Smerst2013 crisci warwick

Social Media and severe weather events: mapping the footprint

Social Media and Semantic Technologies

in Emergency Response

University of Warwick Coventry

15-16 April 2013

Alfonso Crisci - a.crisci@ibimet.cnr.it

Valentina Grasso - grasso@lamma.rete.toscana.it

Social Media

Weatherversus

Weather severe events

Where

WhoWhen

as emergency issue in spacetime imply a WWW

…..as a reality Web

aware & prepared

Towards resilient communities means to be ctizen

Changing climate means changing awareness

Imply the reframing in:

Prepardness & Response

Geographical spreading and magnitude of events are important for awareness

Social media and SEO are the information web rivers available.

Are they useful or not?That is the question ( W. Shakespeare).

A question of time event shape

start

peak

decline

weather phenomena and social/communication streams as "analogue" time delayed information waves

time

…..and geography

real physical process

& information flows

Local dynamic type warping means to be explore the

Time coherence between

[ or its mathematical representation!!!!]

In a multidimesional space or better in every time-varying systems ( as the atmosphere or as the “WEB information seas” ) some structures ever could be detected.

Uncovering the Lagrangian Skeleton of TurbulenceMarthur et al.Phys Rev Lett. 2007 Apr 6;98(14):144502. Epub 2007 Apr 4.

Lagrangian coherent structures (LCS)well known in ecology and fluid dynamics

When two or more time-varying systems are connected a supercoherence could be detected if processes are linked.

The link structure between SM and weather could be done hypothetically by a opportune Hierarchy model (Theory of middle-number systems Weinberg 1975). Social media and weather relationships are surely an Organized Complexity.Many parts to be deterministically predicted, too few to be statistically forecasted.

Agent-Based Modeling of Complex Spatial Systemshttp://www.ncgia.ucsb.edu/projects/abmcss/ May Yuan, University of Oklahoma

To overcome this kind of complexities a 5-point :

road map

• Identify a 1-dimensional time flux of information from SM’s world

• Detection of every local statistical linear association of this one in a parametric –physical- spacetime representation ( time spatial grid of data).

• Mapping the significance in classes previously determined.

• Pattern verification with observations.

• Semantics and textual mining confirms.

Heat wave as a good case

severe weather eventEmergency as consequence of "behaviour“.

Awareness is linked to “perception”.

Weather event: early heat wave on 5-7 April 2011

• investigate time/space coherence between the event extension and its social footprint on Twitter

• semantic analysis of Twitter stream on/off peaks days

Research objectives

Severe weather definitionHeat wave: it's a period with persistent T° above the seasonal mean. Local definition depends by regional climatic context.

Severe weather refers to any dangerous

meteorological phenomena with the

potential to cause damage, serious social disruption, or loss of human life.[WMO]

Types of severe weather phenomena

vary, depending on the latitude, altitude, topography, and

atmospheric conditions. Ref:

http://en.wikipedia.org/wiki/Severe_weather

Target and Products Consorzio LaMMA - CNR Ibimet developed a methodology and a set

of products to quantitative evaluate the social impact of weather related events.

Stakeholders: • forecasters

• institutional stakeholders

• EM communities

• media agents

Products: • DNKT metric

• association of the time vector (DNKT) and a time coupled gridded data stack

• spatial associative map

• semantic analysis Twitter stream:

- clustering

- word clouds

Detect areas where it's worth focusing attention, also for communication purpose.

Target

Data usedHeat wave period considered (7-13 April 2011)Social

- Using Twitter API key-tagged (CALDO-AFA-SETE) 6069 tweets collected through geosearch service for italian area.

- Retweets and replies included (full volume stream)

Climate & Weather (7-10 April 2011)

- Urban daily maximum T° - Daily gridded data (lon 5-20 W lat 35-50)

WRF-ARW model T°max daily data (box 9km)

Twitter metric

DNKT shows time coherence with daily profiles of areal averaged temperature

*Critical days identified as numerical neighbour of peaks (7-8-9-April): social "heaty days"

DNKT - "daily number of key-tagged tweets"

*

**

Geographic associative maps

Semantic based social stream in 1D * time space (DNKT)

Weather informative layers in 2D time* space

LinearAssociation Statisticallybased Verifierby pixel

Geographic Associative Map (2D space)

Impacted areas

It's a weather map at X-rays: Twitter stream is used as a "contrast medium"to visualize impacted areas.

This is not a Twitter map

Associative maps patterns fits

Urban maximum T° over 28 C° on 9 April

where & when

Semantic analysis

- Corpus creationDNKT classification by heat-wave peak days:

heat days ( 7-8-9 April) no-heat days (6-10-11 April).

- Terms Word Clouds (min wd frequency>30)

heat days vs no-heat days

Clustering associated terms

Term frequency ranking comparison

- Hashtag Word Clouds heat days vs no-heat days

R Stat 15.2 Packages used: tm (Feinerer and Hornik, 2012) & wordcloud (Fellows , 2012)

heat days

terms WordClouds (excluded key-tag

caldo-afa-sete)

heat days no-heat days

Terms association clustering

heat days no heat days

"heat" is THE conversation topic "heat" is marginal to the conversation topic

heat days

Terms frequency ranking

no heat N=2608 heat N=3461

oggi 6.0% oggi 8.3% 1°

sole 5.5% troppo

7.7% 2°

troppo 4.1% sole 5.9% 3°

Hashtags WordCloudsheat days no-heat days

On peak days:

- widening of lexical base during "heat critical days" - heat as a conversation topic

- ranking of terms (i.e.:adjectives as "troppo"!) is useful to detect change in communication during climatic stress

- geographic names appears in terms and hashtags wordsets ("#milano" !).

This fits with recent advances on "social media contribution to situational awareness during emergencies".

Semantic: some results

Snow events

SNA of keytagged social media streams

Begin 10 feb 2013

End 11 feb 2013

The Graph metrics of SM streams are dynamics.

The graph centrality analisys of Media and Istitutions may provide very useful parametersforWeather Event follow-up.

#firenzeneve

conclusions- Methodology for a social "x-

rays" of a weather event: semantic social media stream as a "contrast medium" to understand the social impact of severe weather events

- Methodology social geosensing is able to map severe weather impacts and overcome the weakening in geolocation of social messages and eliminate the bias due to "social fakes".Weather as a key emergency context where it's worth working

on community resilience - also with the help of social insightful contents.

Reproducible R code

Github Master class socialsensing Code & Data

https://github.com/alfcrisci/socialgeosensing.git

Wiki Recipes in

https://github.com/alfcrisci/socialgeosensing/wiki

#nowquestions(slowly please if is possible)

www.lamma.rete.toscana.itwww.ibimet.cnr.it

#thanksContacts:Alfonso Crisci & Valentina Grassomail: grasso@lamma.rete.toscana.it a.crisci@ibimet.cnr.it

Twitter: @valenitna @alfcrisci

Code and data Alfonso Crisci alfcrisci@gmail.com

www.lamma.rete.toscana.itwww.ibimet.cnr.it