It is always better to have a parsimonious model, having only 5 to 6 important variable which have significant influence on yield (based on the physiology of a particular fruit). Further, for this type of study, most desirable thing is that one should have long term data, at least of 50 years.

Hi Bhim,

Maybe you can go faster when asking yourself the question which climatic variables are absolutely NOT involved in fruit production and climatic impact on it.

That question is much more closed and difficult than the one you put forward. Difficult questions give more progress when solved.

Just a suggestion. Proof can also be provided by asking yourself whicg factors are NOT involved in certain processes.

Cheers and good luck on the hunt for non-impacting factors on fruit production.

Frank

You may look the issue through data driven approach by the use of data reduction technique (Factor Analysis) or theory driven as per expert view, or a hybrid one.