@teixeirak FYI, 1999 at Harvard Forest is going to look a bit weird. This was a pretty tough drought year so DOY25 is happening earlier (interestingly, it seems to be impacting DP tree timing more), likely as a consequence of less growth due to drought. This is essentially what Ref 1 was saying about the DOY variables being dependent on one another. I've tried reducing the SD in the DOY timing cleaning category but it doesn't improve much (though, I think we should reduce this to 2 SD which we can discuss in the data cleaning issue).
We already removed 1998 so removing 1999 as well would leave us with 4 years of data for HF. We have two options:
1.) We can leave it in, explain it, and move on.
2.) Remove the year
I think we would be fine leaving it in, but would like your input.
@teixeirak FYI, 1999 at Harvard Forest is going to look a bit weird. This was a pretty tough drought year so DOY25 is happening earlier (interestingly, it seems to be impacting DP tree timing more), likely as a consequence of less growth due to drought. This is essentially what Ref 1 was saying about the DOY variables being dependent on one another. I've tried reducing the SD in the DOY timing cleaning category but it doesn't improve much (though, I think we should reduce this to 2 SD which we can discuss in the data cleaning issue).
We already removed 1998 so removing 1999 as well would leave us with 4 years of data for HF. We have two options:
1.) We can leave it in, explain it, and move on. 2.) Remove the year
I think we would be fine leaving it in, but would like your input.