Talk & transcript
Climate Hackathon
NYC Open Data
Transcript
Download textAutomatically generated transcript. May contain errors. Select a timestamp to watch that moment on YouTube.
a few strategies like increasing the can cover the diversity of the species of the trees that are planted and the health of the trees increasing soil moisture quality through composting and distribution of that compost tree bets this is application for specifically for tree uh sorry Street trees in New York we did a little research found that 90% of the the city are not sweet trees meaning not the ones that are on the corner and the sidewalk but that means they must be in Parks so um when we're talking about maintaining um trees around the street that's one thing but that's not to miss
the bigger picture which is massive quantity of trees that are absorbing carbon and doing collusion are going to be in your concentrated areas your par so um one thing you talked about is not expecting trees to solve all the problems of plant of car and and pollution that are happening when a lot of addressing the problems going to be at the source one example would be development so one of the U problems in the city specifically New York is uh warehouseing new apartments um new development being used as a way to address uh problems of um housing shortages when we have plenty of housing
in giant apartment buildings that are are just sitting empty because of a very Twisted Inc me it building owners uh are incentivized to keep those uh Apartments off the market we're building new buildings and we have plenty of Housing and available buildings that have already been built so we can't expect trees to solve carbon footprint problems of that are the result of these twisted uh incentive model so address the root Source uh for the tree volunteer programs a lot of the data you probably see if you looked at the tree data is entered by volunteers thee that was done 2015 a lot of that is people everyday people just downloading an app
volunteering to go U measure trees in their neighborhood and going out and saying I see a tree I'm going to take a picture of it I'm going to enter the data and little by little that's a lot of the um entries in the 700,000 records of tree uh Street trees that are in that massive public data set so increasing the tree volunteership uh and stewardship programs uh we've all noticed that trees uh that are planted on the street may or may not survive so they may have a barod and a tag means they were planted pretty recently but you may be looking at a dead tree and that may be the second or third iteration for that particular tree bed that means there's a stewardship problem that means the person the ownership of who's supposed to be taking care of that
tree is either not clear or not forc so you got to improve that um direct City resources to concentrations of the trees in the Parks it may be that it's better to direct a particular resource towards maintaining the he of a gigantic old tree and a park or a cemetery for example versus using those same resources to plant a new tree which may or may not survive due to the stewardship problem and tree data sets should be public and updated New York's got a great um uh the the tree data that exists now is a great uh model to begin with but we also noticed that the the dates on the um data set were all from
2015 2016 meaning that's when the original sensus was taken unclear when we look at like let's say something that should be updated like the diameter of that tree today if that entry in the database is um from 2015 or if that's from today that should be clear in the data set too that's something that could be improved um other things in the data set like being able to understand from The public's point of view what the carbon sequestration capacity is for particular tree species in particular soil types in particular uh stage of grow that would be all very valuable I think for the public to be able to look at last thing I'll mention is we looked at just as a comparison um for what is
the capacity of trees to sequester carbon uh in versus the whole carbon footprint of New York City it's main ERS it's at least a couple order magnitude in terms of difference in other words the capacity for treas to sequester carbon is less than 1% of the carbon Li of the city that's with a city that's done a pretty good job of planting new trees so we can't have outside expectations in terms of what trees can be done but we want to make sure that that program is still successful and that's
our ra your a tree we want to go like counterwise
so we did a lot just I guess to first show off what we did um we kind of generated a map just using the data and the long latitude to show off like um where the trees are and you can zoom in and it shows where all the trees are um the data does not include Park trees so that's why Central Park doesn't look like have any trees um yeah and then we also use the NTA codes to kind of map the tree health of um Tre is like by NTA so like kind of sub regions in New York um and also the tree
density so that's kind of what the interactive map shows on a high level but uh more gran and showing like the specific measurements of tre Dy we show that here and generally from our findings one interesting thing was that there are actually no Street trees um around Midtown if you see around here I don't know if everybody can see that I'll try but yeah there are no Street trees um but when you map that against like a population map just from the government or something um actually where the highest pollution is which makes sense
