I have a list of thunder flashes, and I'm trying to find if a meteorological variable (CAPE) is a good predictable variable (or proxy) for theses flashes. In my thoughts, I want to evaluate this proxy and find the best threshold (for example 1000 J/kg-1 is maybe the threshold where I have the best score).
My actual methodology is very simple for now: I look at, for every hour of everyday, the value of CAPE variable, and I look if there is a thunder flash at the same time. So in my mind:
- if CAPE value is > threshold and there is a thunder flash, then I have a hit
- if CAPE value is > threshold but no flashes, then false negative
- if CAPE < threshold but thunder flashes then false positive
- if CAPE < threshold but no flashes then true negative
I originally thought I can do it with ROC/AUC (see thread here : Make ROC graph with my problematic) but it seems it's not possible for my case.
What is the best tool I can use to evaluate this proxy as predictable variable ?
