Presumably, someone has already checked this, but... it seems like temperature might be the primary influence. As the temperature increases, the heavier radioactive gases, like radon, as distributed more evenly, and higher up in the atmosphere. Was this tracked?
The signal appears seriously weak relative to the noise...
As an example, radon concentrations are known to fluctuate seasonally, as has been noted in Ref. [10], and it was suggested that the decay of 222 Rn could lead to a seasonally dependent charge distribution on the experimental apparatus. However, this effect is extremely small given the low counting rates that typically arise from radon background [11], and in any case, the PTB data shown in Fig. 3 were corrected for background.
The signal appears seriously weak relative to the noise...