Open AccessTime series analysis on precipitation with missing data using stochastic SARIMA
Author(s) -
M. Sharma,
Omer Mohammed,
Sarah Kiani
Publication year - 2021
Publication title -
mausam
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.243
H-Index - 12
ISSN - 0252-9416
DOI - 10.54302/mausam.v71i4.45
Subject(s) - autoregressive integrated moving average , missing data , precipitation , climatology , time series , multiplicative function , series (stratigraphy) , stochastic modelling , meteorology , data series , econometrics , environmental science , statistics , mathematics , geography , geology , mathematical analysis , paleontology
This paper presents an application of the Box-Jenkins methodology for modeling the precipitation in Iran. Linear stochastic model known as multiplicative seasonal ARIMA was used to model the monthly precipitation data for 44 years. Missing data occurred in between for 34 months for some reason. To fill the gap a SARIMA model was fitted based on the first 180 available observations and the missing observations were substituted by the forecasts for the next 34 months. Then a SARIMA model was fitted for the full data. The result showed that the fitted model represent the full data well.