The atmospheric pollution alerts to the need to promote the
development of air quality monitoring systems. Equipment with
low implementation cost and energy consumption are highlighted
for highly scalable applications. The datasets generated by the
sensors feed machine learning algorithms for the weather
conditions analysis in urban centers. However, sensor technology
used in these systems tends to be less accurate than complex
equipment, resulting in a strong correlation between the collected
data and the climatic elements variation. This paper studies the
influence of temperature and relative humidity in low cost air
quality monitoring systems. This knowledge can generate new
approaches to the monitored metrics conditioning and increase
the robustness of the low cost systems.