cuz the more touristy areas are around um Midtown uh Time Square area and that's where the highest pollution is and this also where less trees are because you don't have the area to put the trees um and generally from our uh just discussions we kind of want there to be or wanted to create a map of tree density and health to show off those Tree locations and the Ant data and also use that against the upcoming 2025 n tree sensus to kind of extrapolate Trends see where proper allocation for funding of trees and resources should be put for example in these more polluted areas where trees
are lacking um and yeah we didn't get to go too far in depth uh there's a lot more we want to do after the 2025 Census Data comes out though yeah [Applause] [Music] okay hi everyone I'm Ethan from our team and uh Mike was all of us so uh what we did is uh we started with thinking about how we can use time series data so uh every decade is recorded 2005 there's 95
and there's 2015 as well so uh the only thing is the number of fields that were captured every decade have kind of increased right U because the level of complexity um so we went into time series to see okay how is the health of a tree declined um in every decade right uh we kind of hit into a you know wall with that because the tree ID is changed so we said okay let's abstract it one more level and go uh into a neighborhood right so finally the data that we could uh you know kind of analyze and we also got a nice visualization which I'll show you all uh is basically uh for every neighborhood right we kind of got uh what's the average threee Health based on the health that has been captured by the people who recorded this data and
also uh what's the average tree count in that neighborhood so the neighborhoods thankfully have stayed the same uh and they've been very consistent over the last 20 years um so yeah we got some interesting stats um I I I want to call out a couple of things 2005 Battery Park Lower Manhattan and central Harlem had like the highest tree Health um and a good amount of tree um so they were probably the top two precincts to be healthy I would say in terms of tree Health but 2015 it's changed Hudson yards in Chelsea um and it's Upper West Side in terms of the average health and U you know the number of trees so that's an interesting statistic one flaw or like a risk that I may be kind of missing out is 2005 as many trees may
not have been recorded maybe just because of lack of you know Human Resources so that's one thing and then finally uh yeah we kind of put all of that on a visualization I know it's a small screen but basically these are all the different precincts or the neighborhoods right so if you see this is a good sign which is average tree count in every Precinct has increased right uh from 20 2005 to 15 um and if you look at the U health it's kind of Remains the Same depends on Precinct to Precinct so you can really understand okay which Precinct is done well which versus which is not so that's an interesting Insight we got so our whole final idea was to really create a Precinct uh report card so yeah like every Precinct has a Tre report card and it tells you okay yeah it basically
focuses on you know okay what's the performance of your Precinct from the last you know decade and how has it been doing what are the goals you should set you know you spoke about sequestration so all of those things right U how you performing as a neighborhood because people generally associate themselves with their neighborhoods a lot more than actually the burrow itself right and yeah and for Simplicity of this exercise we we just took Manhattan data because the all five BS was a lot of information set but yeah thanks
you can have everybody this is here we wanted to uh not only limit ourselves to the tree data that we had we also wanted to use the other data sets we figured out that in the open NYC open data we have data about the traffic automated tra the traffic which is taken by the automated traffic cameras all around the neighborhood and we also found out that there's a good amount of data related to the air quality index of the entire inv city and we wanted to create a heat map and overlay all the three data sets together and to figure out how the most congested areas in the
city and the least congested areas in the city are impacted the health of the tree re in the city and uh how that is in turn impacting the air quality and this is what our idea was we didn't have much time to figure out the entire data set but we going to talk more about what were what the insides are yeah so uh what we basically did was we first took just the tree data set and how we overl it based on three factors of the three like just one factor and red yellow and red orange anded green or sorry I'm color blind so I just I am I can't really figure out but it's green or and red based on the uh tree health and uh yeah so uh we the major Insight was when
we overlay the tree Health Data along with the traffic data we can see that the areas which have a higher traffic volume passing through throughout the day just like the whole 59 Columbus Circle area it doesn't really have that many trees so that is something that can be improved and secondly Midtown has a lot of trees which are either dead or are in a very bad situation so you might want work on that and that is something so if we get a better we had more time we would also want to Overlay the aqi data set but uh unfortunately the data type was a bit different so it took me some time to figure that out you don't U but as surprising inside which I would want to share here is just by looking at
