Breakout Summary Report
ARM/ASR User and PI Meeting
10 - 13 June 2019
12 June 2019
10:30 AM - 12:30 PM
29
Dan Lubin, Andrew Vogelmann, Johannes Verlinde, Ann Fridlind, David Bromwich
Breakout Description
The AWARE breakout session will address how atmospheric science over the Antarctic continent fits into the larger picture of understanding global change and improving climate model simulations. This session solicits presentations from both modelers and observationalists from all disciplines within ASR, including radiation measurement, physical meteorology, and aerosol science. Over the past year significant progress has been made with AWARE AMF2 remote-sensing retrievals of cloud properties and in understanding their influence on the radiation budget. AWARE case study data from both the McMurdo and WAIS Divide sites are also being used to examine the performance of global and regional climate models. In addition to covering what the AWARE campaign and related modeling studies have specifically accomplished, the session will address current issues and future directions in Antarctic science and the unique challenges of atmospheric science over the Antarctic continent. We will emphasize the significant contrasts with the high Arctic. The two general objectives are (1) to determine how AWARE data can continue to serve current model development and validation, and (2) to begin strategizing about future collaborative ARM/ASR efforts in Antarctica. We will schedule short presentations from all interested participants, followed by substantial time for discussion.Main Discussion
This session presented the past year’s work and immediate future plans for researchers with major efforts involving AWARE data. AWARE data have been used to assess several major regional and global climate models, and are also being analyzed to reveal new insights into Antarctic aerosol, atmospheric thermodynamics, and cloud microphysics. There were eight presentations from researchers at Scripps Institution of Oceanography (SIO), BNL, Pennsylvania State University (PSU), Ohio State’s Byrd Polar and Climate Research Center (BPCRC), The University of Wyoming, and the University of Cologne. We circulated a sign-up sheet that collected an email list of 29 attendees, most of whom we recognize as being involved with ARM/ASR high-latitude research (e.g., NSA, MARCUS, MOSAIC). Audience participants had various questions and suggestions. particularly regarding interpreting climate model performance against AWARE data. At the end there was some discussion about future ARM/ASR work in Antarctica, potentially at year-round stations on the Antarctic Peninsula that might provide better sites than McMurdo for sampling aerosol sources and aerosol-cloud interactions.Key Findings
Dan Lubin (SIO) gave a brief introduction, mentioning that there are presently 16 publications published or in review based on AWARE data. The most recent is a BAMS article describing the AWARE campaign and giving preliminary climatological analysis and climate model applications. One highlight from this article is an analysis of the annual cycle in aerosol abundance and chemical composition prepared by Lynn Russell (SIO). This work analyzed arctic in situ observations as well as AWARE data. Whereas the arctic seasonal cycle shows highest sulfate and organic mass concentrations associated with springtime haze, the antarctic seasonal cycle shows a summertime maximum and is driven by phytoplankton (sulfate), seabirds (organics) and wind-driven sea spray (salt). Dr. Lubin also described a new NSF-supported field campaign that will deploy at Siple Dome in West Antarctica during December 2019 and January 2020. The goal is to miniaturize and increase transportability of the type of surface energy balance (SEB) equipment that was successfully deployed by AWARE at WAIS Divide, providing more versatility for SEB and remote-sensing missions throughout Antarctica. This campaign is provisionally called Surface Energy and Local Forcing following AWARE (SELF-AWARE). Dr. Lubin also summarized the potential for improved aerosol and cloud microphysical sampling in future antarctic fieldwork using either the US Antarctic Program’s Palmer Station or the British Antarctic Survey’s Rothera Station, and to begin new discussions suggested a provisional campaign name, Antarctic Low Cloud Interaction with Natural Aerosol (ALCINA).
Keith Hines (BPCRC) presented an assessment of how cloud properties and their related surface radiative fluxes are simulated over West Antarctica by the Polar-optimized version of the Weather Research and Forecasting (PWRF) regional model and by the Antarctic Mesoscale Prediction System (AMPS) forecast model. In PWRF the cloud microphysical parameterizations were interchangeable, and four were tested: the single-moment 5-class model, the Morrison 2-moment model, the Thompson-Eidhammer aerosol-aware model, and the Morrison-Milbrandt P3 model. AMPS is only able to use the older single-moment 5-class model. Throughout the January 2016 AWARE surface melt event, AMPS seriously under-predicts cloud liquid water and the resulting anemic cloud radiative effect leaves the surface too cold, even at time of greatest atmospheric moisture. The newer microphysical models give significantly better surface longwave flux, but still have a deficit in cloud liquid water content that during the warmer period can yield differences from the AWARE observations by up to 20 Watts per square meter. Presently it is not clear which of the newer microphysical parameterizations performs the best. This work is in revision with the journal Atmospheric Chemistry and Physics.
