What's New in Teams Calling, Meetings and Devices March 2024
Social+media+in+the+2011+victorian+floods
1.
Social Media in the 2011 Victorian Floods
June 2011
Sponsored by:
2. Contents
1. Executive Summary ..................................................................................................1
Executive Summary Continued... ..................................................................................2
Executive Summary Continued... ..................................................................................3
2. Background to the Floods ........................................................................................4
3. Background to the Review of the Use of Social Media in Flood Events ..................5
4. Objectives of the Review..........................................................................................6
5. Methodology ............................................................................................................7
5.1 Methodology Summarised ......................................................................................................7
5.2 Detailed Methodology............................................................................................................7
6. Classification of comments ....................................................................................10
6.1 Some Examples of Comments from Different Sorts of Message Spreaders..........................11
7. Profile of the Event from SES Information Releases ..............................................12
7.1 SES Message Out .................................................................................................................12
7.2 The Response to the Floods in Social Media .......................................................................13
8. Nature of the Commentary ....................................................................................15
8.1 Examples of Personal Comments.........................................................................................16
8.2 Content of Commentary .......................................................................................................18
8.3 What’s Missing.....................................................................................................................21
8.4 Victoria Police Tweeting......................................................................................................22
9. Learnings for Future Social Media Monitoring ......................................................25
9.1 Some Guides for Monitoring ................................................................................................25
9.2 Detailed Content Analysis....................................................................................................25
10. Conclusions...........................................................................................................27
10.1 Communications Models ......................................................................................................28
11. Suggestions for Future Work................................................................................30
Appendix 1: Locations Covered...................................................................................31
Appendix 2: List Of Top Social Media Accounts..........................................................32
Appendix 3: Sample Of Community Facebook Pages (Public Pages)..........................35
3. 1. Executive Summary
Introduction
This report analyses social media commentary about the Victorian floods from 10
January to 4 March 2011.
In this period there were two floods – one in rural Victoria (January) and one in
Melbourne and surrounds (February).
Objectives
Wide ranging objectives were established for the review:
• Document social media mentions during Victorian Floods
• Analyse comment by location and other characteristics
• Ascertain the nature of comments
• Establish flows of information
• Establish the nature of sharing of warnings and other information
• Explore the database for other useful insights
• Recommend approaches for future events
Methodology
Social media comment for the period 10 January to 4 March 2011 was collated using
social media monitoring provider Buzz Numbers, and Twitter and Facebook feeds
from Rowfeeder.com. The data was then analysed by Alliance Strategic Research.
Data was cleaned from over 320,000 comments to those relevant to the Victorian
Floods with an end number of 12,405 comments analysed. Comments were
classified in a way to understand the nature of the comment. The classification was:
‐ People in the Floods
‐ Official Posts
‐ News
‐ Message Spreaders ‐ Warnings
‐ Spectators/Commentary
‐ Message Spreaders ‐ Relief/Donations
‐ Message Spreaders ‐ News (story about floods)
‐ Message Spreaders ‐ News (story not about floods)
‐ Message Spreaders ‐ Information
‐ Recovery/Post‐flood issues
‐ Social media
‐ Commentary ‐ not directly about floods (but including reference to
floods)
1
4.
Executive Summary Continued...
Key Findings
• Social media has been used to comment about the Victorian floods – much
of it stemming from news sources
• Data captured for this analysis possibly represents a fraction of the total
commentary. The data collated is entirely dependent on key terms and unless
these are used, the comment is not captured
• During the floods SES information releases strongly focused on warnings
• Different social media channels have been used for different types of
communications and at different times in the events
• Social media comment and Twitter in particular has the capacity to be highly
reactive, and possibly even “ahead of the news”
• Comment via Twitter was most frequent. News services also use Twitter to
promote stories.
• There is evident willingness to spread information about crisis events –
including fundraising
• Well wishers and ‘spectators’ also have a reasonably high profile in the
commentary
• Commentary drops away quickly after the events. Recovery and community
commentary declines also
• Although news commentary drops away after the event, this source persists
for longest
• Commentary in metropolitan areas was more personal, and there was a lot
more volume in a short period of time
• Discussion of recovery issues as a share of comment increases after the event,
but is at low levels overall
• Specific locations get mentioned frequently, larger centres more so than
smaller. The duration of inundation will also impact the comment for specific
areas
• The nature of the comment is generally helpful and positive in its nature. The
ironic or cynical comment evident in many other topics in social media is not
evident in the comment about floods
• As the crisis is occurring there is commentary, as it abates so to does all
comment
• Each and every event is likely to have a different profile
• The communications task will become more complex as one too many
broadcasts fracture to become one to one message spreading, not only
through known sources but individual to individual
2
5.
Executive Summary Continued...
Suggestions for future work
• Successfully monitoring future events in social media requires comprehensive
data collation as the events unfold
– Established relationship with a social media tracking software provider
is the most time effective way to collate information
– Understanding the limitations of data tracks is important
– Key terms need to be identified early and modified as needed
– Key publishers established and followed
– A human resource required to monitor and modify
– Establishing hash tags early and monitoring if they change is essential
to track activity
• Pre establishing tracking terms for categories of events (fire, flood) and
setting them up in a monitoring services ahead of events occurring will
facilitate start‐up of monitoring as the event breaks.
– Detailed analysis content analysis is painstaking, so planning for that
analysis is also recommended
– This will include planning of key words to be tracked and specific
research questions
• Consider using SM aggregation tools for publication back into the SM stream
– Eg. Paper.li
• Many mobile phones are SM enabled (Facebook and Twitter) and may be the
last way people in a disaster have to communicate (as demonstrated after
Yasi). As dependence on mobile technology increases the need to respond is
likely to grow
• Apart from disaster communication there appears to be considerable scope to
tap the willingness of people to help and contribute. Some targeted recovery
promotion and activity may fall on fertile ground in social media.
