 I'm pretty sure that this workshop is going to be a fantastic experience for everyone. We've brought in diverse perspective on what is going on with communication about risk communication about extreme tropical cyclones. My name is Anne Bostrom. I'm from the University of Washington Evans School of Public Policy and Governance. And I would love to be able to introduce the committee members to you in a second here. I think we have a slide for this. Is this correct? And in the meanwhile, I'll start with my other messages. So welcome, everybody. We have a wonderful workshop planned to really lively half days of exciting events, including breakouts and including panels with very diverse perspectives, a broad array of perspectives on tropical cyclones. We have an extremely busy schedule, so we're going to try to keep on schedule and be ruthless with time keeping. And that goes for everybody. So please remember, you're a time limit. And now are we ready for the slides? Excellent. So I'm going to introduce the statement of task and the committee. You can see here the members of the committee. Oops. Thank you. And I'm chairing. Derricka Carol Smith participated very actively in the committee and is unfortunately unable to be with us today. Brad Coleman is sitting here, William Craig Fugate, Mike Lindell, who is also unable to be with here today, but will be online. And Andrea Schumacher from the National Center for Atmospheric Research, Marshall Shepherd and Jeanette Sutton. All names that are probably familiar to all of you. The people who really made this committee work are listed on the right side, the National Academy staff. Hugh Walpo, who has been a study director and been fantastic to work with, John Benzuelo, who brought, whose name I can't pronounce, and who brought the experience of a longer time at the Academy, Rachel Silverne, who I've had the great pleasure of working with before as well. So Sanchez, who helped us quite extensively in the beginning parts of this committee and is no longer here, Rita Gaskins, Rob Greenway, and Eric Edkin. We thank you all for your support for this committee. It's been wonderful to work with you and we're looking for you to see how the proceedings turns out. So with that, I'd like to introduce the statement of task for the committee, perhaps. Uh-huh. Unfortunately, I don't have it memorized. Anybody else have it memorized? Ah, okay. It's a hidden slide. Well, the task was not hidden to us, so I can see it here on the left. Excellent. So this was an ad hoc committee. This is just to get you all to relax. You won't have a few technical difficulties you can't relax. So we can look at it this way. So ah, perfect, perfect. Thank you very much. So this committee planned this workshop to explore challenges and learning opportunities around actionable and understandable risk communication with decision makers for extreme weather events. There's a lot of details here. I'm going to skip the details, but we have three main topics first exploring the current understanding of effective communication practices and features to convey to decision makers uncertainty and probabilistic information about risks associated with discrete extant extreme weather events, such as that that we are experiencing in California right now with unprecedented high winds, which makes this workshop extremely timely and emphasizes the importance of learning more about this as we go forward. We were also tasked to examine risk communication and decision making challenges posed by extreme weather events that are unprecedented in nature or scale for the affected locations. And finally to explore opportunities for learning from synergies, successes and challenges across multiple hazards and decision making contexts and applying them to the hurricane context. And the breakouts this afternoon will especially explore this. So we're really looking forward to learning a lot. All of us who have been working in this domain for a while from other domains as well as from each other. And with that, I'm going to pass it to Andrea Schumacher. So we will be able to stay on time. Thank you. There's some things. I don't think so. Try pressing the button. The red light is always the good, good thing to see. So I'm Andrea Schumacher and I have the pleasure of introducing our first panel today. So the first session we were going to be having today, starting things off, is communicating risks of atypical tropical cyclones, lessons from Henri and Hilary. There are many, many other tropical cyclones we could have included in this list and we don't want to, we want to encourage our panel members to speak on anything that they choose. Those are just two recent examples. The goal of this session is to gain an understanding of the unique challenges, opportunities for innovation and lessons learned in communicating evolving tropical cyclone threats. And our panel members joining us today from the Weather Prediction Center, we have Alex Lamers. And from the WFO, which is the Weather Forecast Office in Los Angeles, Oxnard, we have Rose Schoenfeld. And I believe we have Robbie Berg on the line. I'm here from the National Hurricane Center. He is a hurricane specialist who also works with their social science outreach. So I would actually like the panelists now to each introduce themselves with just a couple of minutes and give us an idea of what their perspectives are and their experience they'll be bringing to this discussion. So start with Alex if that's okay. Yeah, thanks Andrea. I'm really glad to be here. So my name is Alex Lamers. I am the Warning Coordination Meteorologist at the Weather Prediction Center. And we're one of, you know, about 10 National Centers in the National Weather Service. And each National Center sort of focused on a particular forecast challenge. So for the National Hurricane Center, the challenge is pretty obvious. For the Weather Prediction Center, it sounds a little bit more generic, but we're really focused on temperature and precipitation extremes. And those can both come into play before and after hurricanes, but really the thing you think most of is the extreme rainfall. So my role is really to coordinate with our partners and that's internal to the Weather Service and external and work really, really closely with the National Hurricane Center on the rainfall aspects of tropical cyclones. And that partnership has grown in leaps and bounds in the last five to 10 years. So I'm happy to be here. Hello, hello. My name is Rose Schoenfeld and I work at the WFO in Oxnard, California, and we cover the Los Angeles, Los Angeles County, which is currently seeing quite the weather right now as well as supported by the Weather Prediction Center. We're focused on everything in our specific area. And one of the most notable weather events that we have had was Hurricane Hillary and that's really only being topped right now by this atmospheric river that we're seeing currently. I'm really excited to be here and to talk about how that event was significant and how it raised so many alarms because of how scary the thought of a hurricane hitting Los Angeles was. Robbie. Hey, good morning everybody. Sorry I couldn't be there with you today, but I'm really honored to be a part of this panel. So as Alex was mentioning, as the National Hurricane Center, one of the 10 centers within the National Weather Service that looks at kind of the large scale weather. And we have a responsibility for looking at all chopped cyclones across the North Atlantic Basin as well. Okay. We're good now. Thank you. Could you start over Robbie? Sorry, we have difficulties. Sure, no problem. So good morning everybody. My name is Robbie Berg. I'm a senior hurricane specialist at the National Hurricane Center to say I'm sorry I couldn't be with everybody there this morning, but I'm still honored to be a part of this panel. As Alex was mentioning, the Hurricane Center is one of 10 national centers where our goal is to essentially be monitoring and forecasting and communicating chopped cyclones that occur within the North Atlantic and the Eastern part of the Pacific Ocean. So while this panel is specifically talking about atypical chopped cyclones, because we have such a large area, we see a lot of atypical chopped cyclones, not just on Reid and Hillary as we're going to be discussing. I work in a unit where we have, when we're fully staffed, 11 hurricane specialists and it's our job collectively to make the forecast for the hurricanes, tropical storms, other chopped cyclones, coordinate working with folks like Alex at the Weather Prediction Center and Rose at the local WFOs and then also provide briefings and other communications to emergency managers and the media who are also communicating the message of chopped cyclones. So we're in some sense, you know, we're kind of the starting point of the forecast and how things get communicated down the road. So our interactions and relationships with all of our partners are extremely important in these events. Great. Thanks, everyone. So I'll sort of kick things off. We have some prepared questions that we'd love to get answers from our panel on and then we will open it up to questions from the in the audience and on online. So let's kick things off and I'm happy to take whoever wants to jump in or I can pick someone. So what challenges and or barriers have you encountered in communicating the risks expected with relatively rare tropical cyclone events, particularly in the the arena of communicating those risks? I know there's a lot of language associated with hurricanes and tropical cyclones that is common to other types of events. I'm sure you've worked. So I can start with that for for our region in Los Angeles where we rarely see such an event. Something that is a challenge is a lot of the words for hurricanes aren't really going to apply to us. We tried to to focus it and change our message to compare it to weather events. We do see like like an atmospheric river like Santa Ana winds on what we we called it a wet Santa Ana a lot because a lot of that concept of tropical messaging doesn't really apply for us. We're not really thinking about storm surge. We're not really thinking about a lot of the impacts that that will be