Analysis of 396 and 397 historicals

[islasimpson]

Here's a place to discuss any issues or differences we see in the 396 and 397 historical simulations

[islasimpson]

Global mean temperature: In the project meeting we had some concerns about there being a difference in the response to Krakatoa compared to our previous historical simulations. Here is a plot of global mean temperature compared, with a spatial masking in the same way as observations comparing LENS2 with our previous iterations of historicals (315, 316, 330_branch) and our latest historicals (396 and 397). The global temperature response to Krakatoa does appear to be muted compared to what we had previously seen (red compared to blue), but it's now more aligned with LENS2 and perhaps with observations.

Looking at the globally averaged surface downwelling shortwave (FSDS) there is a bit less of a decline but it's not totally convincing that's the cause as 315 is pretty similar to these new runs.

Here's a plot of the globall averaged AODVISst from the old and new forcing files

It's not obvious that the imposed Krakatoa forcing is really different. Our background volcanic forcing is, however, a bit higher. @tilmes - is it expected?

[islasimpson]

Actually, I just noticed there's an extra volcano in the late 50's. I suppose having quite a bit more volcanic forcing in 1850/1860 could make the Krakatoa anomalies look less dramatic

[islasimpson]

By 50's, I mean 1850's

[islasimpson]

Attaching here a presentation from Ilaria, via @tilmes on volcanic forcings. I think a relevant figure is this one where the "CMIP7 ne30" forcing is different than "CMIP7", aligned with what we see above. It would be good to hear from the experts that they are comfortable with a rise in background piControl stratospheric AOD that looks to be about 50%

CESM_WG_2026 (1).pptx

[islasimpson]

@ilaria-quaglia the relevant question is here. Do you know why CMIP7_ne30 has such a higher background volcanic emissions compared to CMIP7 and is it OK to have this large of a difference?

[islasimpson]

While looking at this, I tried to look at the AODVIS variable, but it seems what is included in there may have changed. It looks like perhaps this previously did not include stratospheric AOD but now it does. I doubt these are real differences. I just want to make sure everyone is aware and it is intended...

[tilmes]

Yes, we included the prescribe volcanoes to show up in AODVIS diagnostics now.

[islasimpson]

I also took a look at the AMV to see if there's any evidence for surface impacts of the difference in AMOC variability that WIlliam showed today (I don't know his github handle - perhaps someone else can tag him who knows). The AMV is the difference in SST between a box over the North Atlantic and the global average, following Trenberth and Shea (2006). I'm only masking out land. I'm not doing anything fancy to deal with sea ice and I'm using TS so, as you will see, this isn't purely a measure of ocean variability and can be impacted by the lab sea freeze.

Here's the AMV, 10 year running mean on the left and then detrended versions of 396 and 397 on the right

You can see a wild trend in 397 but that's nothing to panic about. It's what we already know, that the lab sea is freezing. Maps of the temperature trends from 1850 to 1910 are below where you can see the big cooling over the lab sea in 397. But if we detrend, we can compare the other low frequency variability between 396 and 397 in the right panel on the plot above. I think you can see that the oscillations in 396 kind of align with the AMOC variability in William's plot (pasted below). While the amplitude of this variability is a bit larger in 396 than in 397, they're not drastically different, so I'm a little surprised that the variability in AMOC is so different. If someone wants to send me the AMOC index, I could relate these two things more carefully.

Maps of TS trends (land masked out):

William's plot:

[cecilehannay]

I am tagging @ilaria-quaglia about the volcoano questions