Use can use Pearson's correlation method for finding if there is any relationship present is the two datasets.

Moreover, you can also go for the ANOVA test, which is used in this present study to determine whether there is any statistically significant relationship between the two variables.

You can refer to the following publication

Article Spatial variation of surface urban heat island magnitude alo...

Responses of NDVI to both T and P will likely/potentially be nonlinear. I suggest to use a scatterplot first, then a correlation as suggested before. However, I would likely use Spearman rather than Pearson (dep. on scatterplot).

Moreover, there will be interactions of T and P that can be tested (g-test).

Finally, I'd suggest to use Generalized Boosted Machines :

NDVI ~ T * P

You may also have to add quadratic terms to account for nonlinearity.

Cheers,

R

Rainer Ferdinand Wunderlich thank you. When plotting on a scatter graph, which should be my independent (x axis) and dependent (y axis) variables? I am assuming that NDVI should be the dependent variable and temperature/precipitation/greenhouse gases should be the independent variables?

The orientation of the scatterplot is basically irrelevant as it is just a graphic representation of your data allowing visual pattern analysis.

(

Please note, that this is not the case for evaluation (R², etc.):

https://www.sciencedirect.com/science/article/abs/pii/S0304380008002305

)

Cheers,

R

NDVI results will be categorical, whereas the other factors, such as temperature and precipitation, are continuous. So you would use tests such as point biserial correlation, logistic regression, and Kruskal Wallis H Test :)