the data because we couldn't overlay the air quality index on top of the traffic and tree data that we have but this is something which I could see from the data set that we had was uh just near the r fairy rable tram terminal there's there are very few trees lot of traffic and very bad air quality so I mean we see that relation and this is very obvious and intuitive to see that relation but I guess we can work more on the data and we can figure out the exact impact on [Applause] quality which other groups are trees
so that was based on the 2015 tree sensus that means the next tree sensus is next year so can I got a show head who's going to volunteer to be part of the 2025 tree census no no come on come on you know you are there we go come on is that everyone there we go you get a little measuring tape see how many rocks are in the tree Ro pretty exciting rip good time all right raise your hand part of okay let's shall we start yeah we'll start and and the whole group can come up or just certain members depending on what you want should we stand here uh it's
to hi everyone I'm p 2 and we've got Molina and AR soina and we worked on this project but we're very excited to share with you um as you see there's already so much data about the flooding system but I'm just going to share what we wrote Because it's more thoughtful um and and then we're going to the rest of our group each person um some of the key questions is how might we improve the data that the NYC open data team has today for First Responders and New York City residents
it's the first question how might we improve communication channels between First Responders and New York City residents with an evacuation mandate especially and how might we improve communication channels right for those who don't need to evacuate maybe they want to give their home uh to someone to stay and then how do we create better tools and systems to give the people who need to evacuate a place to say so we focus on creating a smarter datadriven approach to emergency evacuation by integrating realtime weather flood and historical data and our Innovations include targeted communication via mobile alerts Geo fencing coupled with
tools that provide live updates on evacuation routes shelter availability Transportation options by leveraging these improvements we can significantly reduce Panic streamline the evacuation process making the entire system more efficient and resilient so our goal is to improve the data to predict through the ml model and better prepare NYC residents so arsh is going to talk a little bit about um what analysis he did but before he does that we going to talk about the data sets um Molina is going to talk the data
sets that are going to improve uh what is the insights that we have yeah so for open data we we're looking at the the flood vulnerability index but I thought it would also be nice to just make the data set more robust because the more data you have the more you can do with it so um we I was looking at a bunch of different data sets that we could potentially use so some of the things that came up were FEMA they actually have a blood insurance rate map so that would be interesting for G companies uh we also have the we had the open NYC ones which were blood vulnerability storm water blood Maps as well as a sea level rise projection so those three can
be definitely used for the NC case and then something that was nice to also incorporate was the NOA hurricane data set so just in case like as the climate worsens we can use hurricane data to try to predict and like see how that'll affect the city as well so those are the DAT that for trying to do the uh trying to make our flood prediction tool and I could give it to AR to talk about the DAT okay so um there some exploratory data analysis yeah for the Jupiter notebook uh made some deductions so this is just
um visualizing the data and like some n values throughout the years and then if we go to like um the dat app that was created it's um it's showing some distributions throughout the fs HRI index and um then we could see like how those projections would be created throughout those years and then I did some correlations and um through that there's a three indexes that there was there was 2020s 2050s and 2080s so uh for for that F FSH valy and through those different indexes of years I create some correlations and this distribution so this is just showing a normal distribution for the fbii so
depending on those storm searches it'll um it'll show like the amount of occurrences that could possibly happen of floods throughout those indexes of years but then uh something interesting was the storm surch vulnerability over time so I don't know if you can see that it's pretty small but there's um coloring so for blue it's the present uh orange is the 2050s and then the green is 2080s so we used the uh we PL visualizations so um comparing the present projections to like projections from like the later years in the data set uh it show realize that there's a lot of missing data which is why it looks like why it looks like this but not only that like it it would still be
lower than the present but um it's like a broader spread of vulnerability levels so throughout time we basically found that it suggests like some future surch risks that would happen later on in the years and um unfortunately we could d D exactly why those would happen but then I made a histogram um for theab index it this represents it but in a different way um unfor du to time I would have just done the work analysis we like developed a machine learning model create an interface to like you know do like a testing training data configuration and try to deduct why exactly those um indexes arise when the years yeah and you and he also mentioned
