Wednesday, February 15, 2017

Notes from AMS 2017 Annual meeting

AMS annual meeting notes and pseudo report:

Before I start, let me say that there were a million talks I wanted to see. But because the meeting is huge and I can't just bounce between rooms, because it would require me to move at the speed of light and/or become Schrodingers cat or Jesus (being in multiple places at the same time). I am neither (insert sadface.animatedgif).


I like the annual meeting because its got a great social science grass roots movement which is innovative and interesting. There is much new light being shone upon the work being done, how it impacts people (publics', forecasters, researchers), and how new perspectives and angles of attack can help make old problems new again. Wait... even more so, tractable. Meteorology has a hammer and everything looks like a nail. I said so 2 SLSs' ago, look it up.


I could normally care less about a theme at conferences, but observations from the social sciences perspectives brought some new ways of knowing and new ways of seeing to the conference and our field. So I spent my time finding talks I knew would have that flavor.


Panel on "Observations of forecasting" :
https://ams.confex.com/ams/97Annual/webprogram/Session42150.html

-Dr LaDue had some good quotes, namely: "How will the world be different if we get this right?" & paraphrasing: no single discipline can answer all the questions from all the angles.

-Dr. Friedman built upon this noting that in order to build hypotheses, you need observations. In essence, need to observe the forecasters in the wild. In his ethnographies, its important to have the context of how forecasters view their job and how that relates to them making the forecast.

-Dr. Daipha went even further by noting her immersion into the WFOs, including midnight shifts! Also interesting: "being relevant is a part of the forecast!"

-(honorary Dr.) Rick Smith verified some of these observations: "We did the May 20 panel at AMS 2014. I didn't recall some of Jack Friedmans' observations of our office, in the heat of that event."

In the discussion that ensued a few other points were made:

1. Some forecasters already think like social scientists.
2. To brief others, learn their language.
3. We aren't always situationally aware even while trying to be situationally aware.
4. Be careful automating anything - you can rob humans of maintaining or developing expertise.

Observations of Communication:
https://ams.confex.com/ams/97Annual/webprogram/Session42152.html

- Dr. Henderson: study the experts not just the publics'!

- Dr Joslyn: people can and do understand probabilities. Framing can have a positive or negative impact on understanding.

- Dr. Carr: People receiving products wanted cues too. It wasn't just the river height they wanted but what the expected rate of rise would be.

- The five main points from this session in a tweet: "Add soc scientist to your team", " people are smarter than you think", "study experts often", "ask users what they want", "involve user"

"Where we go from here: observational needs" :
https://ams.confex.com/ams/97Annual/webprogram/Session42156.html

Session I moderated on Research with NOAA forecasters and partners:
https://ams.confex.com/ams/97Annual/webprogram/Session41993.html

- Friedman noted that collecting data is just deciding to do it. You have to know when to collect data and what you get might be a function of time as he found out. Literally researching his research design in the field.
"the story is reg shared, refined, rationalized, and normalized in the in the hours and days before & after event"

-Sherman-Morris: most burning Question WCMs wanted to know -- "how to get people to respond to a warning"

-Berry & Obermeier PHI & broadcasters:
1. Coverage decisions were complex. reluctance to cut in during the bachelor, for example;

2. How did they attach meaning to probabilities? they began experimenting with phrases to convey this meaning;

- Eosco:
1. WWA products are the last thing people hear. already been messaging for DAYS;

2. many wouldn't shed a tear if the word advisory went away; "there is a need for more flexible consistency ... "

3. 73% of EMs approve of the current system, tho forecasters and broadcasters want some change;

4. Perception of confusion is not the same as the publics' being confused ... further investigation needed. ;

5. Do we all have a shared definition of what "IMPACT" means?

New initiatives at NOAA in support of resilience and risk comm:
https://ams.confex.com/ams/97Annual/webprogram/Session42026.html

Klockow: we sit in a sea of met data, but human behavior data is not sought much less collected for every event even big events

Sprague: "we do a poor job communicating all the soc sci work we are doing"; need to make studies and reports more accessible.

Marsh: what if we had watches that followed the threats? What is optimal lead time for a watch?; what if the short term outlook was the watch?;

Lots happening at the SPC in the future. Anything from continuous probabilities on our maps, to 4 hour outlooks, to perhaps amoeba-like watches. All in the realm of making finer scale risk assessments.





