To remove technological impact on crop yield, many de-trend methods (linear de-trend, first difference, moving average, etc.) had been recommended. If a de-trended series of yield still show a significant increase trend, does it mean that the selected de-trend methodology is inappropriate? if all de-trend methods had been conducted, and the results still show significant increase in yield, what shall we do to remove the technological effect on yield?
Dear Zhao, if you have the technological level of each measurement I recommend to do not use detrending. Instead, give weights to each measurement according to the technological effect.
This is because many other variables which you want to measure their effects may be parallel to the technological effect, and by detrending their effect will remove too.
In some cases, we use two or three methods of detrending, in which we commonly apply a linear detrend and in next step we apply a smooth splin.
Hope it Helps
Dear Khedive
Many thanks for your professional answer. In my case the technological level can not be measured, and some variables surely may be parrallel to the technological effect. It seems likely that your twice detrending method is very valuable. Please send me a copy of a publication which had been conducted the same detrending method to remove technological effect.
Happy New Year to You!
sincerely, ZZ
Dear Zhao, Thank you, these are two articles on detrending tree rings. Although the fields are different, but, both are time series data. Besides, I think it gives you an extended insight on time series.
Have a good year
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis