from five blocks of Bhagalpur district in Bihar.Multistage sampling was adopted and at each stage samples were selected randomly.Data collected from these sampling procedures, 1024 regression models were developed. To establish the relationship among Yield (Y), Average Plant Population (X1), Average Plant Height (X2), Average number of effective Tillers (X3), Average Length of Panicle (X4), Nitrogen (X5), Phosphorous (X6), Potash (X7), Number of Irrigations (X8), Pest and Disease Infestation (X9) and Average Plant Condition (X10), a questionnaire based on biometrical character and farmer’s appraisal for rice crop was developed. The variable Y was used as dependent and all other X’s were as independent.Out of these models, ten regression models were highlighted with least RMSE (Root Mean Square Error) value. Further, five regression models were selected for minimum RMSE. Regression analysis was performed for each model. All five models were highly significant. Out of five selected regression models,Model-V i.e.Y ̂=16.74955+1.55753X_3-0.03582X_5+0.10947X_6+1.22290X_8had the minimum Standard Error of Mean Predicted (1.08670) value. Its residuals value was zero and RMSE value was 3.36722. From these analyses it was reflected that Model-V is the best whichwas used for pre-harvest forecast of rice yield.By using this model pre-harvest forecast of rice yieldin Bhagalpur is about 43.80967 (q/ha) for the year 2016-17 based on biometrical characters and farmer’s appraisal.
Description
Keywords
null