15 Apr 2020 NERSC Directory Change Notice

Due to project's name change at NERSC from 'ACME' to 'E3SM' and NERSC’s file system update, the directory '~~/project/projectdirs/acme/~~' is now '/cfs/cdirs/e3sm'

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

Click here for instructions to fill up the table below ......

The first table in Design Document gives overview of this document, from this info the Design Documents Overview page is automatically created.

In the overview table below 4.Equ means Equations and Algorithms, 5.Ver means Verification, 6.Perf - Performance, 7. Val - Validation

Equations: Document the equations that are being solved and describe algorithms
Verification Plans: Define tests that will be run to show that implementation is correct and robust. Involve unit tests to cover range of inputs as well as benchmarks.
Performance expectations: Explain the expected performance impact from this development
Validation Plans: Document what process-based, stand-alone component, and coupled model runs will be performed, and with what metrics will be used to assess validity

Use the symbols below (copy and paste) to indicate if the section is in progress or done or not started.

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	Improved Dust Aerosol Physics
2.Owner	Yan Feng
3.Created	03 May 2019
4.Equ
5.Ver
6.Perf
7.Val
8.Approver	Chris Golaz
9.Approved Date	27 Jun 2019
V2.0	Pending
V1.0	Declined

Click here for Table of Contents ...

Table of Contents

Title: NA2: Improved Dust Aerosol Physics

Requirements and Design

NGD Atmosphere

Date: 2019-05-03

Summary

Dust aerosols play an important role in Earth's energy balance and atmospheric interactions with biogeochemical cycles. It is critical to quantify dust life cycle and radiative effects in the coupled E3SM. The development and evaluation of dust representation in the V1 model have mainly focused on the global constraint of the dust optical depth (AOD) with observational estimates. However, further analysis of the dust model simulations with V1 suggested that the dust shortwave (SW) absorption is too strong compared with observations and other GCMs. While the direct radiative effect of dust at the top atmosphere is constrained to a large extent by AOD, the heating effect of dust in the atmosphere is overestimated. This could lead to changes of the lower troposphere thermostructure and the subsequent cloud and precipitation changes.

The second issue identified is with the dust size distribution at emission used in the model, which leads to over-predicted fine-mode (<1 um) dust particles in the atmosphere as shown by Kok et al. (2017). As a result, the overall cooling effect by dust at TOA may be over-estimated.

The third issue is related to the increasing of model vertical resolution from 30 vertical layers to 72 layers. It leads to an over-deposition of dust in source regions and less long-range transport. Despite the global AOD is constrained, global dust emissions increase by nearly a factor of 2, dust layer height is low-biased compared to observations, and the lifetime of dust is also short than other global models.

Therefore, the following dust updates proposed for V2 will address the above three issues. First, we will update the dust refractive index in the SW with the AERONET (obs)-derived values from the OPAC data (Hess, 1998) that is used in the default configuration. Second, we will update the fractional dust emission fluxes in accumulation and coarse modes, calculated with the size distribution by Kok (2011). With these updates implemented, dust emissions need to be re-tuned for the globally constrained AOD.

To fundamentally address the over-deposition bias in dust, it requires some surgical change in the code - the aerosol turbulence transport term needs to be moved from CLUBB and add to the aerosol emission/deposition equation. This issue might also need affect other aerosols and water vapor, which probably requires extensive model evaluation. In view of the timeline for the V2, a simple fix is thus proposed. We will implement an alternative option that distribute dust evenly in the PBL, instead of the lowest model layer. As insufficient dust transport is common in large-scale models for various reasons, similar approaches have been adopted in other modeling studies to enhance the dust vertical transport. We propose to implement this simple fix as an option in the namelist so that we will have a short-term solution, while we are working on a more physically-based solution.

Requirements

Requirement: (1) update namelist options in the default compset; (2) update size distribution of dust emission fluxes; (3) update the dust emission distribution in the boundary layer

Date last modified: 2019-06-27
Contributors: Yan Feng

The first two updates have been implemented and tested in V1. The third update is implemented and tests are not completed.

