I am working on a project that involves monthly time series climate data for 33 years. I want to use the Mann-Kendall test to examine the trend of the variables. However, I encountered a problem with the correlation test. The data are significantly correlated, which means I need to pre-whiten them before applying the MK test. I used the formula (Yt)=(Xt)-(B)*(Xt-1), where Yt is the pre-whitened value, Xt is the original value at day t, B is the correlation coefficient, and (Xt-1) is the original value at day t-1. However, this resulted in negative values for rainfall and minimum and maximum temperature, which are not realistic for a hot climate arid region. I am not sure if I made a mistake or if there is a better method to pre-whiten the data. I would appreciate any advice on this issue.