ANALYSIS OF CLIMATE FACTOR VARIATION USING THE HURST EXPONENT AND DYNAMIC CONDITIONAL CORRELATION METHOD

• Main Author: Herdiansyah Pratama, Muhamad
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/68246
• Institution: Institut Teknologi Bandung

Each entity in a system must have dependencies between one another. This dependence can be described by the tendency of changes in the system in a certain time interval. The purpose of this study was to characterize the nature of climatic factors based on their tendency to change using the Hurst exponential method; determine the dynamic correlation between changes in climate factors, predict the beginning of the dry season, and predict the beginning of the rainy season using the DCC method; and determine the relationship between the Hurst exponential method and the DCC method. The filter used in this study is a filter based on the lowest unobserved system change described by the standard error of the Hurst exponential method and the standard deviation of the DCC method. The probability that a system is affected by a random system is described by the pvalue. Climatic factors that have persistence are rainfall and duration of sunlight. Climatic factors that have anti-persistence properties are maximum wind speed, air temperature, and relative humidity. The system that has an average relationship with a negative direction is relative air humidity with air temperature, while those that have a very low relationship with a positive direction are air temperature with maximum wind speed and relative humidity with rainfall. The rest have a very low relationship with a negative direction. The season start prediction error ranges from 14 days to 179 days. The p-value of the Hurst exponential method has a very strong correlation with the p-value of the DCC method. With this research, it is hoped that it can provide motivation for other researchers to explore further the complexities that occur in nature.