Algorithmic Formulations, and Design Implementation

Design Solution:

(1) Change dust optics (dst_a1, dst_a3, dst_c1,dst_c3) in the namelist &cam_inpam to dust_aeronet_rrtmg_c141106.nc

(2) Change fractional emissions in size bins: dust_model.F90. The new dust emission size distribution is calculated based on Kok (2011). dust_emis_fact is updated for ne30 and ne120 to constrain the global dust optical depth to 0.026~0.03

Date last modified: 2019-06-27
Contributors: Yan Feng

Updated on 2020-05-28 by Yan Feng: updated the dust optics for RRTMGP to dust_aeronet_rrtmg_c141106.nc as well as mam1 and mam3 radiative properties used by RRTMGP

Planned Verification and Unit Testing

N/A

Planned Validation Testing

Validation Testing:

Implemented the new parameters and ran sensitivity simulations with V1. Evaluate the model outputs with observations

Date last modified: 2019-06-27
Contributors: Yan Feng

Table 1 list the model sensitivity studies and configurations. [updated by Yan Feng on 2020-05-28: global scaling factor is estimated from the simulations based on ERA40 reanalyses winds;

dust emissions are retuned for V2 using the model simulated winds, cf. /wiki/spaces/EWCG/pages/1045233888]

	Horizontal resolution	Vertical layers	Physics configuration	Dust emission size distribution	SW optics	Global scaling factor
LRes	ne30	72	FC5AV1C-04P2	Default (Zender et al., 2003)	Default	2.05
LResTN	ne30	72	FC5AV1C-04P2	Kok (2011)	AERONET	0.95
LResTNs	ne30	72	FC5AV1C-H01A	Kok (2011)	AERONET	0.95
HResT	ne120	72	FC5AV1C-H01A	Kok (2011)	AERONET	1.2 (Default is 2.5)

Model outputs

LRes is the control run with V1. Analyzed the last 5 years of the 11-year free model run (by Qi Tang). The model outputs are in:

/global/project/projectdirs/acme/yfeng/AcmeRuns/20161118.beta0.FC5COSP.ne30_ne30.edison/rgr/*ANN*

LResTN is a 4-year run with V1 nudged by the ERA reanalyses (2009-2012) after 1-year spin up. The model outputs are in:

/global/project/projectdirs/acme/yfeng/AcmeRuns/Dustv1_anvil_5d_H_NdgHyb/rgr/*ANN*

LResTNs is a 2-year run with V1 nudged by the ERA reanalyses (2009-2010) after spin up. the model outputs are in:

/global/project/projectdirs/acme/yfeng/AcmeRuns/Dustv1_edison_NdgHyb_dsd_extAN_mamAN_dstemi095/rgr/*ANN*

HResT is a 1-year free model run after spin up. The model outputs are in:

/global/project/projectdirs/acme/yfeng/AcmeRuns/Dustv1_anvil_ne120_H_Hyb_dstemi12/rgr/*ANN*

Evaluation:

(1) Dust aerosol optical depth (AOD) and absorption aerosol optical depth (AAOD) are compared with the AERONET data (INV Version3) averaged between 2006-2015 for sites near the dust sources. LResTN and LResTNs show better agreement with the observed AAOD

Dust AAOD Dust AOD

	AERONET near source regions	LRes (Correlation with AERONET data)	LResTN	LResTNs
AOD	0.311	0.299 (0.91)	0.284 (0.96)	0.282 (0.97)
AAOD	0.017	0.04 (0.69)	0.022(0.72)	0.022(0.74)

(2) Kok et al. (2017) showed that the new size distribution (Kok, 2011) compares better with recent emission flux measurements (Fig 1c in their paper) than other size distributions.

(3) Impact on global dust AOD:

Dust AOD	Total	Accumulation	Coarse
LRes	0.026	0.0093	0.0171
LResTN	0.026	0.0040	0.0214
LResTNs	0.025	0.0040	0.0214
HResT	0.04

Kok, J. F. (2011). Does the size distribution of mineral dust aerosols depend on the wind speed at emission? Atmospheric Chemistry and Physics, 11(19), 10149-10156. <Go to ISI>://WOS:000296357300009

Kok, J. F., Ridley, D. A., Zhou, Q., Miller, R. L., Zhao, C., Heald, C. L., et al. (2017). Smaller desert dust cooling effect estimated from analysis of dust size and abundance. Nature Geoscience, 10(4), 274-278. <Go to ISI>://WOS:000398162800014

Zender, C. S., Bian, H. S., & Newman, D. (2003). Mineral Dust Entrainment and Deposition (DEAD) model: Description and 1990s dust climatology. Journal of Geophysical Research-Atmospheres, 108(D14). <Go to ISI>://WOS:000184611000001

Planned Performance Testing

No changes in performance