Precipitation variations in the central Vietnam to forecast using Holt-Winters Seasonal Additive Forecasting method for 1990 to 2019 trend

Abstract

Precipitation played a vital role in the landslide events, water cycle, irrigation management, agriculture yield. The major factor is the extreme weather such as temperature played a vital role in it. This study applied climatological data (30 years of span) of two different geolocation at Vietnam, and applied time series analysis, Holt-Winters Seasonal Additive Forecasting model. The highest Precipitation i.e. 260.83 and 112.96 mm in the month of September and November for Danang and Pleiku, respectively; and total sum of 66550.51 and 34118.38 mm over the period for Danang and Pleiku, respectively. Danang is with higher Precipitation in lesser number of events, and Pleiku is with lower Precipitation though more number of events. Holt-Winters Seasonal Additive Forecasting method revealed the potential forecasting method for the different statistical characteristics data set. Lag plot showed the higher positive as well as the negative lap quartile. Residual error plot present to support the applied model feasibility. Weakness of the study and the future objective of the study drafted.