greater for flatter terrain and places like that. One thing I think just generally speaking about extreme and rare tropical cyclones is that when you get into the tails it sometimes it's hard to know how the meteorology will map on to impacts and how it will translate to impacts right because you may be seeing something that you haven't necessarily seen before. And so that can be a bit of a challenge for us and saying you know for instance hurricane Hillary is an example we're getting all this rain you know Death Valley could potentially get more rain than they see in a year and this sort of thing and it's like well I don't quite know exactly how that's going to manifest into impacts in these areas we know it can it probably will be bad but the question is how bad and so that can be a challenge and likewise I think another challenge is just the constant drumbeat of extreme events that we have now and the pace that's increasing with climate change and getting people to register each one of those and it just doesn't blur into one thing and sort of lose its impact over time. So those are two just general observations that I would say about communicating extreme events and I would add to all of that too is noise you know we have a pretty tried and true process with the hurricane center with all chop cyclones that we forecast and then the communication with emergency managers and other partners and the message that we're putting out often gets conflicted with other noise that's out there whether it be on social media whether it be on traditional media some examples of that include things like over focusing on landfall and a lot of these systems landfall does not matter and oftentimes in the headlines you see a scroll in the bottom of the news it's it's focusing on these what I call gee whiz gee whiz characteristics of the storm it's the first storm to do this it's the first storm to do that landfall is expected to be here and that kind of attention or hyper focus on those type of characteristics really detracts from the message that we're trying to put out about what each of the hazards of a tropical cyclone may do the risk that each of those poses so I think for us is just trying to wade through that noise and have the message the most important message get through to those who need it and if I can add on to what Robbie just said actually you know I think the gee whiz statistics one thing I like to say is I think there's a lot of focus in the lead up to a storm as it's approaching land on what I call the storms of vital signs you know the hurricane centers are hurricane hunters are almost like doctors they're taking the central pressure and the max wind and all this and there's these this hard data that I think people latch on to which for good reason of course and that's what I think because you're getting a regular tempo of new data in that respect it can lead to a focus on those things but of course we know in the last decade actually statistics compiled by the hurricane center actually rainfall induced flooding as the deadliest hazard in the United States and so getting people to focus on that when the average person even in the driest parts of the United States will see thousands of rainy days in their lifetime getting them you know to to register that this is going to be a meaningful rainy day rather than just one of many that they've seen that that is a real challenge to break through really good point there for Los Angeles we definitely get an absolute media frenzy whenever anything happens and we get a lot of like really leading questions that frankly us at the WFO aren't really the people to speak on it we're not really the ones to make that call whether this is going to be more common with climate change that's not really something you can determine with one single event that we're focused in on but that's that's really the thing that can get really distracting a lot opposed to what we're trying to tell people is during this event it's really important to try not to drive which sometimes is the the main message that we have opposed to something that's longer scale and yeah that point of what what rainy days are really important because rain can be really notable for LA whenever it happens but it doesn't always mean that that we are asking people to make changes to their lives thank you these are really great responses I wanted to sort of dig a little bit deeper since you all have roles that are working with emergency managers and other types of decision makers in your CWA's or or your your constituencies how have decision makers reacted to warnings for rare tropical cyclin events you've worked through and what was expected or unexpected about decision maker responses in these events I can start if everyone's okay I guess so you know I'll start by saying that at least along the east coast United States in the Gulf we try to foster really strong relationships with emergency managers in coastal communities and inland as well every year we post through FEMA workshops at the hurricane center we also go out in the field and provide workshops and seminars to emergency managers and we know that having those relationships beforehand helps when the time comes to make a very heavy decision before a tropical cyclone so in some sense I think I'll take Henri for example in the northeast even though they don't get hit by tropical cyclones all that frequently because we already have a network and in a process there by way we work with the emergency managers and we educate the emergency managers they still as far as it looks to us have made appropriate actions ahead of Henri now on the other side with Southern California and Hillary and maybe Rose can speak more in this specifically our relationship with those folks isn't as strong we don't see tropical cyclones hitting Southern California that frequently so we don't have that pre-established relationship with many of those emergency managers on tropical cyclones themselves so I think that's one of the things that comes into play is what kind of relationships do we already have in place and I'm very sure WFO Los Angeles has those relationships with EMS there but from a national standpoint the Hurricane Center's relationship with those emergency managers isn't as strong as it may need to be in the future for Southern California we have really strong existing relationships with our partners and and we'd at least like to hope that that that relationship for a more routine things like atmospheric rivers Santa Ana winds other hazards fire weather will carry over towards more more unprecedented newer more rare events like like hurricanes or tropical tropical storms during this event our emergency managers in our region took it very seriously they activated EOC's for all of the counties we had people deployed to them to help them make those decisions and largely they acted well and by the end of the event the the general consensus amongst our emergency managers was what is was that it was generally a success that people weren't driving as much that there wasn't too many too many swift water rush rescues due to getting people out of those areas and warning ahead of time and there wasn't too many roadway flooding issues in terms of cars actually got running into hazards so at the end of the day we feel pretty fortunate in terms of how the impacts worked out for for life for Hurricane Hillary there was definitely property damage but some of that is a little harder to mitigate since it since it's the roadways are there the mudslide is not really a stoppable as protecting life yeah I think actually was there zero fatalities in Hillary or I think so yeah I mean I don't know of any there might be some but yeah I thought it was zero or it's very close to it which is pretty amazing right you have a really unusual event just the not only a tropical cyclone but the amount of rain that it delivered especially to inland areas and you're talking about hundreds of millions of dollars of damage to some of these inland counties and it was either their costliest hazardous event on record or one of the costliest and talking about a significant percentage of their county GDP or close to it so to get zero fatalities out of that is pretty impressive so clearly there's some connection there that's going right at the local office you know some one of the surprising things that I've noticed more recently is evacuating people out of harm's way for the rainfall aspects of these storms not necessarily a whole scale evacuations of a major metro area like that would be a huge undertaking at Robbie can speak more on that with how they do that for the storm surge but at least in a targeted way towards vulnerable populations so after you know the back to back of honoree and Ida in the same basically in a few week period in New York City they had a new flash flood action plan it built in targeted actions like evacuating people who live in basement apartments and ways to get to those people and get to them in advance in Hurricane Hillary we saw them evacuating some homeless populations out of vulnerable areas in advance hundreds of people and I think that's pretty interesting it means that there's a confidence in these decision makers in the rainfall forecast that we're providing and the information that we're providing and that they can take action and so I think understanding a little bit more of you know how they're making these decisions sort of the clearance times that they would need to evacuate these vulnerable populations would really help inform us as we try to improve these forecast over time yet to add what Alex is saying the one thing that we really have to keep in mind too is exposure you know looking back some of the storm surge events from from top cyclones you have a cyclone like Ian that hits Southwest Florida a couple of years ago and obviously Florida is you know well used to talk cyclones yet we had a lot of deaths from storm surge and I think it points to the fact that even though a region might be familiar with us with top cyclones hurricanes that doesn't mean a locality is familiar the top cyclones and hurricanes Fort Myers Beach had not been hit by a storm like Ian for years they did have Charlie in 