# Analyzing time-series of flux data and day-of-year identification

1. ## Analyzing time-series of flux data and day-of-year identification

Hi there,

I am currently looking at a time-series of CO2 data, logged at 30-minute intervals over the course of the entire year. The are a total of 17521 points in this dataset spread over 365 days.

I need to find the exact day in which the uptake of CO2 (which is represented by negative graph values), crosses below the red line I have drawn (which is at a value of -4.352), and when it crosses back up again. I have included where these 2 points are in the example spreadsheet to help.

The other alternative is if there is a way of capturing the lower or upper envelopes of this very noisy dataset with a curve? In which case I could subsequently use that curve for analysis.

I am currently taking a pretty visual approach to the problem - zooming in 400% and trying as best as I can to identify when this date is but that is highly unscientific.. I would greatly appreciate any help and advice.

Thank you! =)

2. ## Re: Analyzing time-series of flux data and day-of-year identification

Add a formula ...

 Row\Col A B D E 1 -4.352 2 Day of Year NEE Crossover 3 1 0.3502 4 1 0.3473 0 D4: = (B3D\$1) + (B3>D\$1) * (B4

... and filter
 Row\Col A B D 1 -4.352 2 Day of Year NEE Crossover 5163 108 -4.526 1 5165 108 -1.697 1 5251 110 -4.497 1 5252 110 -3.483 1 5350 112 -5.142 1 5352 112 -3.941 1 5395 113 -4.364 1 5408 113 -4.073 1 5410 113 -5.649 1 5411 113 -2.892 1 5597 117 -4.544 1 5598 117 -4.123 1 5841 122 -5.691 1 5842 122 -3.822 1 5843 122 -4.648 1 5844 122 -3.264 1

3. ## Re: Analyzing time-series of flux data and day-of-year identification

From your graph, looks like that second point is not the first time the value rises above -4.352... but it's the last?

Day 108, step 25, is where the value is first below that value - but two rows down the value is -1.697 - but that's not where your second blue arrow is pointing.

Day 311, step 28, is the last time the value falls below the line - is that what you are trying to identify? Looks like it's up and down across line all the time - but past this point it oscillates above the line only.

Find first one with this function: {MIN(IF(D2:D17521<H1,ROW(D2:D17521),10^6))} - "array function" Confirm with cntrl+shift+enter

Find last dip with this function: {MAX(IF(D2:D17521<-4.352,ROW(D2:D17521),0))} - "array function" Confirm with cntrl+shift+enter

See attached.

4. ## Re: Analyzing time-series of flux data and day-of-year identification

Or, (I noticed the point you found was not the last time, but second to last)

Try this: (you see day 291 is where it was still happening a lot - 311 was an outlier)

example2.png

5. ## Re: Analyzing time-series of flux data and day-of-year identification

I used a couple of helper columns that returned TRUE or FALSE.
First helper. Enter in C2 and fill down:
Formula:
`Please Login or Register  to view this content.`

Second Helper. Enter in D2 and fill down:
Formula:
`Please Login or Register  to view this content.`

Day of the year:

Formula:
`Please Login or Register  to view this content.`

Cell with day of the year:
Formula:
`Please Login or Register  to view this content.`

RESULTS:
 F G 33 Day 112 34 Cell A5351

#### Thread Information

##### Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts

Search Engine Friendly URLs by vBSEO 3.6.0 RC 1