From several early research works, I have noticed the negative correlation between temperature and PM2.5. However, in my recent study using Wavelet Coherence (WTC), I found a positive correlation in high-spectrum regions with a delay of approximately one month. What may be the reason behind this? Should we consider the influence of the other meteorological factors?