2003 but Charlie was very different characteristic wise and so you know a lot of deaths there but then you take a storm like Adalia last year again also hit West Coast of Florida up near Appalachee Bay no storm surge deaths at all there so it might be quick to say oh well the storm surge warning program that we have it's working but it's it's hard to say it over a year you know we have a change I think it's because of exposure and the populations in those areas and what they're used to what they're knowing whether or not they're at risk of the storms and taking appropriate action so you know it's it's it's hard because every storm is different and you have to take into their the storms characteristics and then where that storm is actually threatening and I think that's one of the aspects of Henri and Hillary is you've got two areas where the frequency of storms is not high and so we know that the locality is they're not going to see those types of situations in hurricanes that frequently great thank you so much I think it's a really nice theme that we can think about these next two days also is experience and how that varies not just with people but with time as people come into contact with some of these new hazards so now when I turn it to asking some questions specifically about these events and have you used any novel or innovative approaches while you were communicating these rare events and how successful do you feel they were for for our office something we did that I don't know Robbie might have remembered is our MIC or meteorologist in charge really pushed to get the timeline for products issuance accelerated so normally there's there's a standard window that tropical storm watches are issued tropical storm warnings are issued but since Southern California has never seen that product issued before ever we really push to issue that earlier and we did issue that the tropical storm watch and warning ahead of what the normal hurricane center schedule would be to try and give people some more time to process those new products that also those products might have even been in some ways it could have been somewhat of a distraction from the message maybe going forward it won't be anymore but with that first one that novelty of it definitely could have distracted people from the main message being this is more rain than usual for August by a long shot this is that that's the main thing to prepare for besides like whether or not it's the first warning or not you know I think one thing that really resonated with me with what Robbie was saying earlier is the importance of laying track with partners before the events and making those connections and part of that is you know establishing the relationship but part of it is understanding the products and that sort of thing and so one one thing that we do with rainfall as we use the excessive rainfall outlook here at WPC to communicate the threat it's for people who don't want to get into the details it's a simple tiered outlook it's kind of stoplight pattern and people can sort of intuitively understand that the high end of the scale the reds and the magentas are bad and but but to use a slightly different storm as an example Harvey in 2017 the policy was we could only issue the highest category of risk up to two days in advance day one and day two but our forecasters were looking at that they were sitting on the operations floor just the model forecasts of an astronomical amount of rain and they were saying look if we're ever going to issue a high risk three days in advance this is when we should do it like this is the event and they were so confident that by the product definition that probability there that probably did exist of that extreme of an event that they coordinated a change in real time to weather service policy it happened in about 10 minutes it was coordination from the local office to WPC to the hurricane center to weather service headquarters and our sue just on the fly made a change to the coding to allow that to happen and I think that's a really cool example of the way that you know when we feel that an event is potentially really really impactful you can get everybody on board really quickly and singing the same tune and consistency of message I think is really important in helping it sink in so we knew the local office was going to really want to hit that hard and just gave them another tool to do that saying hey this is something that the weather service has never done before with this particular product and hopefully it made a bit of a difference. Yes, so real quick I'll say kind of piggybacking on what Rose said with Hillary we did issue the watches and warnings in Southern California a little bit earlier than what the typical lead time is it's 48 hours for talks like watches excuse me in 36 hours for warnings and that's not abnormal with other events we've seen in fact lots of times for some of the more major hurricanes one of the issues we have is when when storm surge and hurricane warnings go out it also sends out we alert messages to folks and we know that sometimes if those warnings get issued say at 11 p.m. or 4 or 5 a.m. There's a little bit of consternation among the public when they're getting woken up to by these alerts so we've actually worked with the local WFOs to slide some of the issuances of those warnings to a more appropriate time to communicate them so that people are already awake it's not going to bother them at times overnight so we do have a little bit of that flexibility even though there is a specific policy times with those watches and warnings the other thing I was going to mention as far as of novel approaches I would say for these two storms specifically from the National Hurricanes perspective again when we're looking at large scale what we did during these storms is not too different from what we do with any chop cycle in the threats threatens the United States we have policies and procedures that we use every time there's a threatening storm we increase and grow the message messaging as we get closer to the event two things that we've done recently past few years is one we've introduced key messages and I know this has been starting across the weather service now and I think it's great because it's really setting the agenda and setting the message for folks in the media emergency managers to carry that message forward with their constituents and who they're speaking to and then also we've done a lot more live streaming on Facebook on our YouTube channel and we've found that even some media folks are tuning in to those live streams that we do to take what's the most important message at that time and they're re communicating that message to the people that they're serving so again it's not specific to these two events but I think it also proved to be beneficial in these events by us being out there providing these kind of chunk messages that are easy for people to digest and re-communicate during an event. Thanks so much this last question I had for the panel maybe a little touchy I'm sorry so feel free not to you know how to dig in too much but I wanted to know if there was anything that you wanted to share with us that you would have done differently during these events these recent events and you know that can include on re and Hillary but also you know we've had IDAida which was a really large rain event and Northeast so just curious if there I guess this is where we would love some lessons learned if you have some I can go first so I think just generally speaking in the weather service at any office we try to learn from every event regardless of how well or bad it went and and you talk to your partners we've gotten I think better at doing that in the last decade and they'll give you comments they're going to give you honest feedback and say well could you do that or could you do this I really like that but can you do a different way and so trying to learn from that and you know on re and IDAida the back to back in New York City was sort of I don't know that it was like really a well it was a bit of a lesson learned is just how important the rain rates are in urban areas I mean that's the specific act aspect of vulnerability and exposure like Robbie was talking about and so having a way to allow these large cities where you have huge amount of pay surface and people concentrated in one area our director like to calls likes to call it the payment and pipes problem and so you know we're we're working we have a project we're working on developing an urban rain rate dashboard that will give decision makers in large cities a chance to see when these chances of extreme rain rates are spiking up in their metro areas so that's not a finished product yet but we're saying okay we understand that there's a fundamental issue here that people want more information on and then trying to develop a solution so that that'd be one example something that we've learned from this storm would be to just just trying to learn how to communicate and better communicate that uncertainty with the tropical system that's always always a challenge but it definitely like our feedback at from less so a lot of our partners because they were working and seeing more impacts like across the the region but from the general public a lot of the feeling that we've gotten is that people feel that hurricane Hillary was over hype that it was that it was nothing like it always just rain oh there wasn't really anything I didn't feel any wind where I was so something that would be helpful going forward would be to try in once we have a better idea what the the track is doing because when the track is a little more uncertain when that the because this this storm made greater impacts for Baja California but we the whole time we didn't know exactly where some of those greater impacts we're going to end up surfacing so once we start to see some impacts being less important trying to communicate the new focus of hazards to people so that they don't still feel like we're telling them anything could happen and then they only see some rain or something like that so to keep that warning warning fatigue a little lower because we do want people in the public to listen when we say