Breakout Summary Report

ARM/ASR User and PI Meeting

2 - 6 May 2016

Session Title: CAPI low warm clouds
Session Date: 5 May 2016
Session Time: 8:30 AM - 10:30 AM
Number of Attendees: 0
Summary Authors: Steve Ghan, Graham Feingold, Cheng Zhou, Jim Hudson, Yangang Liu, Jiwen Fan

Breakout Description

A breakout session to discuss aerosol effects on warm clouds and the implications for climate model paramaterization.

Main Discussion

Ghan discussion: Relationships driving aerosol effects on warm clouds estimated from PD variations are apparently not a good proxy for PI to PD change. Why? Meteorology -aerosol covariations? Some metrics (SPOP vs LWP) have better correlations (Wang et al GRL 2012) constraining forcing with recent regional changes. Example: global brightening (1990-2005) Cherian et al GRL 2014. Sort by cloud regimes, e.g., Zhang et al ACP 2016, cloud feedbacks in different regimes. S. Schwartz: concern over “uncertain then irrelevant now” question; factors undiscovered and put into models.

G. Feingold discussion: Approaches to quantifying ACI (Feingold PNAS2916). Bottom-up: large error and error propagation; uncertainties compound; Lebo and Feingold ACP 2014) top-down (albedo-cloud fraction relationship Engstrom et al 2015, relationship between rCRE, cloud fraction and cloud albedo. Use two sets of LES simulations demonstrating differences between the two approaches. Detectable but small aerosol effect in scene albedo-f in set 1 but not set 2. Highlight the need and usefulness for quantifying macroscopic relationships in top-down and link to microphysical relationships in bottom-up.

C. Zhou discussion: Compare SCAM5.3 and GCE CRM driven by same forcing for a sc case during MC3E. SCM has wide range of LWP but GCE stable when CCN changes >> CAM little changes in evaporation unlike GCE where evaporation and autoconversion changes balanced. GCE evaporation occurs mainly near cloud top. Coarse res GCE (50 m vs 100km) gave LWP vs Na closer to SCAM due to reduced entrainment. Liu: Suggest doing more resolution sensitivity study with GCE and test with more autoconversion schemes.

Jim Hudson discussion: two experiemts: MASE (Sc) vs ICE-T (cu). Hudson et al JGR2015 s from CCN Hoppel minima. Rank CCN bimodality from 1 to 8 (unimodal). Opposite relationships Nc-modal ranking (cu positive vs sc negative). Cu primarily collection-processing vs sc chemical and Brownian processing. Parcel model show different sensitivities to w for mono-and bi-modal spectra over different diameter ranges. CCN modality (associated with cloud processing) enhances both AIE in sc but reduces in cu. Ghan: averaging scale; answer (~km) worry more about the gaseous precursors. More CCN in sc are cloud-processed >> cloud-processing tend to reduce AIE, buffering factor; clouds make the best CCN.

Y. Liu discussion. Going beyond traditional ACI regime paradigm based on adiabatic and cloud droplet concentration dependence on aerosol properties and updraft velocity to consider dispersion and entrainment-mixing in ACI parameterization. Graham: Need to consider collection process; Steve G. asked about the effect of other aerosol properties like aerosol distribution width on the results. On entrainment, Steve G. pointed out relevance of Barahona and Nene’s work on entrainment parameterization. Liu is extending to consider dispersion and coupling. Steve S. on potential use of measurements of reduced light scattering in clouds.

J. Fan discussion: coupling spectral bin microphysics in WRF-Chem (SBM-MOSAIC) to study aci and ari (aerosol-radiation interaction) changes: 1) activation: map 4 MOSAIC bin to 33 CCN bins in SBM; 2) aerosol-resuspension to interstitial aerosol bins; 3) in-cloud wet removal. Test VOCAL sc cases. Examine the effects of the major changes and demonstrate their importance. The fully coupled model improves simulations.

Key Findings

Large diversity in all factors driving anthropogenic aerosol effects on cloud radiative forcing.
Spatial and temporal variability in clouds, aerosol, and cloud radiative forcing are not necessarily useful in constraining factors.
Co-variability of aerosols and clouds with meteorology continue to challenge isolation of aerosol influence on clouds.
LASSO could provide an effective framework for simulating the co-variability.
Improved understanding of dependence of droplet dispersion on aerosol and updraft velocity.
Aerosol processing by shallow cumulus clouds is primarily by collision and coalescence, while processing by stratiform clouds is primarily by aqueous chemistry.
Cloud-resolving models are useful for addressing dependence of interaction on resolution, but results depend on treatment of cloud microphysics.
Bin cloud microphysics has now been fully coupled with bin aerosol microphysics in WRF-Chem.

Issues

Studies of aerosol effects from surface measurements are more difficult when the cloud is decoupled from the surface.
Spatial and temporal variability in clouds, aerosol, and cloud radiative forcing are not necessarily useful in constraining factors.

Needs

CCN at Barrow.

Decisions

Apply same microphysics treatment in cloud-resolving models and global models.

Future Plans

Extend LASSO to simulate cloud-aerosol interactions.

Action Items

Further analysis of why spatial and temporal variability in clouds, aerosol, and cloud radiative forcing are not necessarily useful in constraining factors.
Add dispersion effect to global models.
Use autoconversion representation in models that have a weaker dependence on droplet number concentration.