Lady, it is possible that the site is influenced by the buildings in the surrounding areas. First, I suggest making sure that each station is somehow measuring unperturbed flow conditions and that temperature and humidity sensors are not heavily influenced by buildings and the contrasting vegetation/built footprint.

Correlation or regression approaches among the sites, as suggested by other RG members, may work, but the influence of the rather strong diurnal cycle on surface station parameters will obscure a deeper insight of such regression approaches.

I also suggest looking carefully how the sites are located relative to the aspect of the complex terrain (if any).