something but that it's always a concern that then they won't take the next thing more seriously like for example this atmospheric river we we want to convey that this is really significant people kind of having that impression that oh well since Hillary wasn't too bad this is not going to be bad so trying to yeah narrow in on what what the new impacts are as we as the system keeps evolving and I'll circle back to the noise that I brought up at the very beginning you know and also this kind of touches on what Rose is saying here is we often struggle with these systems and it's not just on re and Hillary but if a system is a strong hurricane or five days away people think and see that hurricane is being the same entity when it reaches their locality we call it a hurricane if we people have a mental image of what a big bad hurricane does and then when it's weakening even if we're expecting it to weaken as it approaches the coast and the hazards like rainfall become you know the bigger biggest concern that's not what people have in their mental image of what the storm may do and so it's how do you get through the noise of what it's called it's it's a hurricane now cat three but it's expect to weaken how do you get people to focus on the watches of warnings that are actually in effect where they live regardless of what the storm is doing right now and to understand all the different hazards that that storm may pose rainfall storms urge wind tornadoes even rip currents we see so many deaths along the coast from rip currents from storms that aren't even going to make landfall so I don't have an answer here hyper focused on each of the individual hazards of a storm what the risks are of those hazards individual localities states and trying to deemphasize for example what's the category of the storm because in the end that may not be so important for the story thanks so much but this time we'd like to open it up for questions I don't know John Ben do we have anything worth jumping in yet or should we keep it in the room and we can take some questions from the room first and then we'll move on the virtual okay yes go ahead hi I'm Sarah McBride from the USGS I'm just curious for the panel how useful scenarios like ArcStorm have been in terms of communicating this massive event currently in California if that's been helpful if that's been difficult to explain the difference between the ArcStorm scenario and what is currently happening to my understanding that ArcStorm term is this I don't I'm not sure how new that is but it definitely took off recently for us in terms of describing California's big disaster being some really significant storm that's will cause massive flooding and shut down the state for a long time in comparison like Pacific Northwest thinks about like Cascadia so that's kind of like California's major version of that we've generally found it a bit distracting because that is something that people want to try and catch catch clicks attention because of how scary and cool sounding that name is when when we're trying to hone in like on on our specific area and then say like these three counties are at the greatest risk so it it can be distracting but and another sense it may be having that comes concept of the worst case scenario is helpful for emergency preparedness or for emergency management to have that idea what the what the the big one could be so it's kind of a mixed bag yeah I'll agree with Rose I think there's a big difference between the public communication aspects where I think it's probably more of a hindrance to have people throwing around worst-case scenarios and saying well this is the one you know um but I think for stress testing uh emergency response uh and preparedness I think that worst-case scenarios uh imagining what it could be is really useful I think that's actually one of the ways we talk about in the weather service support services and a lot of the focus is on real-time support but of course there's a whole EM cycle right and there's other aspects of that that are super important so um having that real deep connection uh in blue sky days and saying you know well in this particular community here's about how bad we think the rain could get you may not have actually experienced that yet are you actually ready for that what would you do um one example we did an exercise we WPC we partnered with the uh local office in New York City and we there was an event in 2014 on Long Island where they had like 14 inches of rain or something and the hourly rates are like 5 inches per hour so that was worse than it was in Ida in New York City proper we're like if you move that over 30 or 40 miles it'd been right in New York City and you know the results would have been really really bad and what would you do in that scenario so that's an example of I think how you can have that deep connection with partners and stress test the system in a theoretical way and hopefully help them be a little more prepared yeah and I'll say there you were already thinking ahead um for example if you look back at hurricane Otis last year hit Acapulco uh the devastation that occurred there we haven't seen our hurricane catfire strength hit a major metro area in the United States uh in a long time Andrew did buy me area but Andrew was also small um it could have been much worse uh so we're already starting this conversations with emergency managers this off-season to start thinking about what would you do if you were in charge of a major metro area and you had this type of storm threatening your area um so I don't think we'll have the answers uh but you know we we want EMS and other folks to be as prepared as possible because there is always going to be uncertainty in these events we're never going to know for sure if it is going to be the worst case scenario but oftentimes you have to prepare for that possibility there's so many questions here I'm sure there's a lot to go through here um Alex you brought up the payment and pipes problem and that you're developing this um urban rain rate dashboard will that apply everywhere and I'm just thinking about how little we know about some of the pipes in some cities in Atlanta there were a lot of big sewage pipes they didn't really know where they were or how big they were um and I think it's similar in some other cities yeah it's a great question um I think the initial so the practical matter I think it initially it'll be you know several dozen cities and then it may scale up from there um and yeah in terms of the practical impact of you know do cities fully understand the pipes in their city I I mean they probably varies and I I think modeling that in a hydrologic sense is a challenge um so you know we have colleagues at the National Water Center new the newest National Center in the National Weather Center that I think that's a tough nut to crack but I think they're going to work on that to try to not like can we get the modeling right in urban areas for us we're trying to communicate the meteorology of the situation and and generally speaking they have an idea of what sort of rain rate will cause problems in that city just based on past experience of course if you're off the scale um you you may not know but um usually they'll know like an inch per hour or two inches per hour what that will cause Karen Florini Karen Florini with climate central I'm wondering if if you have anything you'd particularly like to mention in terms of the rapid intensification storms where you know all the sudden thing that was barely a tropical depression is now a category whatever and how that affects warnings yeah excellent question so I think the research showing that rapid intensification events are occurring more frequently um we have not cracked the nut however we we are a little bit encouraged in that some of the statistics we've been compiling on our forecasts seem to just suggest that we are actually getting better at forecasting these events again not perfect by any means I mean we totally missed Otis um you know it went from a tropical depression tropical storm up to a major hurricane in the span of of the day or two so we still have events that we're not capturing but with a lot of the investments that have been made in the National Weather Service in NOAA on uh on intensity forecasting essentially we have seen the models providing more skillful guidance to the forecasters and so we've seen a shift where we're we're capturing these events more frequently I think that's a positive um but you know talking about the Otis example in the US how do we get folks to realize that even if the forecasts are getting better they're never going to be perfect and we still have to prepare for these events especially as before they happen so excellent question and it's definitely on our radar you know as we speak we have several questions from our virtual attendees these are directed to the entire panel does strategic risk communication such as best practices look different on other continents and do we learn from one another I I can start there from the chop cyclone realm so one of our duties at the hurricane center is to coordinate very closely with first for storms that are threatening that the region we're not responsible for putting our out hurricane watches and warnings for those countries they are ultimately responsible but we coordinate with them every forecast cycle to determine actually make our recommendation and they determined whether or not to put those watches and warnings out and to my knowledge our region is the only region across the world where that occurs there are other national hurricane centers in other parts of the world there's one in Japan Australia Fiji and so forth India but they don't have the framework that we have in our part of the world where they're coordinating with other countries around them quite as closely as we are and so not that there are places we can get better but I think what we do in our part of the world really it shows the strength of the partnerships we have with our region and I think that could be something that the other regions should probably be focusing on and strengthening as well in order to get those risk messages out to other countries that aren't