Breakout Summary Report
ARM/ASR User and PI Meeting
10 - 13 June 2019
How ARM Meets the Needs of ASR Science Goals? (Panel Discussion)
13 June 2019
10:30 AM - 12:30 PM
80
Shaocheng Xie, Jim Mather, and Jennifer Comstock
13 June 2019
10:30 AM - 12:30 PM
80
Shaocheng Xie, Jim Mather, and Jennifer Comstock
Breakout Description
The goal of this panel discussion session is to facilitate two-way feedbacks between the ARM infrastructure team and ASR science team to improve the understanding of priorities on both ARM data developments and ASR working group research activities and address the critical ARM data needs of the ASR working groups. The panel discussion will include a few 5-minute talks to highlight major ASR working group activities and their urgent data needs, as well as major ARM data developments and their potential use in support of ASR science. The majority of time will be dedicated to panel discussion. The panel members consist of representatives from both ARM and ASR.
Panel Members: Sebastien Biraud, Jennifer Comstock, Ann Fridlind, Mike Jensen, Steve Klein, Allison McComisky, Andy Vogelmann
Main Discussion
This is the first time that ARM/ASR has organized a breakout session in a “Panel Discussion” format to solicit input on how ARM meets the needs of ASR science goals. The panel members represented the ARM science data product team, the UEC, LASSO, ASR working groups, cloud modeling, and the lab’s SFA projects. The breakout was well attended with around 80 participants. The session started with a few short talks from the panel members to provide a background on ARM science products and critical data and measurement needs from aerosol, cloud, and precipitation working groups, as well as their perspectives on how ARM could better address the data and measurement needs of the ASR science community. The majority of time was dedicated to open discussion, which was well engaged and covered a wide range of topics including instrument calibration, data uncertainty and quality, data discovery, data mining, machine learning for filling in data gaps, better utilizing scanning radar for cloud profiles, building a model simulation case library suitable for various resolution models, connection between ARM and GCMs, communication between the ASR science community and the ARM infrastructure team, etc.Key Findings
A brief summary of key findings is listed below:
- Data quality and uncertainty was a big topic discussed at the breakout session. Data quality is one key factor that users consider when deciding if ARM data should be used in their studies. Data quality is considered poor for some aerosol measurements and uncertainty is large for cloud retrievals. Providing an error bar for some critical geophysical quantities such as cloud properties is challenging. Data uncertainty comes from both instrument uncertainty and structural uncertainty. How to better communicate data quality and uncertainty to the user community needs to be considered. There was some debate about whether ARM should provide uncertainties if they are not well known (is it better for ARM technical staff to make an educated guess than for a user to do so?). One-page summary of data quality information for each of the ARM data products was recommended. The users are also recommended to read ARM Technical Reports and instrument handbooks for data quality and uncertainty information before ARM data is used in their studies. Weaving together retrievals under different conditions and creating retrieval epochs need more discussion as this introduces complexity into how uncertainties are carried through. Instrument simulators offer an alternative path for climate model evaluation, but uncertainties with instrument simulators due to their inherent assumptions need to be addressed, too.
- Instrument calibration needs to meet community standards. Consider reprocessing with calibrated data (e.g., cloud radar data, in action).
- Long-term ARM data over different ARM sites has become valuable in climate research. Cross-site analysis can be very powerful at yielding insights. It is important to have baseline measurements available in a quality-controlled form from all sites and AMF campaigns. Long-term consistency in ARM measurements and retrievals is critical. Need to better enable data mining of long-term ARM data.
- Supporting modeling is one of the major ARM goals. Developing a case library that includes different test cases for modeling studies across scales is useful. Creating bundle data like LASSO bundle data sets can help other modelers. ARM needs to provides tools for users to quickly identify the cases of interest. CFMIP has high-frequency model output at the ARM CFMIP sites.
- It is valuable to retrieve spatial variability on the mesoscale. Scanning radar seems like an attractive tool to achieve this.
- Online tutorials are a good way to educate new users; tools to facilitate browsing data online. Translators can play an important role in communication between ARM's infrastructure team and the ASR science community.