Definition of AOA1 and AOA2 age tracers

[tmooring]

I am attempting to use the archived CESM2(WACCM6) output available at NSF NCAR GDEX Dataset d651045 for a study of stratospheric mixing and have a question about the tracers AOA1 and AOA2. Their names indicate that they are some sort of age of air, and the distributions of these tracers in the stratosphere suggests that they are both implemented as being conserved except at some boundary at which a time-dependent boundary condition is being imposed.

However, the rather odd tropospheric structures of these tracers as well as their actual numerical values mean that the exact nature of the boundary condition and the units of measure of the tracers are not at all obvious. I am a bit surprised that I have not managed to find anything in the literature or elsewhere online about the precise definitions of these tracers—can anyone here help?

Thanks in advance for any assistance you can provide, and please let me know if you have any questions.

Sincerely,

Todd Mooring
Research Associate
Linz Group
Department of Earth and Planetary Sciences
Harvard University

 

[yhanw]

Hello Todd, I run into this exact same confusion today with a different CESM2 dataset. Like you, I couldn't find it explicitly documented.

Just to add some info from my end if useful: In my case, AOA1 values are in a very narrow range (~2.3-2.4). I tried plotting AOA1 lat-lev field (using a 2.3~2.4 colorbar), and the expected physical structure does appear, so likely some sort of conversion is needed? The description says "Age-of_air tracer 1", but unit in netcdf metadata is "kg/kg".

Hope to echo this question, and I'd appreciate any insight from fellow forum users. Thanks!

Yuhan Wang
Postdoc
Earth System Science
Stanford University

 

[Jakub_Sypek]

Hello both,

Did you figure this out? I'm also trying to use the AOA to determine residence times in the mesosphere. If not, I'd also like to echo this question.

Jakub Sypek
Postgraduate Researcher
Department of Aeronautics and Astronautics
University of Southampton

 

[fairyeszhuo]

Hi All,

Did you figure this out? I am facing the same problem thus I'd also like to echo this question.

Best,
Zhihong

 

[mijeong]

In CESM2/WACCM6, AOA1 and AOA2 are two separate synthetic "age-of-air" tracers, which are designed for quantifying transport timescales. They're diagnostic tracers so they don't affect radiation or chemistry. AOA1 is a constant surface flux tracer that has a value of cflx= 1.e-6_r8 (kg/m2/s, flux density). Since there is no chemical sink, its global burden increases roughly linearly with time once the run pasts spin-up. Age or air is diagnosed by comparing a mixing ratio at a grid point to a reference time series from the model’s output and finding the time lag between them. AOA2 is a linearly increasing surface flux tracer that is driven by a flux that grows linearly with time and restricted to Northern Hemisphere land. Since the source itself has a known, prescribed time-dependence, AOA2 is close to an SF6-like tracer. The ages diagnosed from AOA1 and AOA2 should agree closely with each other and with theoretical expectations (e.g., increasing with altitude and latitude in the stratosphere, consistent with the Brewer-Dobson circulation).

To calculate age of air from AOA1, pick a reference location that best represents the well-mixed source region (e.g., tropical tropopause region). χ_ref(t) — time series of AOA1 mixing ratio at the reference location/level. For any grid point x and time t of interest, find the earlier time t' at which the reference timeseries had the same value: χ(x, t) = χ_ref(t'), AGE(x, t) = t - t'.

References
Plumb, R. A., and J. D. Mahlman, 1987: The Zonally Averaged Transport Characteristics of the GFDL General Circulation/Transport Model. J. Atmos. Sci., 44, 298–327, https://doi.org/10.1175/1520-0469(1987)044<0298:TZATCO>2.0.CO;2.

Yudin, Valery A., Sergey P. Smyshlyaev, Marvin A. Geller, Victor L. Dvortsov, 2000: Transport Diagnostics of GCMs and Implications for 2D Chemistry-Transport Model of Troposphere and Stratosphere. J. Atmos. Sci., 57, 673-699. doi: http://dx.doi.org/10.1175/1520-0469(2000)057<0673:TDOGAI>2.0.CO;2.

 

[tmooring]

Dr. Park,

Please accept my belated thanks for your detailed response, and apologies for not figuring out the intended lag calculation technique myself. However due to the tropospheric structure of the two tracers (see attached figures) I have difficulty believing that AOA1 and AOA2 are in fact defined as you describe, at least for these particular simulations. (The stratospheres seem more or less reasonable, but how is it possible for constant surface fluxes without sinks to create a minimum near the surface in the tropics?)

I understand from an email exchange with Rolando Garcia that the unusual tropospheric distribution was noticed about 5 years ago and that the definition of AOA1 was updated to have a prescribed mixing ratio boundary condition which apparently addressed the "problem". Examination of movies of lowest-model-level age data for the d651045 simulations shows ITCZ-like spatial and temporal variability, which makes me wonder whether these age tracers are also somehow being wet-deposited.

The work I am doing on stratospheric mixing does not actually require calibration of AOA1 and AOA2 to physically meaningful ages, but does assume that these two tracers are in fact conserved (at least everywhere in the stratosphere). Hence my interest in understanding what, precisely, the "chemistry" of AOA1 and AOA2 actually is--results obtained using them seem to be a bit outlier-ish relative to those I obtain from other nonconserved but transport-informative tracers such as N2O. On the other hand, this phenomenon might also be an artifact of numerics issues in my diagnostic technique to which the age tracers are uniquely vulnerable (for reasons that need not detain us here).

Thanks,

Todd Mooring

 

[mijeong]

We appreciate you carefully analyzing the AOA1 and AOA2 and asking valuable questions. I had a discussion with Rolando, Doug and Francis who understand the issues with AOA tracers in WACCM well. Below is the summary of our discussion.

1. The AOA tracers in WACCM do not have any sinks and had a flux lower boundary condition (LBC) in WACCM6, which led to a minimum in the tropical troposphere as your figures show. We do not have a clear understanding of why the minimum was formed in the tropical troposphere, but our guess is that stratospheric air is being transported back to the troposphere. The reference point of AOA is at 100 hPa not at the surface.
2. The AOA tracers are meant to be used in diagnosing the age of air above the tropopause and the troposphere is supposed to be well mixed. The flux LBC turned out to be problematic as it creates a minimum in the troposphere. There also has been an issue with the AOA upper boundary condition (UBC), as the tracers were accumulated at the model top. We corrected the LBC to linearly increasing mole fraction LBC and zero flux at the UBC in CESM3-WACCM7.
3. SF6 is the “natural” tracer most similar to AOA1 but the problem with SF6 for age of air calculations is that its growth rate is not linear, which requires adjustments that introduce uncertainties. For N2O, it has more or less constant values in the troposphere but photolysis in the stratosphere makes it non-conservative.

Please reach out to us if you have more questions. Thank you!