This research provides a comprehensive and statistically reliable analysis of CO2 allowance prices. Also, it aims to identify the diverse behaviors of CO2 allowance prices in different conditions (under various timescales and without timescales, during the crisis period, and before it), utilize different financial variables, and suggest a novel filtering method. This research investigates the drivers of CO2 prices in the European Union emission trading system by applying the multivariate empirical mode decomposition and fast Fourier transform techniques to decompose each time series into three categories and in diverse time scales (short, medium, and long). Also, a novel filtering procedure has been implemented using different statistical methods (principal component analysis, cross-correlation, and Hsiao causality test) to achieve reliable results in final models. The selected data are daily CO2 prices, and its coverage is between January 01, 2013 to 06/30/2022. The considered variables include the stock index, crude oil, coal, gas, selected exchange rates, and some new variables (Euro index, utility index) for CO2 price modeling under the COVID-19 crisis. The results show that new variables used in the research were found to be influential variables for better carbon price modeling. The findings of this research can be helpful for policymakers and investors, particularly when the market is facing an unprecedented crisis like the COVID-19 pandemic.