Andrew Vogelmann (BNL) presented collaborative work between BNL, PSU, and NASA GISS on the AWARE surface melt event as simulated by two global models: the DOE E3SM and the GISS ModelE. Compared with AWARE data, both models simulate too much net radiation into the snowpack. These biases are directly related to cloud liquid water path (LWP) overestimates (in contrast to the regional models discussed above). For turbulent fluxes, E3SM has a slight positive bias in both sensible and latent heat flux, while ModelE agrees better with observations. For E3SM the radiative and turbulent flux biases nearly cancel each other, producing an apparently reasonable SEB simulation but for the wrong reasons. This work is part of the AWARE BAMS article currently in review.
Xioahong Liu (University of Wyoming) presented further evaluation of the E3SM atmosphere model by evaluating four experiments against a control run using the standard configuration. These experiments are: (1) reducing the Werner-Bergeron-Findeisen (WBF) mechanism rate by a factor of 10; (2) replace the Wang et al. (2014) ice nucleation scheme with a scheme by Meyers et al (1992) that nucleates more ice; (3) replace the Wang et al. (2014) scheme with a newer scheme by DeMott et al. (2015); and (4) partitioning the Cloud Layers Unified By Binormals (CLUBB) cloud condensate into liquid and ice phase based on temperature, similar to Park and Bretherton (2009). These experiments mostly have positive impacts on the LWP bias present in the control run. During the warmest part of the AWARE 2016 surface melt event the CLUBB phase partition simulates the best average surface longwave and shortwave fluxes compared with the observations.
Israel Silber (PSU) presented collaborative work with NASA GISS, ANL and BPCRC demonstrating how simulations of cloud properties and water vapor influence longwave biases in the AMPS forecasting model, the new ERA5 reanalysis model, and the GISS ModelE3 global model. As in the BPCRC work, longwave underestimates are most prominent when liquid water cloud should be present. In Dr. Silber’s analysis of the microphysics, the longwave negative biases are also linked to excess production of cloud ice water. The observed atmosphere is “starved” for ice nuclei and exhibits high supersaturations, while the model physics contain rapid nucleation such that the relative humidity with respect to ice rarely exceeds 100% and the atmosphere is quickly dessicated. Interestingly, Dr. Silber finds that the ModelE3, with nudged horizontal winds, produces better simulations of cloud liquid water probability density functions (PDFs) over both WAIS Divide and in selected McMurdo case studies than either AMPS or ERA5. In particular, ModelE3 can simulate the supersaturation PDF quite well when ice nucleation is reduced consistent with INP consumption. Part of this work is in revision with the Journal of Climate.
Fan Yang (BNL) examined case study data from McMurdo to investigate the change in phase partitioning within mixed-phase stratiform clouds as possibly related to decoupling of the boundary layer. This project considered a cold autumn cloud layer (31 March 2016) that showed a rapid increase in cloud ice water content as the lower troposphere temperature inversion suddenly increased in depth from 0.6 km to 1.2 km, signifying a transition of the boundary layer from a coupled to a decoupled state. The question is: is this transition a cause or a result of the fast change in cloud phase partitioning? Running simulations with the System for Atmospheric Modeling (SAM 6.11.2; Morrison et al. 2009), Dr. Yang found that proactively decoupling the boundary layer has minor impact on cloud ice and liquid water content, while proactively changing the ice number concentration drives the simulated boundary-layer structure closer to the observations.
Damao Zhang (BNL) presented a comparison of AWARE-observed aerosol with the Arctic, in a study similar to that of Lynn Russell in the AWARE BAMS article, but with the high-spectral-resolution lidar (HSRL) and Ka-band ARM Zenith Radar (KAZR) as the main instruments. The arctic observations often show vertically extensive dust layers that are not present over Antarctica. Similar to the in situ aerosol abundance analysis, antarctic aerosol has the largest backscatter signature in summer while arctic aerosol has the largest backscatter signature in spring through early summer. On average, surface CCN is twice as abundant at the North Slope of Alaska (NSA) than at McMurdo, but retrieved liquid droplet number concentration is larger in the antarctic clouds than at NSA.
Frederic Tridon (University of Cologne) demonstrated the potential in the first triple-frequency cloud radar observations ever obtained in Antarctica. The availability of three radar frequencies offers pairs of dual-wavelength reflectivity (DWR) that can be used to simultaneously retrieve cloud particle size and distinguish aggregates from rimed particles. In two summertime case studies from McMurdo, clouds show distinct and alternating signatures of aggregates versus riming at various altitudes and temperature ranges. Comparing AWARE observations with geometrically similar clouds observed during the Biogenic Aerosols – Effects on Clouds and Climate (BAECC) campaign in Finland, the clouds show similar tendencies for aggregate particles at temperatures between 0 and -20 C, but antarctic clouds show a much greater occurrence of rimed particles in the temperature range -10 to -30 C. Dr. Tridon also summarized the ongoing development of triple-frequency radar retrieval methods.