To investigate the correlation between two time series (CO2 versus air temperature and wind speed), we used coherency wavelets and phase angle. The datasets used in this study are in the October 16 of 2018, measured from micrometeorological tower in savanna natural area (2°49’30”S; 54°54’35”W; 82 m a.m.s.l.), Santarém (PA). The correlation between with CO2 concentration and air temperature detected high coherence (upper 0.8). When confronted with the wind speed, in microscale, detection a chaotic and disordered region, displaying very intermittent and showed “islands” (statistically significant), justified by the low amplitude between them. Naturally, the use of coherence wavelet and phase is fundamental to emphasize the between with CO2 and meteorological variables ration. It is evident, even for one day only, anti-phase relation between with the CO2 and air temperature signals. The temporal domain less than 32 minutes (microscale), the wind speed, at some moments, exerts a strong influence on the concentration of the referred gas ( 0.8), mainly from 09 to 15 local time.