actually making the forecast themselves Next question from the virtual attendees again anybody at the panel do the strategies change as the lead time changes from a small signal at two weeks to specific actions at two days for example Absolutely with our with our weather on the west coast coming off of the oceans a lot of times there's a lot of uncertainty several days out two weeks out you could get models saying anything so with this atmospheric river here we saw the signal pretty far out for something like this much further out than usual so we we did start talking about it ahead of time we talked about it having a potential to be more significant than the storm we were currently dealing with when we started messaging this we we were putting it at like 30% chance we tried to compare it to previous previous storms things like that further out put that that potential but we we don't hone into the specific details and action items as much until we get to that that closer a few days out but we definitely try and message that there is that chance especially to our partners a little bit less so to the public in terms of messaging several weeks out because then that that could start that warning fatigue of there's always a storm every two weeks out always be prepared and we do want to let our public like have a breather and and relax to in between things but for our partners who don't have nearly that issue we we're we're happy to tell them like there's like a 30% chance right now there's 15% chance right now even in the week and then some time period Yeah I think I'll just add I think the conversations you have with your partners that does definitely change like Rose said that you can have that and then I think there's also value in you know like our national center outlooks of having a consistent methodology across lead times and and a probabilistic framework so you know like even if it's five days out if the probabilities are really high well that's meaningful right because you don't usually see that and theoretically a probabilistic outlook should be able to account for those uncertainties if it's well calibrated that far in advance the other thing I think you know when you're considering lead times as you occasionally run into these issues with the rhythms of daily life right you it may be a Friday right and you're like well people are about to go check out for the weekend so this may not be the ideal lead time with which to put out this message but we we are getting concerned about it maybe we need to try to reach as many people as we can now before people go into the weekend so understanding those things and Robbie mentioned something earlier about sending out we alerts at to a that may not be ideal well we're unfortunately out of time this has been a great discussion I really appreciate all of your time and willingness to talk about this this has been really great to start things off so I guess we can thank our panel and and we'll have our next panel come up and we'll turn it over to and Bostrom everything we're going to move right on to our next panel so if you could take your seats we're going to move forward right now after the next panel we will be coming back together for lunch okay on to the next perspective on communicating risks of tropical cyclones again I'm and Bostrom it is my great pleasure to welcome are we ready is everybody ready it's my great pleasure to welcome our panelists for a researchers perspective on communicating risks of atypical tropical cyclones we have three speakers in this session who will take a more theoretical and and research perspective on this topic first we have Roxanne Cohen Silver from University of California Irvine welcome Roxanne we have Julie Demuth from the National Science Foundation National Center for Atmospheric Research welcome Julie and we have online hi Emma we have Emma Spiro from the University of Washington so a really excellent group of researchers here and Roxanne is going to kick us off you may do whichever you like we're just getting those set up now I think they're wonderful thank you very much and thank you for the invitation to speak to you today on research perspectives on communicating risks of tropical cyclones about six years ago my colleagues and I published an article in risk analysis in which we conducted a literature review a systematic review of decades of research on how people evacuate from natural disasters whether or not they do and the characteristics of research we concluded that there were many conflicting results largely because of the limitations of the the methodological limitations of the research in this area many times people were asked retrospectively sometimes days weeks months or even years after a storm to tell the researcher a little bit about what their decision-making process was and how they were what their behaviors were and we concluded that this research really needed to take a different course we indicated that ideal research on the topic of evacuation should be perspective trying to identify at-risk samples before the event occurred before the disaster occurred ideal research should be longitudinal with immediate as well as repeated post-event assessments ideal research should use representative samples of the population under study and ideal research should be comprehensive we concluded that small cross-sectional studies with non-representative samples are basically inefficient and can provide little new information after decades of research that's been conducted thus far with the generous support of the National Science Foundation or my colleagues and I conducted a study that was designed to address the limitations that we identified in our risk analysis paper we conducted a study of Hurricane Irma in Florida in which we collected data as the hurricane was barreling toward Florida about 60 hours prior to the when it hit made landfall in Florida we had a representative sample across the state about 90% of the participants completed the survey within the first 48 hours we had about 1600 about a month later we can followed up we retained about 90% of our sample collected about one month after the hurricane hit Florida we had about 1500 people you can see from this map that we covered across Florida by wave two we were still covering the vast majority of the state and again these were this was a representative sample um this kind of an design enabled us to investigate psychological responses before and after the hurricane we published a paper in 2019 in Jemma Network open looking at the psychological impact of the hurricane but what I would like to talk to you about in the remaining minutes that I have here is how we came to study people's perceptions of evacuation zones so right before the hurricane was hitting I took this picture of my computer screen was on cnn.com in an evacuation zone leave now of course that presupposes the people know whether or not they are in an evacuation zone and this was also shown on the news right around the same time as you see the entire state of Florida is being enveloped in a bright and this is for me very easy to understand why it might be difficult for people to know whether or not they're actually in an evacuation zone so we collected data both before and after the hurricane I'm just going to whip through this to make one message before the hurricane people said they were not an evacuation zone many of them evacuated anyway before they would evacuate but they did not evacuate when reported one month later in fact some people said that they were in the zone but they wouldn't evacuate but more than half of them actually did evacuate when we assess them a month later under threat people's perceptions of their at vacuations on status may not be accurate and we had the latitude longitude of the participant in our study and were able to map on whether or not they were in an evacuation zone as the hurricane was approaching and just in very quickly the yellow bar are individuals who were inaccurate in their perceptions based on their GIS mapping onto evacuation zones people were both inaccurate in terms of whether they said they were in a zone if they were not in a zone whether they decided to evacuate or not about 40% of our sample incorrectly reported their evacuation zone status about 6% indicated that they were unaware of the evacuation orders and almost 17% over 17% of our sample evacuated unnecessarily what predicts whether or not people evacuated well not surprisingly in this came up in the prior session whether or not people had previously evacuated for prior storm was a strong predictor of whether or not they actually evacuated from this storm independent of whether or not they were in an evacuation zone and also before the hurricane people told us how likely they were to be impacted by the hurricane those pre-hurricane risk perceptions were also a strong predictor of whether or not they evacuated but what predicts the pre-hurricane risk perceptions actually the strongest predictor was media exposure before the hurricane and this was also something that just came up in the prior panel I think it's really important because of the two slides that I showed you in the evacuation if you're in an evacuation zone please leave and then to show a cone across the entire state I think it's really important that we and the research and weather community start thinking about what we can do going forward as we are thinking about making changes in both communicating as well as researching the one message is that retrospective reports of pre-hurricane perceptions may not reflect reality under threat perceptions of risk appear to shift over time perceptions of evacuation zone status also appear to shift over time and perceptions of zone status may not reflect reality finally I think partnering with the media is critical and I again was very gratified to hear in the prior panel the discussion about working with the media in advance of the storm shifting risk rarely results in shifting evacuation messages and perhaps that may be revisited or importantly I think debriefing after storms both with the media as well as with the general public is critical for ensuring a population's understanding of in preparation