A three years trend investigation and spectral behavior of regional climatic temperatures is performed by using fast Fourier transformed signal decomposition technique and spectrogram to the discrete meteorological data. The technique was applied to a custom built weather station data and was set up with the equipment at the premises of Rivers State University, Port Harcourt, Nigeria (4.800085 N, 6.982313 E). The low-frequency components of each years signal show slight changes in the local temperature recorded. Medium frequency aperiodic behavior observed may be attributed to localized factors within the region of study. These findings could help climatologists’ gain good understanding of the short term mechanisms that could cause climate change.