B11: Emissions-forced simulations design document
- Bryce Harrop
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, - competed, 
- in progress, 
- not done
Overview table for the owner and an approver of this feature
1.Description | B11: Capability to run emissions-forced simulations |
| 2.Owner | Bryce Harrop |
| 3.Created | |
| 4.Equ | |
| 5.Ver | |
| 6.Perf | |
| 7.Val | |
| 8.Approver | Katherine Calvin (Unlicensed) |
| 9.Approved Date |
Title: B11: Emissions-forced simulations
Requirements and Design
E3SM Atmosphere Group
Date:
Summary
Code additions necessary to allow reading of surface and aircraft CO2 emissions data files. These codes will allow for emissions-forced simulations to be run. Success of this code means that the CMIP6 data, after appropriate processing, can be read into E3SM on model atmosphere grid and used as a tendency for the CO2 and CO2_FFF tracers.
Requirements
Requirement: Read in surface emissions
Date last modified:
Contributors: Bryce Harrop , Balwinder Singh
Surface fluxes provide the bulk of the historical and projected CO2 emissions. CMIP6 input4MIPs data are available that specify these fluxes at monthly timescales on a 1 degree grid. We require that the annual, global total CO2 emissions are conserved between the input data and the fluxes used by the model. Offline conservative remapping is done to meet this goal.
Requirement: Read in aircraft emissions
Date last modified:
Contributors: Bryce Harrop , Balwinder Singh
Aircraft emissions are a non-negligible contribution to the total CO2 emissions. Like the surface emissions, these are provided by the input4MIPs data catalog. The same requirements for conservation of the surface emissions is applied to the aircraft with emissions with the additional requirement that total column CO2 is identical between the model and input file, while preserving as closely as possible, the vertical distribution of emissions.
Algorithmic Formulations
Design solution: offline conservative regridding; extensive variable regridding from Jöckel (2006) for aircraft emissions
Date last modified:
Contributors: Bryce Harrop
Conserving surface fluxes can be accomplished by regridding the input data with a conservative remapping algorithm. Aircraft emissions are provided on a fixed height grid and must be converted to the model pressure levels at the time the data are read in. The extensive variable regridding method of Jöckel (2006) is used to perform the regridding.
Where G is the remapped variable, F is in the input, A is the grid for F, and B is the grid for G. Essentially, the emissions are divided up onto the model's pressure levels based on how much overlap that level has with the input height coordinate and the file emissions.
Design and Implementation
Implementation: Modify co2_data_flux.F90 and add aircraft emissions module
Date last modified:
Contributors: Bryce Harrop , Balwinder Singh
New code was added to read in data on the atm grid (both SE and FV grids available). For surface emissions, a complete overhaul of co2_data_flux.F90 was done. For aircraft emissions, a new module was added, aircraft_emit.F90. Additional logic for handling the time coordinate was included in input_data_utils.F90.
Planned Verification and Unit Testing
Verification and Unit Testing: Compare model output emissions to input
Date last modified:
Contributors: Bryce Harrop
Surface emissions and aircraft emissions are both output by default when CO2 is active in the model (aircraft emissions are summed in the vertical for the default output, though the full 3D representation is an available history file variable). These emissions can be directly compared to the input file for agreement in spatial and temporal patterns as well as conservation of total emissions. The input data file assumes all months are equally weighted, which is not the case for the model. The time stamps of the input files are modified to counteract this discrepancy, but an error of roughly 0.2% is introduced in the process.
Planned Validation Testing
Validation Testing: short-desciption-of-testing-here
Date last modified:
Contributors: Bryce Harrop
The data are prescribed so validation is met under the same criteria as verficiation.
Planned Performance Testing
Performance Testing: short-desciption-of-testing-here
Date last modified:
Contributors: Bryce Harrop
No performance testing was done during implementation.