for the next storm I'd like to thank the generous support of the National Science Foundation as well as my collaborators and facilitators thank you thank you Roxanne Julie is there that's going the wrong way there we go okay so Roxy and I are going to compliment each other really beautifully with our set of slides here I'm just going to get a little bit more granular with what I'm going to talk about so just to begin thank you for the invitation here I'm Julie Demuth I'm a research scientist at the National Center for Atmospheric Research and I want to dive right in but I want to thank my many wonderful collaborators on this effort the research team that is listed here and we have generous support from Noah and many wonderful Noah collaborators that are listed down there at the bottom so like I said this is going to compliment what Roxy talked about really beautifully but just on a more granular scale this is the research question that really is the foundation for the work that we want to investigate looking at a given hurricane event so not across events like Roxy was doing but for a given hurricane event how do the risk information how does the risk information that people use their risk perceptions and their protective actions that they take that they take how do those evolve as the real world hurricane itself is evolving and as the risks that it is posing are the hazards which Robbie talks so nicely about as those are evolving as well this seems like it's a straightforward question to investigate but in reality it's quite tricky we did do a prospective longitudinal panel design where we surveyed the same people repeatedly during a given event as it was threatening an approaching landfall we got three waves in during the predictive phase when the hurricane was threatening an approaching landfall each of those waves was in the field for about 24 hours so really short turn around with about a 24 hour interval in between and we did that three times the hurricane was threatening then we also did a post storm survey so that we could see what people's actual experiences were some more details there I would be more than happy to talk about the methods in great detail if people are interested we did this for Henri but before that we actually did this initially in 2020 for hurricanes Laura and Marco which were threatening the Gulf Coast around the same time and then we also did this a year and a half ago now for a hurricane Ian and this is just an image of where we fielded for each of those storms anywhere from 50 to 100 miles inland in the the black areas are the zip codes of the respondents who completed the surveys for each of these events we did get in three predictive waves and you can see we had a pretty good sample of anywhere from 800 to over a thousand people who were responding to all of the waves for a given hurricane so even though we did this for three different events we didn't survey the same people for all the different events of course so really focusing on how people are responding within a given event and just a few more details we talked a lot about Henri of course it made landfalls a tropical storm but these different events had different characteristics so Marco never actually made as a weak tropical storm but then we had Laura and Ian which both rapidly intensified to category for a hurricane so it's interesting to look at how people are responding across these different storms which again have very different characteristics but just digging into Henri a little bit more because this will become important this shows what the cones and I know we shouldn't just focus on the cones but just showing what the cones are looking like at each of the waves when we fielded and one thing I really want to draw your attention to is when we fielded that first wave wave one you could see the cone was actually kind of showing that parts of New England were threatened but halfway through wave one that cone shifted a little bit west those hurricane and tropical watches went into effect and at that point in time Henri was forecast to make landfall as a hurricane and then you can see kind of what happened at wave two and way three but that shift within that first wave will become really important with some of the results I'm going to show one more quick methodological detail here is because the areas that are at risk are shifting over time we decided we wanted to categorize people as exposed to the hurricane risks or not and we did this for each of those events and for each wave based on whether people were in the cone if they were in watches and warnings and if they were in areas where tropical storm force wind speed probabilities were greater than 30% or if they were in an evacuation order and so what I'm going to show are results just of the people who are in these exposed areas so that we can compare these across the different events and so just to dive in actually one more quick detail because we were talking so much Roxie mentioned this too about I don't have time to dig into the details here but the on re residents who were exposed have lived there longer about 10 years longer on average but they have fewer past hurricane experiences and people who responded for Laura Marco and for Ian okay so just a few details here so for all of these different slides I'm going to show different kind of plots that are the different questions the different dependent variables we want to look at we're going to show the on the x axis are the wave so we're showing the means over time at each of those waves talking about the number of times per day that people are getting information from different sources in oranges on re blue is Laura and Marco and then black is Ian so just a few things you'll see from these results here if you look at on re you kind of see the increase in terms of how frequently people are getting information from wave one to wave two again potentially because that cone is shifting and those watches went into effect and you see this across the different sources that we that we're looking at here but for the most part compared to Laura Marco and Ian people are getting information about as frequently across these different events from all these different sources the one difference we see here is for the National Hurricane Center where if you look at that orange line on average overall the different waves people on on re are getting information less frequently from them compared to Laura Marco and Ian the other thing we measured was environmental cues we know that this is really important to people and so interestingly by wave three for on re people are looking outside more we collected data for so that's one set of results let me see if I can get this to move forward we also wanted to look at how important different types of information were and so this is a similar set of results here is what I showed before but here now we're looking at how important people think the cone is or the track are different wind speeds and so forth and again if you look at that orange line you see kind of that bump from wave one to wave two for on re so people think that the information is more important from wave one to wave two and we see that for the most part they think that cone is really important the track and the wind speeds this is really where we see the biggest increase from wave one to wave two but other than that again not a lot of difference across the different storms right went on to measure risk perceptions which we measured in a lot of different ways but one of the ways we measure this was by asking people how likely do you think it is that the area where you live will be affected by the different hazards just touches nicely on what Robbie talked about trying to communicate these different hazards again you see that big bump for on re from wave one to wave two across all they're most likely to be affected by those strong winds less so by rainfall flooding less so by surge flooding for on re overall people thought it was less likely that they would be affected by tornadoes which ultimately I think played out we also wanted to see people's perceptions of how likely they thought they would experience different negative impacts or their perceived susceptibility again you see this theme comes up over and over again this bump from wave one to wave two and people thought overall that they were most likely to experience power outages and road closures this was true across all of the storms and then on re people were less likely to think that they would experience emotional impacts or financial losses compared to the other storms and then just one more quick set of results I know I'm going through this really quickly we also wanted to see what protective actions people were taking at those different waves across the different storms and so here now focus on that orange bar that's on re and really look at the areas where that orange bars is shorter than the the blue or the black bar which are Laura Marco and Ian and you can see the overall on re residents are getting about that the number of people who are following the forecast information moving things or doing other home preparation about the same as for the other storms where you really see a lower percentage of people doing this and on re is for people who are getting supplies gassing up their vehicles boarding up or evacuating but otherwise we're seeing really similar kind of protective actions for some of those different behaviors so just to kind of summarize I know I went through this quickly yes on re with aty typical in a lot of ways but the respondents maybe not so much there's a lot of similarity across the different storms or a few things that are different there that I just highlighted but just to kind of underscore this broader implication and again I think this follows nicely on Roxy's talk this kind of longitudinal perishable data that we're collecting during a hurricane event as it's actually threatening is really essential to understand the dynamic processes that people are going through and that if we really want to assess how well the risk communication is working and identify where we need improvements we have to be able to collect this kind of information across