Estimación del rendimiento del cultivo de Passiflora Edulis (Maracuyá) a partir de modelos estadísticos

Abstract

Crop planning needs tools to support to making decision to establish its economic and environmental sustainability, as proposed this article presents a statistical analysis for historical data from 2007 to 2014 of the independent agroclimatic variables of the station code 21055020, located in the coordinate latitude 2,378 and length -75,89 municipality of La Plata, Huila, Colombia. Objective estimates a performance forecast using two mathematical models ARIMA and Multiple Regression. These are recommended in the literature. Finally, the results are compared to understand the system and establish their interactions, as support for decision making. The crop selected for the analysis is Passion Fruit (Passiflora Edulis) This crop has an impact economic and social for farmers in the area. The Multiple Regression Model underestimates the peaks of higher performance, besides the adjustment of the model is very low, which implies that this model is descriptive and not predictive while the ARIMA model is recommended given its adjustment to the time series analyzed for this study.