Breakout Summary Report

 

ARM/ASR User and PI Meeting

10 - 13 June 2019

Process-driven Sampling Strategies for ARM Instruments
10 June 2019
4:00 PM - 6:00 PM
80
Scott Collis, Marcus van Lier-Walqui, Brenda Dolan

Breakout Description

In the last 10 years ARM has invested in increasingly complicated instrumentation. These instruments (primarily radar and lidar) are highly configurable. This allows the operation of the instrument to be tailored to observe specific atmospheric processes. The Southern Great Plains X-band scanning experiment, carried out in the summer of 2018, set up a rotating schedule that alternated the operations of the X-Band network between a mode targeting dynamics and microphysics of storms (multi-Doppler mode) and a mode that looked at fast evolution of microphysics just above the freezing level (microphysics mode). Instead of trying to observe the totality of the physics of interest to ARM stakeholders, these modes are tailored to address specific science questions.



This session, which will be a mix of talks and a panel, will use the experiment to drive conversation on the benefits and costs associated with specialized operations of ARM infrastructure. The goal of this session is to deliver, through the session report:




  • An overview of progress from the SGP experiment.

  • A discussion on process-driven processing (certain parameters in processing VAPs are not applicable to all science questions) and targeted sampling. What are the benefits of this approach? What are the costs?

  • Identification of where else (instruments, sites) targeted science-driven sampling would be worth pursuing. Is this the role for ARM or are targeted experiments better left to other agencies? What are the trade-offs between trying to observe as many phenomena as possible versus targeting specific phenomena?

  • The relationship between process-driven sampling and data epochs.

Main Discussion

Presentations by Collis, Dolan, and van Lier-Walqui covered the summer experiment carried out at the SGP. The experiment captured many events but there was a major issue due to a bug in the IRIS software, which was not saving the raw phase. Marcus and Brenda showed multiple analyses and outlined analysis plans. Brenda gave a key example of how we would alter the configuration in the future.



Further talks were given by Paytsar Muradyan (adaptive profiler), Ann Fridlind (Arctic experience including inputs from Sergey Matrasov), and Steve Nesbitt (CACTI experience). Steve made some very good points on how one needs to be careful when letting students “Just take control” of the radar. The CSU C-band system and the C-SAPR performed very well in CACTI/RELAMPAGO. A question was asked: Can the C-SAPR2 do what the CSU radar system did? Paystar made a comment that essentially her changes to implement adaptive scanning with the RWPs were at no extra cost to ARM outside of some mentor effort and time to modify the ingest. The sampling is based on the wind and precipitation types based on one year of data to identify thresholds, and she successfully made the changes in 2018. Ann discussed what pol radar provides in terms of looking at column microphysics and ice crystal properties, including possibly ways to look at axis ratio and characterize sizes. She brought up the point that in the Arctic the revisit time needs to be much less due to the longer time scale.



Very robust discussion ensued. A summary of key questions and findings is in the key findings section below. It was very clear there is a lot of interest in the community in these modes. We made clear the whole community is invited to analyze the data and encouraged them to get in touch with Brenda and Marcus. Some ASR PIs voiced their wish to be involved and wanted to know how to get involved.


Key Findings


  • Lessons have been learned from SGP. Want to run the experiment again.

  • Is C-SAPR2 a better focus? Can we involve the C-SAPR2 when it is at the SGP site?

  • Process study people don't want to see "random samples of data." They want "meat."

  • There are important trade-offs between Statistics, Processes, and Context.

  • Mesoscale meteorologists want context + winds.

  • What is that you want to measure, and what are you willing to throw out?


Decisions

Brenda, Marcus, and Scott continue to analyze data and report back to ARM from the summer experiment. The team will work to ensure that any action in the future is as inclusive as possible.

Future Plans

The team (which will be expanded) will continue to advocate for a second sampling experiment. We have learned a lot from the last one and several issues arose that have since been rectified.