different hurricane events sort of a screen capture of a recommendation that we put into a report and I'll go to this final slide here where we made recommendations as we piloted this methodology for how NOAA and the broader weather community can expand on on this kind of data that we've collected here so with that thank you yep thank you Julie and now we'll turn it over to Emma good morning all I'm gonna ask to share slides from here so I'm gonna does that look like it is working? Yes looks good perfect okay and I apologize if my video is flickering it seems to be going in and out a little bit but it's lovely to be here all with you all virtually apologies I couldn't be there in person so my name's Emma Spiro I am an associate professor at the University of Washington information school and one of the co-founders and incoming faculty director for the Center for an informed public here at the University of Washington so this talk I think is also going to pair really nicely instead of talking about sort of surveys and survey data I'm going to be talking a little bit about online social media data so let's see here so since network social media sort of hit the scene especially in crisis communications sort of in early 2000s our team has been studying how so this could be a natural hazards as our focus is today but it also includes things like civil or political unrest domestic terrorism and breaking news events like elections so you know this is this is our sort of bread and butter all that noise that Robbie was mentioning in one of the earlier panels our work is really grounded in the social sciences so aiming to understand sort of how information know a lot about how this works during crisis events we know people come together we know they come together to try to make sense of what is going on around them and they need to make these kind of critical decisions about what to do and when as we have heard from our other speakers so they need information to do this and we know it's also very common in these kind of situations for rumors to emerge and these kind of rumors have been studied for many years in the social sciences there we go and a rumor is a sort of term that I would sort of define as information unverified at the time it's being talked about and this sort of concept conceptually has been the focus of our work over the past decade here at the University of Washington and previously at the University of California Irvine where I was as a graduate student so rumors can be really useful in situations they allow people to come together together alleviate their anxiety and take protective actions and we know that people use online systems to access information we know that these vast amounts of social and behavioral data are were and are available to researchers and so we can learn a lot about how people engage with information during crisis events but the challenge here is that we also can see where things go wrong where this collective sense-making process that I talked about before could be strategically manipulated and result in mis and disinformation so our team at the University of Washington collects millions and millions of social media messages across a handful of different social media platforms and then we're able to compare things like how information flows and how people engage with information across hazards. We look at message content at format about how things are designed in social media spaces to understand why some things spread and others do not. So you know things that are about hazards and their impact are sort of highly spreadable and can go viral online things that are sort of more about thank yous or directed at particular entities or accounts online far and wide. So I think one of the really interesting things in this space is that over the years we've kind of developed a much better sense of what I might call the sort of rules that govern information flow in these systems and you know from uncertainty and trust to social networks to emotional valence to the actual words that are present in these messages to the larger narratives that are sort of part of people's world view in different communities that are affected by disasters. And so we know a lot about the features of different information narratives of rumors that contribute to their virality in online systems. And I think it's really important to recognize in these systems also that people engage and they spread information because they want to help each other right. There's a sort of well meaning I'm going to pass on this emergency evacuation notice because if I don't someone could lose their life. So especially in crisis situations it is really important to remember that people share information because they're trying to help. Now lately despite all that we know these information environments are becoming more and more challenging more challenging for researchers which I will talk a little bit about at the end more challenging for responders and more challenging for the public. And that is because they become rife with miss and dis information false and misleading messages and bad faith actors that are trying to manipulate the information environments that we all live in. And we're all vulnerable right. This is something you know you talk to researchers of missing disinformation and they've all had an experience where they have shared message that turns out to be questionable or perhaps not framed in the right way. And this is a really key aspect because not only are we all vulnerable but we all have to participate in order for information to spread online. This is a very participatory kind of phenomena and of course these two things in combination makes it very challenging. These problems are only growing in complexity with the introduction of generative AI tools so artificial intelligence tools which is like this of this so you know claimed pentagon bobbing we have challenges around not only the increased quantity and quality of information right and a very low bar barrier to doing that but also increased personalization of information persuasion of information and other kind of open pathways for these tools to involuntarily generate false or misleading information due to what are called hallucinations in the system. I do want to point out that this is not necessarily a new problem just because we have for example weather images that can be created by artificial intelligence based tools you know that this has been a problem in sort of the crisis space for many years so you are all probably familiar with this image every you know tropical storm hurricane that we have we see a shark swimming down you know a national highway but these tools also offer interesting opportunities especially in the research and sort of prediction domain for how they can be used to improve our response to crisis events so from AI tools being used to prediction to various kind of interventions that are being designed in the research community to help support the collective sense making problem process to help people make decisions during these very uncertain situations. I do want to leave plenty of time for us to discuss so let me wrap up a bit with a few of the sort of challenges I see especially coming from a research perspective over the next few years you know we've made lots of progress over the past decade plus but we're really at a challenging sort of inflection point at the moment where social media platforms in particular are rolling back previous access options for really has affected our ability to collect curate archive and share data all of those things have been significantly restricted in the past year plus we also see that the online environment the social media space is becoming increasingly fragmented so there are many many platforms and users sort of move across and within them they move information and they have different presence on these different platforms this makes it really difficult for researchers it also makes it challenging for users to navigate and find the information that they need and for officials to know where and when to share information in these spaces we also see a lot of what I would call multimodal information so text plus audio and video is becoming much more widely used you know the rates of YouTube and tiktok in these newer video based platforms especially in younger populations is significant and rising and there are many new research challenges there and then you know maybe perhaps most importantly we're operating in a world now where we see trust in information and trust in institutions is at risk and this last element I think is really critical to address because in fact many of the sort of phenomenon we study here at the university of Washington the goal of these activities is to really undermine trust in information in institutions and in each other and so and so and this is a really critical element and I hope we were able to discuss how this and some of the other things fit in with what we've seen so far this morning thank you Thank you Emma Hey we're going we're going to kick off I'm going to kick off with a question and then welcome your question so please put questions in Slido if you have them or raise your hand and so I'll start with the first question which is that we'd like to identify how the experiences of the managers and forecasters that we heard about in the first panel connect which agendas and what you have shared today so we can you draw that connection for us and maybe we should start with Julie Roxanne I mean I think kind of an obvious connection so Anna and I were talking about this prior to us kicking off this morning I mean I was saying and you know are we are we all in agreement on what we mean by the term risk communication when we're coming into this workshop because I think so much of what the way that we tend to think to think about risk communication I say that not to diminish it but it's not just the messages that are going out to the public or to the emergency managers but it's also understanding what are the tools that the forecasters have at their fingertips one of the questions I wanted to ask in the previous panel because Alex you are also talking about the rain rates I mean