uh it would be very useful to come up with right better predictions about the data so yeah that was wonderful um we also have some final Solutions just to summarize so in terms of um if you want to just read we're just going to briefly snap through and then move on so just to improve the data set as we are going to share expanding the weather and flood data sources that we have um improving the data visualizations that exists already and automating alerts and improving some of the alert mechanisms whether it's through Amber
Alerts or other possible solutions so if you want to um yeah because we were also talking about like maybe using either an app or web application depending on what's more accessible to other to people our users and so in a similar ve think about social media alerts uh using geofencing important for people who are into area maybe they don't really know how things are so like having that local knowledge right away is very useful and then also like having people needing two-way communication to confirm that you have that people who are evacuating have Lefty
areas and yeah so pretty much um to reiterate something else that came up a lot I just speak to those last tools systems yeah actually I something else so we were also thinking about like API conjunction so through like um using like a weather API for live times like obviously a lot of correlations between like what the current weather is and then also like flooding happening so if you use those both those TOS in conjunction like let's say we created a machine learning model through that data and then we use like an API uh weather model uh you could make more correlations between those and then like you have like interface we could uh like
predict if that flood would happen where would happen depending on like the geographical location and like past uh floodings so yeah that's basically it yeah so yeah thank you so
much thank you
clockwise all right I'm um and so our solution for the was a New York City surge City dashboard and the name is brought to you by sobi Factor like I didn't get too much into it but includes like a lot of um you know data in that number so there's like exposure to the hazard there's a susceptability to harm from the exposure and the capacity toover so one example um if you look at like hospitals and um fire departments there's like less of those um less of that infrastructure in areas that are
more likely to so what's the need um dashbo should include like a notice of if you're at risk of the flood how to get help how to help so volunteering and then financial support so things like vouches to housing and then help with insurance C so we'll go to the dashboard
yeah so this like a search St um so we have like a map that that link so can yeah number here and this emergency help so like um obviously you have an emergency one but the voler Box like uh you kind check on people uh have yeah and there a priority list so like you put your St yeah I think a big part of this is the volunteering
um yeah um as for tree data so that's sort of like the main feature that we want yeah yeah just like how we like Mobil realiz like volunteers should take three centes we think about doing a similar thing for like a FL vulnerability individuals so like the way we envision this dash for is that individuals want to volunteer the community they can go out and like report the data on like all the individuals who might need help in times of vaccination or dur times of emergency could be anyone elderly to anyone with a disability or anyone need accessibility isue um and then they can submit this data onto our dashboard which is not
going to accessible hopefully only two that people give access to in the times of emergency and what this going to do is it's going to help us mobilize as volunteers to do Wellness welfare check on those individuals when there's a time of evacuation when there's a blood warning or any type of like tropical weather coming they can go in and the wellness check on them to the app or to the dashboard and report on them and if they need require any additional help they can basically call 911 or get actual responders on the location to get this guys out into a location and basically yeah and just to das basically how overall coordination
with volunteers like if there's lack of volunteers location we can send out flyers or like notifications to say hey we need more people here can someone go and go do some welfare check and lo sort of like providing infrastructure for like community help with the Hurricanes like a couple weeks ago like all these vogs in in in uh yeah in North Carolina where like most of the help most of the support that's being provided to people need are like there and our neighbors so for those people TR and of course like money that's what really people need at the end of the day so um for for citizens in like high FBI areas um you know checking if they have
insurance getting a deduction on their plan or maybe subsidizing you know their premium and then getting help fing claims or Diss to the insurance compan like financial support so that people who are effective can recover fin overall I feel like I'm volunteering but I feel like the volunteering just like people are saying there's no enough data the volunteering basically can help us impr theity might help census data and all the other public don't [Applause]
