P=9T^2 + 296T + 645

where P=precipitation (mm a-1 ) , T=5.68 +/- 1.5°C

My thought process was to calculate the error for '296T' using the multiplication by constant rule. Then, calculate the error for 9T^2 using the power rule. Finally, I would use the sum rule to combine these errors. However, this article states that "the errors being propagated must be uncorrelated". So now I am totally confused on how to get the error for this equation.

To add more complexity, there is a standard error on the above equation of +/- 200mm a-1 and I am unsure on how to factor this in.

Can anyone point me in the right direction on how I should treat this error analysis? I would be so grateful!