Yes, with the spikes it does make sense because "null" is interpreted as 0, so when you have say 25 and then suddenly no value, it would interpret that as 25 deg difference.
But the pressure outlier doesnt make sense, because outlier is defined in case of pressure by exceeding the max, so if no data is available, it checks your defined max pressure with 0 and that 0 should not be greater than or equal to the max P limit for outliers