Abstract

VILLAZON-BUSTILLOS, Daniel et al. Time series analysis to forecast drought in the northwest side of Chihuahua, Mexico. Ecosistemas y recur. agropecuarios [online]. 2016, vol.3, n.9, pp.307-315. ISSN 2007-901X.

In 2011 it was presented one of the most devastating droughts in northern Mexico that seriously affected the activities of the productive sector. The aim of this study was to predict the next event of drought in the northwest region of the state of Chihuahua, Mexico. Information of 40 years was used in four weather stations. The ARIMA models with nine years seasonality were used, and artificial neural networks with 50 neurons and 24 delays were used. Precipitation from 2000 to 2012 was simulated and predicted from 2013 to 2024. For the Abraham Gonzalez station is forecasted for 2019 and 2022 less than 200 mm precipitation will occur. While in the Bachíniva station is forecasted for 2019 about 140 mm precipitation. While at the station Cuauhtemoc is forecasted for 2020 a rainfall of about 200 mm. While in Namiquipa station, a precipitation of around 160 mm is forecasted for 2018. It concludes that between 2018 and 2019, a between moderately and severe drought will occur in the northwestern region of the state of Chihuahua.

Keywords : Drought; ARIMA models; NARX neural networks; forecast.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )