A1 Ozone Hole: Design Document
The Design Document page provides a description of the algorithms, implementation and planned testing including unit, verification, validation and performance testing. Please read Step 1.3 Performance Expectations that explains feature documentation requirements from the performance group point of view.
Design Document
In the table below, 4.Equ means Equations and Algorithms, 5.Ver means Verification, 6.Perf - Performance, 7. Val - Validation, 
- completed, 
- in progress, 
- not done
Overview table for the owner and an approver of this feature
1.Description | Ozone-hole-Linoz-v2: Interactive stratospheric ozone and ozone hole |
| 2.Owner | Philip Cameron-Smith |
| 3.Created | |
| 4.Equ | |
| 5.Ver | |
| 6.Perf | |
| 7.Val | |
| 8.Approver | Shaocheng Xie; (Phil Rasch (pnl.gov) |
| 9.Approved Date | |
| V1.0 | Accepted |
Title: cameronsmith1/atm/Ozone-hole-Linoz-v2: Interactive stratospheric ozone and ozone hole
Requirements and Design
ACME Atmosphere Group
Date: 2015-9-4
Summary
The Antarctic ozone hole changes the solar heating in the atmosphere, and it is fairly well established that this impacts the Southern Annular Mode winds (SAM) and hence surface stress on the ocean and thereby ocean currents that may have implications for sea-ice and ice-sheet melting (ACME v1 cryosphere science driver).
The interannual variability of the Antarctic ozone hole is quite large, so this variability may impact the probability of warm water reaching the base of the ice-sheets and triggering rapid melting.
To implement an Antarctic ozone hole that is consistent with the ACME model state and provides variability in the future requires an interactive stratospheric ozone capability. Specifying stratospheric ozone with monthly-mean data files (the current ACME default) will not provide ozone that is consistent with the model state, and it is not clear how to provide the variability in the future.
Many stratospheric ozone and ozone-hole capabilities have been developed, but most of them are far too computationally expensive to even consider for the high-resolution ACME v1 cryosphere simulations. However, there is one scheme (Linoz) which is well tested and computationally fast: it only requires the computational cost of a single advected tracer.
Linoz was developed by Michael Prather's group at UC Irvine, and Linoz-v2 was implemented into CAM in 2008 for the LLNL_super_fast chemical mechanism by Philip Cameron-Smith and Daniel Bergmann, and subsequently incorporated into the NCAR trop_mozart mechanism too. Hence, Linoz was used extensively as part of the Atmospheric Chemistry Climate Model Intercomparison Project (ACCMIP).
The goal of this task is to implement Linoz for ACME-v1, and tune it to reproduce the statistics of the historical Antarctic ozone hole.
This task is covered by the ACME plan for atmospheric chemistry: /wiki/spaces/ATM/pages/2523491.
Requirements
Requirement: Combine Linoz-v2 and MAM3 in a new chemical mechanism.
Date last modified: Contributors: Philip Cameron-Smith (add your name to this list if it does not appear)
Create a new chemistry mechanism that adds Linoz-v2 to MAM3.
Requirement: Implement Linoz-v2 into the default MAM4 chemistry configuration.
Date last modified: Contributors: Philip Cameron-Smith (add your name to this list if it does not appear)
Create a new chemistry mechanism that adds Linoz-v2 to MAM4. Make this the default for ACME v1 simulations.
Requirement: name-of-requirement-here
Date last modified: Contributors: Philip Cameron-Smith (add your name to this list if it does not appear)
Tune simulation to reproduce statistics of Antarctic ozone hole and stratospheric ozone hole (trends and variability).
Algorithmic Formulations
Design solution: short-description-of-proposed-solution-here
Date last modified:
Contributors: Philip Cameron-Smith (add your name to this list if it does not appear)
Linoz-v2 is documented in Hsu and Prather, JGR, 2009 <http://onlinelibrary.wiley.com/doi/10.1029/2008JD010942/abstract>.
Linoz-v1 is documented in McLinden et al., JGR 2000 <http://onlinelibrary.wiley.com/doi/10.1029/2000JD900124/abstract>
The key equation (from Hsu and Prather) is:
f = ozone concentration.
T = local temperature.
c = overhead ozone column.
(P-L) = Net chemical tendency due to production minus loss.
o = Climatological tendency or sensitivity at the model climatological equilibrium point (for the UC-Irvine CTM).
Note: Only a single tracer is needed (f), since c is calculated by integrating f above each grid-point. Hence, errors in ozone concentrations can be caused by incorrect atmospheric circulation.
The equilibrium ozone concentration will not be in chemical equilibrium, because transport provides an additional tendency to f.
The ozone hole is implemented using the polar stratospheric cloud (PSC) parameterization of Cariolle et al. [1990]. When the temperature falls below 195 K and the sun is above the horizon at stratospheric altitudes, the ozone loss scales as the squared stratospheric chlorine loading (normalized by the 1980 level threshold), which will quickly destroy ozone under typical ozone hole conditions.
Documentation for the ACCMIP project, in which Linoz was used for the LLNL-super-fast simulations, can be found among the papers listed under /wiki/spaces/ATM/pages/2523491.
Design and Implementation
Implementation: Create new chemistry-aerosol configuration.
Date last modified:
Contributors: Philip Cameron-Smith (add your name to this list if it does not appear)
Planned Verification and Unit Testing
Verification and Unit Testing: Algorithms and code-base are already well established, so verification and unit-testing are not needed.
Date last modified:
Contributors: Philip Cameron-Smith (add your name to this list if it does not appear)
Linoz has been published (see above), and the code has been used by multiple modeling groups for several years, including at LLNL and NCAR, and was accepted by the CESM chemistry-climate working group and been released as part of CESM. Hence the algorithms and code-base are already well established. Thus, the risk is deemed to be minimal, so verification and unit-testing are not deemed to be worth the time investment for v1. However, a unit-testing capability with inbuilt verification would be desirable for the future.
Planned Validation Testing
Validation Testing: Validate against historical satellite record of stratospheric ozone.
Date last modified:
Contributors: Philip Cameron-Smith (add your name to this list if it does not appear)
The ozone simulated by ACME with Linoz will be validated by comparing the simulated ozone field against historical satellite observations, including
- atmospheric ozone columns (monthly-mean maps),
- zonal-mean ozone distribution (monthly-means),
- Area of Antarctic ozone hole (historical trends and interannual variability),
- Degree of ozone depletion (historical trends and interannual variability).
The criteria for matching is if the difference compared with observations is comparable (or better) than the ACCMIP simulations with the LLNL-superfast mechanism.
Planned Performance Testing
Performance Testing: Determine performance by running ACME with Linoz turned on and off.
Date last modified:
Contributors: Philip Cameron-Smith (add your name to this list if it does not appear)
We will test the performance of the Linoz capability by running ACME with and without Linoz. We expect the timing to be equal to the time required to advect one extra tracer.