so hi everyone I'm cash and um and we looked at the the same SL vulnerability index data set but one one number that we HED in on is the fshr number you may have heard that mentioned what is it exactly we dug into a couple of papers that go into the definition of what that is this is so small you like there's no way you this but FSH stands for flood susceptibility to harm and Recovery index and that takes into account more than just the risk of a flooding but also things like the percentage of elderly population elderly people living alone even things like language isolation so there are a lot of
immigrant communities where people may have not evacuated due to not understanding the warnings um and in very in some of these papers you'll see things like the deaths the residential fatalities happen in areas not where there's the highest R of flooding but where the FSH as inly people living alone and so on are the highest so what we did is we analyze the data set focusing on a couple of these areas so we have um asse so this so this this map shows the I think
a similar visualization to what other teams have been showing this is the FBI for um the various counties or the various whatever the subdivision is called I'm not like americ whatever GE geometrical shape this is um but something that's really interesting that we that we looked into is what are the areas which have the largest change in flood vulnerability from now to 20180 because these are the areas where we think uh less attention is being given and we need to do some work in these areas and we have some more we have some interventions we have in mind that um we're going to talk about um just a last note on some of the
data sets we didn't manage to incorporate so there's the census data which we looked at um which includes things like you know a more detailed look into elderly population and that kind of thing and the population to be of those areas yes um and the storm Events St Bas is something that's been used in all of these papers we looked at um but we didn't manage to incorporate that the storm events will change over time as well tar those so I will hand off to one one of you so you you can go more on so basically that we thought of is that um
when he was talking about like the Immigrant communities in New York city so um a lot of the areas with increased of flood so that would include like areas like U around brighten uh Beach there's like a lot of Russians and uh South Bron there's a lot of uh Hispanic people so we're thinking uh that we could create like um R uh radios and that the radios would have uh translations with um you know Spanish Russian and um all kinds of langu so um it could be like inclusive and um people are more likely to you know listen to something that you know they feel like um safe with and um also another idea that we came up with is
that um um you can increase the amount of First Responders in these areas and um vehicles that could um take seniors cuz um like seniors that you know can't really like help himself so like you know like people wheelchair people um that need help like assist walking and um yeah yeah so going off of that these um people that might need ass staying out of their homes we have IDE of maybe possibly building like flood bunkers in densely populated areas that are vulnerable to floods that can maybe car you know C these people off to um these bunkers or um just getting them there
maybe building a bunker maybe every half mile mile in these densely populated areas um we had were um improving the drainage system in these areas whether it be um increasing the size of the catch basins and giving more time for people to respond just because our Focus was how do we increase the time for people to um get the safety wherever it may be um so increasing the drainer system or making some more pors like roads or um part just because it increases the amount of time for water to drain away for people um and other one other idea we had was
often free emergency training uh flood training protocols just whether it be in the parks on the weekend where you you know incentive people get free birthday to or lunch and then people may um better understand what to do when it comes to flood because um I don't know about you but living in New York we don't get as worth of flood as many other people so I wouldn't know what the first reaction might be to do in response to like really bad um and figuring out and getting these training free access to like people would I think help uh understand how to what to do when play Everything
hi everyone I'm Gabby Dan M and what we focused on with flooding to kind of explain our thought process is like what is one of the biggest issues with flooding and it's like Transportation how do people get to safety get out of the area we considered different topics like commercial vehicle transport or like taxi zones or other areas but I think we all know like the subway is one of the crucial pieces that people use in New York City and we've all seen those videos on Tik Tok of like the flooding that's happening in the Subways and has really old infrastructure so we were thinking like what can we do around that problem um and we kind of learned about like from Sandy and Ida like that is a
