Black hole X-ray binaries (BHXB) exhibit rapid variability across different energy bands, reflecting the complex interplay of multiple emission components such as accretion disks and coronas. Traditionally, spectral and temporal analyses have been conducted separately, making it difficult to disentangle the variability and causal relationships among these components. Omama (2024) introduced a state-space modeling framework for spectral-timing data of BHXB, allowing for the decomposition of light curves into distinct emission sources. Applying this method to MAXI J1820+070 in the low-hard state, Omama (2024) suggested a causal propagation of variability from the accretion disk to the Compton cloud, and subsequently to the soft excess.
In this study, we applied the same model to an independent dataset of MAXI J1820+070 obtained on the same day to evaluate its robustness. Using six datasets, we confirmed consistency in power spectra, time lags, and power contributions of each component. We further extended our analysis to five additional epochs during the low-hard state between March and July 2018. The variability propagation pattern remained unchanged, but we observed a decrease in low-frequency power of both the accretion disk and Compton cloud. Notably, the break frequency of the Compton cloud increased over time, whereas the disk did not show such evolution, implying a gradual contraction of the Compton cloud region.