that's a huge predictability problem and to what degree does the research community actually have the capacity to help you understand what the rain can be managed by the different partners and I think kind of this whole big broad connection in terms of everything that we're doing from the research space whether it's you know from the atmospheric sciences or from the social sciences through the forecasters who use all this information as a source of risk information for them to make decisions about the risk and generate new messages for their person. Yep. One thing that I think is really important is recognizing that although we are interdisciplinary we can all benefit the public and I think that it our work has been very much informed by individuals who are in emergency managers and people in the weather community to help us to figure out what kinds of questions we should ask I want to give a specific example I'm conducting research right now in Lake County California which is a community that is at very high risk of wildfires and in fact 60% of the community has burned since 2015 and in our my research teams meetings with the emergency management they are convinced that everybody knows their evacuation zone and so we collect data on a representative sample of almost a thousand people 60% of the sample at the very beginning of the 2023-24 wildfire season said that they did not know their zone and the 40% who said that they did more than half were inaccurate so I think that it is extremely important to understand that what weather whether professionals emergency managers might think might be going on in the community could benefit from research and we are sharing these data with our emergency management peers because I think that they told me in advance that they were certain everybody knew their zone and they were pretty surprised in the same way I think we can go forward helping as researchers the weather and emergency manager communities in asking questions that I think deserve to be asked Thank you Roxanne Emma Yes thank you so I think maybe just I'll add a brief comment to what's already been said but you know I think the other challenges in this particular space is thinking about events that are unfolding in real time and how we as researchers not only you know should we be working on the kind of longitudinal studies that Roxanne and Julie were describing but also can we lend our expertise to what we're seeing going on and this is more prominent in my own domain in the online space so we work very closely with journalists for example to help translate what we're seeing online into news articles and media articles that can come out and so strengthen these kind of partnerships I think is really great but it's really challenging to think about how research might have to operate in a very long time. Excellent we have time for one or two questions in the room Richard I think you raised your hand Chris. Thank you I really wanted to react to some of the comments that Roxanne was making my name is Richard Allen I'm a seismologist first of all from UC Berkeley I just I was really shocked Roxanne by some of the stuff you presented that people don't know whether they're in an evacuation zone and it's very difficult for them to get that kind of information from my perspective there's no reason to be with a smartphone can't get the direct information you're in an evacuation zone you need to do this that and the other in the case of earthquake early warning that's exactly what we do right we have to draw a boundary around the region that's going to experience strong shaking or weak shaking and then we can tell people you know very short simple messages what to do so I guess my question is why don't we have this message to specific locations telling people when they're in an evacuation zone I'll just say I think that that's a really important question and just in Lake County our emergency managers say that they have told people what their evacuation zones are but the evacuation zone is a very complicated set of numbers and letters maybe that's just not maybe we need to look at what Disney does or you know in their parking lot and you know you know a particular Disney character and that's what you remember but it you know I think that we have to think about how individuals process this information and retain it and you know is it a is it something that is a magnet on the refrigerator I'm not sure there was somebody over in the can I just say the whole point of delivering it to a personal cell is that the evacuation zone they only get it on the cell phone if they are in the evacuation we will come back to this for sure there was one more question here and then we need to turn to one or two questions on Slido yes were you pointing to me or sorry my name is Christina Finch from FEMA and I was actually going to build on that just a little bit in that evacuation zones are not the only ones that are available so there is a lot of demographic boundaries whether that be roadways zip codes other relatable items and part of that difference is because they are developed at state and local levels so there isn't a national standard for evacuation zones there's also not a consistently available resource to find all evacuation zones it varies based on your state and how those zones are defined and whether or not that influences if understanding if you know your zone is effective and the availability of those evacuation zone information so whether that's available on your county website as a map a static map or an interactive map or looking at if it's available on your phone or a know your zone application would be interesting questions I'll just say that it's a website during Hurricane Irma that you could put your address in to find out whether you were in the zone or not and I could not figure it out so it you know sometimes it would say B or C and you couldn't really tell for sure so I think that this is beyond the scope of the kind of research that I do but I think it is certainly an extremely important point about I think my research just says I don't necessarily know their zone and then I think it's you know a great deal of information could dive into the specifics of it maybe that's the kind of thing Julie well I think no I was gonna well we do measure also people's perceived perceptions of whether they're in an evacuation zone we've done this for hurricanes and for flooding and for flash flooding and we do similarly fine I couldn't tell you the numbers off the top in my head evacuation zones are important but I also want to go back to the previous panel who talked a lot about how we do need to be thinking about helping people understand the risk from all hazards not just the ones that the evacuation zones are tied to what it looks like for something like Hillary where maybe there isn't an evacuation zone ahead of time for a type of event maybe there's fire evacuation zones but not something similar for rainfall and I think and you know that's the case that traditionally would be an evacuation zone because those are typically defined based on surge right so I think I just want us to also broaden evacuation zone knowledge is really important but how do we again ensure people are aware of all the hazards that might affect them in a changing climate Alex talked about the tales and beautifully articulated how do we map those impacts for some of these extreme events or these atypical events when we might not I like that thinking beyond the zone let's take one should we take one question from online and then we will break for lunch Thank you directed to the entire panel so far the research perspectives have highlighted limitations and how they address those using longitudinal pre-hazard data collection what comes next what types of work are needed next to continue improving risk communication Emma would you like to start I think at least from my perspective again focused on some of the online and digital platforms I think we need to be thinking about the best ways to across platforms be able to share some of this information recognizing that it not doesn't always have to come from official sources but official sources are great seeds that then can be spread by the communities and so for me I think the thing that we're thinking about next is how can we study how information moves across platforms and gets changed in that process and then also how can we leverage some of these AI tools for good so how can we support the collective sense-making process using these kind of new technologies Emma are you Rossi yes I would like to point out that it is extremely difficult to get the kinds of pre disaster studies that Julie and I talked about into the field and I want to just comment that it requires funding and I have to say the National Science Foundation was extremely helpful in that regard but with my studies are extremely expensive and without pre disaster funding that it's really tough and then also we have to concern ourselves with the institutional review board it requires very very quick assessment and approval of our projects so the kind of work that we're talking about has not been done by many people not because others haven't had the same idea but maybe they just weren't lucky enough to have a cooperative IRB or get the funding in time and maybe I'll just underscore that and kind of characterize it in a complimentary way and in atmospheric science we absorb the atmosphere all the time right we constantly study one or two hurricanes and say we understand how all hurricanes work so we need to be thinking about collecting this kind of social science observational data on an ongoing basis and I just want to really bring in what Emma said too and what Roxy also articulated is maybe continue to facilitate the mechanisms to transfer the knowledge that we're learning back to the operational community in real time I love this or in near real time so if we're collecting this data and we're learning about other service who are media partners so that they might change their messaging accordingly kind of operationalizing our risk communication so and on that operational note we will break for lunch and please be back a minute or two before we start at 1245