key kind of point of you know issue and yeah that so one thing that I saw when researching Hurricane Sandy in 2012 was that there was a lot of problem with Subway they actually spent $5 billion like waterproofing and improving resiliency for things to be able to handle water in the subway systems and then during Hurricane Ida there was more flooding and I think they were shorted up in a lot of ways but they had so much water on the street level that was actively pouring in so even if the Subways are like resilient to water from you know underground or whatever it's like
literally pouring in from the top you can only do so much uh that's the what we learned about the resiliency and what was done in the past yeah and and completely changing the infrastructure is really cost costly so we're thinking of okay what are methods that are a bit easier to integrate less costly so I can go over one you go over the other one um so we were proposing more of like a real time integrated adjustable water pump system because you know a flooding based on the vulnerability index affects different areas differently of course so um how can we help pumps that are you know severely affected kind of mitigate and pump out water versus other pumps
that are not as you know affected so there are a bunch of startups in this space one of them we would recommend is called OPI and they kind of use real time weather data to indicate okay which pumps are going to be affected at what time so we can kind of tell which ones to work harder at What stages and this also kind of reduces you know energy usage if we're smarter about kind of like the pump infrastruct ructure that definitely kind of helps out with the process and it's not necessarily you know adding in brand new pumps it's more of kind of using the iot kind of infrastructure um to make pumps more effective yeah uh so along with uh installing all of these pumps we also
recommend and again uh the other team just touched upon this um we recommend having around these key entrances uh permeable payments um and grein Gardens which which will help absorb a lot of rain water uh before it reaches to these pumps uh in a way it helps to reduce the workload on these pumps and help manage flat fors uh quickly uh another thing what we recommend is underground retention PS uh beneath these green spaces and uh the permeable payments that I just mentioned right now um this could temporarily store run off water and uh give pumps some time to clear uh
excess water uh buing hiding so and our our theory for the mechanism by which this would be enacted is that uh New York government groups who are maybe dealing with hey we have this solution but it's really expensive um can create a call for startups where companies that are just starting out or companies that are trying to create novel Technologies around this area can go and build things that are relevant for them so I think it's just them voicing what they want and what parameters matter to to them for them to go buy a system in the future and hopefully those startups can dramatically you know reduce the cost or improve the efficacy of a solution uh and start to make it more palatable for
a government scaled up
investment let's go that [Music] and can I have all the other flood groups raise their hands if you haven't gone yet got this the last one all
right G and so uh our solution um um is kind of smaller but I think it's uh easily implementable um we want to help with evacuation in in um uh case of floods and help emergency response both with pre pre-event with with education uh during event to help emergency responders and people can CST flood help uh um supplies cover uh so the data um the data um that we were looking at is
vulnerability index and as I think one group behind mentioned that the main index that they use is this FH FSH R it's an index that already rols uh physical flood data with social social uh s of uh factors in a sense we we want to enhance the data because sometimes in this is are good we think for evacuation that's probably not the best uh uh way we actually want to use the tool data set so so we actually want to use the physical indices that they also have
with the population data uh which they have uh but strangely enough they use different indices so we want to do as we're doing it we want to enhance uh uh uh newor open data with providing better geographical indices because they use some indices in one data set and some in the other so so like normaliz act yes so some people use addresses or the names of the neighborhood some people use standard uh uh uh getial uh uh polygrams um so normalize it so integrate the data um and typically we want to as the first
thing we want to use population data set so U this is the map as it looks now we want to integrated with the uh population data set uh so so we'll basically have a joint distribution between the index and population which is number of people and it's also broken up by age and uh uh first language um so so that will create a static data set that basically can be calculate and then it can be used um um um yeah so it can be used uh to plan evacuations so if you're want to send
buses um and and want to plan evacuation how do we optimize what where should we put evacuation centers to get the maximum population that's what that's a vulnerable so that's an optimization like a simple optimization problem um since stting data it's also can be used uh in planning evacuation routes because you know where those buses will be coming from so you can optimize um sort of movement of the popul uh compared to how Google Maps uh optimize the traffic you can actually optimize movement of the bassive population so so you know from sou Brookland you want to move everyone