Shades of Gray (not recorded) on ethics and such:
https://ams.confex.com/ams/97Annual/webprogram/Paper315912.html

Samenow: Our job is gather eyes and clicks, and be compelling & provocative, not hype. Boring headlines make us irrelevant.

Brooks: How do we pass this high specificity forecasts onto our publics?; how do we decide what the best possible outcome is? Referencing health information & decisions...

Eosco: Talking about tor warning comms, at a basketball game: they didn't know about the warning but had right to know, 2 decide; Extreme events communication has a feel of click bait.

Samenow: Back up on being too specific in climate change predictions. Still uncertainty & at same time the risks pose big human challenges

Brooks: Responsible communication is at the forefront, and we are probably behind the curve here.

Samenow: Forecast focus on future blizzard, got sneak attack inch of snow that melted & refroze. High probability low impact that ended up high impact!

Eosco: We are all partners in comm, need local ways to communicate even if communicators are not experts.


Research on risk perception and communication:
https://ams.confex.com/ams/97Annual/webprogram/Session41987.html

Smith: Talked about the increase in storm anxiety. Number of calls to the office are up. Kids in the area have come to visit him at the NWC to show them that they have forecaster-on-duty at all times. Issued the challenge:




Societal responses, risk, resilience to extreme weather:
https://ams.confex.com/ams/97Annual/webprogram/Session42025.html

Karstens: Time series of shelters installed in Norman. Most shelters have been put in new construction on the periphery of Norman, forming a donut. The spikes in the time series are after major events (weather or the removal of public shelters).





Weather warnings: Impact based
https://ams.confex.com/ams/97Annual/webprogram/Session41989.html

Potter: Discussing Impact based warnings noted that although perception of risk was altered, seeking protective action behavior was not much different.






Eosco on IBW: Cues picked up by EMs were from actions of WFO not words, necessarily.



Proposed HazSimp language like "take action" is vague. The definitions or intent were not entirely clear or consistent.

Spinney: In the same vein discussed aligning the definitions of severe and non-severe between publics' and forecasters in Canada.














Bosart symposium II
https://ams.confex.com/ams/97Annual/webprogram/Session41021.html

Skip the science a bit here for mentoring:
1. Just pay attention
2. They don't know any more than you do -advice to Houze as a grad student
3. You have to look before you can see.
4. Close enough for government work.

I enjoyed the story by Houze about the discovery of rear inflow in MCSs'. Basically once the main convection turned linear, NSSL engineers would turn the radar off because the good convection was over. They left it on accidentally and sampled the rear to front flow branch that is a signature of MCSs. Fortuitous.
for more: http://www.atmos.washington.edu/MG/PDFs/MWR85_smul_midlatitude.pdf


Bosart Symposium on Severe Local Storms:
https://ams.confex.com/ams/97Annual/webprogram/Session41020.html

Wakimoto talked about Brian Bosart, which dialed in this session well, its the people that matter, the stories they have to tell, and how we are all in this journey together. One day you may be called upon to lift people up in their time of tragedy or grief. Be there. The real story is not what you read in the journal, but because if we do not put these stories anywhere they can be lost. HOW we make discoveries is probably the most important to educating our students. Roger gave a moving tribute to both Brian and Lance and I was glad I had a chance to see it in person.

I recorded Lances' presentation (audio only) just in case it didn't make it to the web. Probably gonna make future mentees' listen to it.

Vortex SE and Radar:
https://ams.confex.com/ams/97Annual/webprogram/Session42206.html

Rasmussen gave us this slide to think about




after talking about all the field projects associated with VORTEX. This goes to a point Daphne LaDue made about how our methods (or instruments) can point us at certain angles shining light on some things while leaving us in the dark on others. Knowing what you might gain from these different angles (from different disciplines, researchers, or different tools) and working together we stand to learn more than any single angle provides.


Kosiba gave an interesting history of mobile weather radar right up until the explosion of radars at the end (too many to discuss!). It was neat listening to the 1st tornadoes to be observed by these systems. The hallmark being that these tornado radars are now being used to study and explore other phenomena at very high resolution.