New figures

[bpbond]

@AditiGit said:

Probably should not average the flux response over the three episodes. The expectation is that response will change. So an immediate response on inundation 1 can be markedly different from an immediate response after inundation 3. Therefore, plot the median of each response (immediate, post-24, post-48 and post-1week in different colors). So there should be 3 data points per response.

This refers to the first image above right?

For each site, imagining the figure and if it is going to get too busy: Have discreet x axis with immediate, post-24, post-48 points, [CH4] as Y1, [CO2] as Y2, and with median of CC and IC treatments plotted. Alternate: Have a facet plot of seven sites for CO2 and another for CH4, with discreet x axis as immediate, post-24, post-48 points, and post-1week, flux on Y axis, and with median of CC and IC treatments for each inundation. Improve (or not?) the plot shared for AGU.

[AditiGit]

Yes, first image. Also realized, plotting two Y axes on R is a pain. So if we need to separate out CH4 and CO2, that is okay. I was trying to plot two Y axes for a graph...and failing miserably. Google says there is no simple solution...unless I am asking the wrong question to troubleshoot?

[bpbond]

Starting with the simpler one first:

Any adjustment you'd like can made.

[AditiGit]

Thanks. Arrange sites by the following order: BC2, BC3, BC4, BC12, BC13, BC14, BC15

[bpbond]

@AditiGit In the Core-InundationDate-Time.csv file , only the inundation-treatment cores have Inundation1, Inundation2, Inundation3 fields filled out. Okay. But to make your first plot above, the control cores need to be matched up with the inundation cores, right? This may be easier to ask looking at data:

Core	Treatment	Harvested	Date_Inund1	Time_Inund1	Date_Inund2	Time_Inund2	Date_Inund3	Time_Inund3
BC2-14	ControlCore	9/6/19	NA	NA
BC2-8	ControlCore	9/6/18	NA	NA
BC2-3	ControlCore	9/6/18	NA	NA
BC2-6	ControlCore	9/6/18	NA	NA
BC2-4	ControlCore	9/6/18	NA	NA
BC2-1	InundatedCore	9/6/19	7/24/18	12:40	8/14/18	10:25	9/4/18	12:35
BC2-12	InundatedCore	9/6/18	7/24/18	12:40	8/14/18	10:25	9/4/18	12:35
BC2-11	InundatedCore	9/6/18	7/24/18	12:40	8/14/18	10:25	9/4/18	10:00
BC2-2	InundatedCore	9/6/18	7/24/18	9:50	8/14/18	10:25	9/4/18	10:00
BC2-9	InundatedCore	9/6/18	7/24/18	9:50	8/14/18	10:25	9/4/18	12:35

Here are the entries for BC2. For the first figure above, is it correct we want to isolate BC2 flux data within X minutes/hours of the Inundation1, Inundation2, and Inundation3 time points? What is X? Is it 24 hours as shown below, or something longer/shorter?

Thanks for clarifying.

[AditiGit]

@bpbond: Okay. Question, how did we decide which single time point to use for each inundation core (the runs were for 1 hour)? For example, for core BC2-1, Inundation 1 on 7/24, the time is 12:40 (the picaro run was from 11.21-12.27). Is the idea to round it off somehow? This way I can figure out what to write for the control cores.

Also, to your question above, yes, let's keep X as 24 hours after the inundations.

[bpbond]

Sorry I don't understand what you're asking.

[bpbond]

@AditiGit - this is getting close to what you want I think?

[AditiGit]

@bpbond yes!

[bpbond]

Do you like the boxplots? I can also make a version with points and error bars if you'd like.

[AditiGit]

@bpbond I like the box plots. Can you also make the points and error bar, just in case. It is intriguing that generally flux is lower as we increase inundation episodes but cumulative flux for BC4 and BC15 is higher. Need to discuss this with you. But I will incorporate the figures, write the discussion, and then share the draft. Thanks.

