Temporal Analysis Of Temperature Variability In Pakistan Using The Method Of Empirical Mode Decomposition

Abstract

Pakistan is a country with altitudes that ranges from sea level to the world’s second-highest mountain peak in theworld. This distinctive characteristic renders a significant variation in the countries’ climate, for instance, temperature differences and spatial distribution of rainfall. These variations in the corresponding temperature and precipitation are best described and predicted by well-defined classifications of climatic variables and their corresponding spatial distributions across Pakistan. This paper demonstrates the techniques that allow us to predict the average temperature of Pakistan using empirical model decomposing with auto regressive integrated moving average (EMD-ARIMA) and empirical model decomposing with auto regressive integrated moving average with neural networks (EMD-ARIMA–NN) models. This research was applied to a lengthy series of average monthly temperature (report in oC) of Pakistan from January 1901 to December 2016 with 1392 data points. The precision of these models is checked with the avail of statistical analysis and error tests. Comparative analysis shows that the empirical model decomposing with auto regressive integrated moving average with neural networks has a more preponderant forecast