at the same time so you don't want to send them all around the same so that's a static data and also static data can be used for Education right so once we we know most vulnerable area we know their age and primary language so as somebody else mentioned we can reach those people and we can provide PRI information in the native language uh what they should be doing and where what would be possible evacuation points and that data this data can now be combined with the dynamic data like very precise weather weather prediction models um that can be used in actual in actual
planning uh and uh it can be used in simulations for First Responders that they can plan for different events and uh so I would just add um a couple of the groups already alluded to this but the data is kind of missing in terms of um its values so if you look at the data um and I'm talking about the um flood vulnerability index data um if you do a drop ni so if you drop the rows that are missing data you come down to about 80
rows versus like 2,000 um so trying to do correlations it it looks great but it doesn't really help so what we we're looking at instead was the um FS the FSH HRI um and doing an injoin with the population information and um suggesting that the the data really has to grow and instead not focus on the other columns but instead focus on the fs HRI um something to note about the FSH though is that um there really isn't any you
know sometimes it's really help to have like qualitative information about why one specific location has a five and all the other ones have a five and we're talking about like let's say one point in Manhattan has a five but you know it doesn't have any values but right next to it you know um the other the other columns are filled in and they have fives and I'm talking about a one to five like a ELC scale kind of thing um it indicates that perhaps the FSH is really high because other locations nearby are really really high and you know something that we talked about as a
group is you know right now for instance with hurricane alen and hurricane um forgetting the name of it but with 10 um hurricane hurricane melon perhaps you live in downtown St Petersburg you're at a really high rate of or really high likelihood of being flooded but someone nearby like two three miles away is not going to be affected as much so nonetheless the FSH is going to be a five it's kind of the same thing um and so we wanted to really emphasize that the FSH um is really important key thing to
look at regardless of what any of the other rows and columns are telling you and we need to have an inner join um with the population uh data that um is based on the geography of New York City and Alex you want us to add anything yeah um actually something that you mentioned is um the definition of sech that you brought up that you're actually comparing you know this sege data to other reports and it's like what are we actually comparing so I think something it's just going back to as we par the data having standard definitions for like when we talk about SE what is sege we measuring Sur that you know it's like apples to apples in terms of looking at
and something else I think we talked about was uh looking at I mean other groups mentioned this but using languages in in particular areas to make sure that messaging about um you know evacuations that we're covering the languages but something I thought of and I don't know if this is currently what happens when we get messages um it's actually storing um you know you have a a link to a message or something um but you have multiple translations stored within that so you can you know translate the content to whatever language that you speak so not sure what is the current status quo on that but that would be interesting is to having kind just a better content management system for um you know these kind of
emergency notifications and making sure that all the translations are covered one one thing I want to quickly add is so during s a lot of evacuations were actually done on or not evacuation but a lot of the PO was done on basis soot ofs because Russian BR was affected um so so that you know knowing that uh knowing that information could help in recovery because you can organize basically W help through
specific EIC groups because they you know they will be going [Applause] that's everything um I hope everybody enjoyed their time glad you're all down to hang out and build stuff um if anybody is particularly interested in NYC open data whether it's these data sets or something else uh who wants to get involved with like submitting something uh for the March like some kind of a demo or exhibition in March raise your hand if you're like down okay we should
like get all of your names and like create uh some kind of a chat so we can figure out like who's in um so we'll like have a list like over there and uh make sure you get your name down if you want to like get involved with that I think that'd be pretty cool they said the deadline is November 1st to submit a thing but I talked to the people at NYC open data and they said we could probably fck the deadline a little bit so like we'll we'll make something work in November probably but yeah that's the that's the goal um the next climate hackathon we're talking about one in late November to close out the year and then there's one really cool one planned for January um so join the WhatsApp if
you want to stay involed um and yeah we're just continuing on let's uh take a giant group picture uh and we'll take it on these uh steps on the other side
there is some help from