[bpbond]

@bpbond to do:

1. QC: make series of cumulative C with incubation periods highlighted
2. Methane plots
3. Distribution of 24-hour change numbers
4. CO2:CH4 ratio

[bpbond]

Sorry day got away from me a bit...coming

[bpbond]

@AditiGit is it correct that BC4 and BC14 have multiple-day inundations:

This is present in the inundation data in Core-InundationDate-Time.csv:

BC4-2 InundatedCore 9/6/18 7/24/18 15:40 8/14/18 BC4-14 InundatedCore 9/6/18 7/24/18 15:40 8/14/18 BC4-6 InundatedCore 9/6/18 7/24/18 15:40 8/14/18 BC4-5 InundatedCore 9/6/18 7/24/18 15:40 8/14/18 BC4-4 InundatedCore 9/6/18 7/31/18 15:40 8/21/18 BC14-4 InundatedCore 9/13/18 7/31/18 14:30 8/21/18 BC14-1 InundatedCore 9/13/18 7/31/18 9:20 8/21/18 BC14-3 InundatedCore 9/13/18 7/31/18 9:20 8/21/18 BC14-7 InundatedCore 9/13/18 7/31/18 14:30 8/21/18 BC14-8 InundatedCore 9/6/18 7/31/18 14:30 8/14/18

Are these correct? Or are those last BC4 and BC14 rows flipped, i.e. BC4 was all done on 8/14 and BC14 on 8/21?

[AditiGit]

@bpbond ack. Sorry. BC4 were all done on 8/14 and BC14 were all done on 8/21

[bpbond]

Ah, I wondered. OK, that was causing problems. Thanks!

[bpbond]

I'm assuming that all the BC4-4 and BC14-8 dates should match the others in their site grouping. (The other inundation dates were similarly off.)

[bpbond]

It's tough to visualize CH4 - so bursty, and both positive and negative.

[bpbond]

Figure 1. 24-hour fluxes by site and incubation number

Figure 2. Overall fluxes by site and treatment

Figure 3. Treatment versus median(control) for each incubation + site (i.e. the distribution of 24-hour changes as requested above)

[AditiGit]

Oooh, pretty figures :) Thanks @bpbond. For Figure 3, can we visualize this for each inundation? So there would be three plots like these, one for each inundation episode.

[bpbond]

Like this?

We can also plots as a histogram though I don't think that's very useful.

Glad to see that yes, inundation suppresses CO2 fluxes and increases CH4! A useful sanity check.

[bpbond]

This is pooling all cores together, and just plotting by inundation.

[bpbond]

I was wondering whether there was a trend as you got farther into the experiment--e.g. if subsequent inundations would provoke larger responses. Not really 😞

[AditiGit]

@bpbond

1. Do we need to update figures, after removing the duplicated data files? Let me know if you have done it, otherwise I can. Don't want to duplicate our efforts. :)
2. Did Figure 2 change (cumulative flux by site and treatment)? Can we also have this as a box plot instead of the bars we have right now, and arranged as BC2, BC3, BC4, BC12, BC13, BC14, and BC15?

[bpbond]

Hello Aditi - I'll get on these.

[bpbond]

Whew. Am jammed up with the coastal proposal right now. ASAP!

[AditiGit]

i understand! working on things at my end. so this is progressing.

[bpbond]

@AditiGit With respect to the cumulative fluxes, going back to the question of time, the control and treatment cores have flux data for different amounts of time. Do we want cumulative fluxes at the end of the control times, so everything is the same? I'm a little confused.

[AditiGit]

@bpbond Woah! look at BC3. Okay, first thought, why would there be different amounts of time. Can't seem to figure out why. If we keep this perplexing thought aside for a minute, then yes, cumulative fluxes at the end of the control cores make sense. Still confused though.

[AditiGit]

@bpbond: can you also share the flux numbers for the seawater? There were a handful. I want to check what the response was for just water.

[AditiGit]

@bpbond: Adding things to do here so that we don't lose stuff in the emails.

1. adding the uptake and emission numbers that I need from you.
2. Seawater flux data.

[AditiGit]

@bpbond : Reminder about this.

1. adding the uptake and emission numbers for each gas
2. Seawater flux data.

[bpbond]

I will look at this in next 24 hours and may get back to you with questions. Thanks.

[bpbond]

Well that was a fail on my part. 😞 Going to check.