Geographical Based Prediction of Carrot Yield Based on Registered Farmers Details

Abstract

Geographical Information System (GIS) is one of the most important computer systems used in variety of industries in Sri Lanka. Carrot is a seasonal crop and it is cultivated monthly wise. Accurate carrot yield predictions can be used to determine the extent of the carrot crop and the next crop of the farmers and time period to change from one crop to another crop. Also, the number of carrot farmers can be determined to fulfill the market necessity. The purpose of this research is to identify the nature of the yield fluctuation of the carrot crop and avoid the existence of the carrot excess in the market. A Statistical package is used to generate the models and predicting the yield under time series analysis. Monthly yield is considered as the variable for analysis. The most highly accurate model is used to fit the model for data set. Each and every district in Sri Lanka is covered by the system. 11 districts were selected and data were collected based on the registered farmers details from 2010 to 2019. SARIMA(0,1,0)(0,1,1)12 was the best model for Badulla, SARIMA(0,1,0)(0,1,1)12 was the best model for Nuwara Eliya, SARIMA(0,1,0)(1,0,0)12 was the best model for Ratnapura, SARIMA(1,1,0)(0,1,1)12 was the best model for Kandy, SARIMA(0,1,1)(0,0,1)12 was the best model for Matale, SARIMA(1,1,1)(0,1,1)12 was the best model for Monaragala, SARIMA(1,1,1)(0,1,1)12 was the best model for Vavuniya, SARIMA(0,1,0)(1,1,0)12 was the best model for Mullaitivu, SARIMA(1,1,0)(2,1,0)12 was the best model for Mannar, SARIMA(1,0,1)(0,1,1)12 was the best model for Kilinochchi and SARIMA(1,1,0)(0,1,1)12 was the best model for Jaffna. Language was used to create an algorithm to determine number of farmers and expected yields for supplying the market necessity.