3
6. 2. Background to the Floods
In January and February 2011 Victorian experienced two flood events.
January
The January floods were widespread encompassing large areas of western and north
western Victoria. Other areas of the state also experienced flooding at the same
time, but in a more localised way.
The January floods were regional and rural in nature affecting sparsely populated
farm land, and regional towns. This flooding was, for most part, from rising rivers.
Locations anticipated to flood, were advised in advance.
The floods occurred between 10 – 20 January and some areas were inundated for
some time.
February
The February floods were largely in densely populated urban areas in and around
Melbourne.
The flooding was more sudden than the January floods, and resulted from torrential
rain.
The floods occurred 4 – 5 February 2011.
4
7.
3. Background to the Review of the Use of
Social Media in Flood Events
Online and social media outlets were used extensively during the Victorian and
Queensland floods by media, services, communities affected by the floods, and by
other concerned and interested people.
Social media use adds a different dimension to emergency communications. The
traditional communications model of announcements being conveyed through
known media sources can potentially be circumvented with people getting their
information from each other rather than official sources.
Understanding the way in which social media was used in the Victorian floods was
thought an important starting point in developing policy and approaches to the use
of social media in future emergencies.
Research was undertaken to inform social media strategies for all Victorian
Government agencies including VICSES.
5
8.
4. Objectives of the Review
Broad ranging objectives for the review were established and are summarised as
follows:
• Document social media mentions during Victorian Floods
• Analyse comment by location and other characteristics
• Ascertain the nature of comments
• Establish flows of information
• Establish the nature of sharing of warnings and other information
• Explore the database for other useful insights
• Recommend approaches for future events
Some of these objectives have been more fully answered than others.
6
12. 6. Classification of comments
Each comment was read by one reader (thereby ensuring consistency in
classification) and one of the following classifications was attributed to each
comment.
People in the Floods
People experiencing floods (incl pics/videos & RTs2 of these)
Official Posts
Emergency Services & Government Departments and Agencies
News
News outlets/sources as defined by Buzz Media
Message Spreaders ‐ Warnings
RTs of official and unofficial warnings (inc road closures)
Spectators/Commentary
Spectators, well‐wishers, commentary about floods, observations about floods,
any comment about floods from individuals (inc a lot of blogs)
Message Spreaders ‐ Relief/Donations
Information and RTs about donations/concerts/fundraising/help available to
victims
Eg: Lifeline/mental health
Message Spreaders ‐ News (story about floods)
RTs and relayers of news articles about the floods
Flood updates (put out by news outlet), news articles,
Message Spreaders ‐ News (story not about floods)
RTs and relayers of news articles that just mention/make reference to the vic
floods (but the article is not about the floods)
Eg. Articles about global disasters/climate change
Message Spreaders ‐ Information
RTs and information about floods
Eg. flood maps, relief centres & meetings, SES volunteer and other phone
numbers
Recovery/Post‐flood issues
Economic implications, inc for various industries, house & food prices,
government response to floods ($), political fallout
Social media
References to social media's role in communicating the floods; "twitter talk" ‐ the
frequent message spreaders saying goodnight to each other, etc
Commentary ‐ not directly about floods
People (personal and organisations, but not news outlets) who make reference
to the floods but are not talking about the floods specifically, inc (a lot of) blogs
2
RT = Retweets. Another’s message is passed on by using the ‘retweet’ function in Twitter. Eg. If I follow
@Victorianpolice and get one of their messages I can retweet it and all my followers will receive the message
(although they do not follow Victorianpolice). It effectively multiplies the message.
10
13. Classifying the data in other ways and along themes is possible, but have not been
explored further given the budget of this study.
6.1 Some Examples of Comments from Different Sorts of
Message Spreaders3
• “come on queenslanders please give generously to the victorian flood
appeal!! “ (Facebook) Message Spreaders ‐ Relief/Donations
• “A flood watch is current for the greater Melbourne catchments of Werribee,
Maribyrnong, Yarra, Dandenong and Bunyip.” (Facebook) Message Spreaders
– Warnings
• “RT @VictoriaPolice: Police aware of flooding in Keysborough, Hampton Park,
Dandenong and Narre Warren areas. Pls don't ring triple 000 ...” (Twitter)
Message Spreaders ‐ Information
• “Want to help cleaning up the vic floods. Anyone looking for a small group of
friends to come and help clean this weekend? #vicfloods” (Twitter)
Recovery/Post‐flood issues
• “Australia's top musical talent to perform for #vicflood relief tomorrow night
in #Melbourne.. How good!! http://bit.ly/xxxxxx “ (Twitter) Message
Spreaders ‐ Relief/Donations
• “After the devasting floods in Queensland and Victoria many organisations
companies and crafters are helping out with the relief effort. Many craft sites
are auctioning goods and services with the proceeds going to the flood
appeal. Here are some of the sites: ...“ (Blog) Message Spreaders ‐
Relief/Donations
• “Australia warns of fresh floods: Residents in the flood‐hit city of Brisbane
brace for a king tide as more towns in Victoria are eva...” (Twitter) Message
Spreaders ‐ News
3
Quotes are as written by author. Twitter identities have been concealed
11
14.
10
20
30
40
50
60
70
0
7.1
0
10
20
30
40
50
60
70
11/01/2011
10‐Jan
12/01/2011
11‐Jan
13/01/2011 12‐Jan
14/01/2011 13‐Jan
15/01/2011 14‐Jan
16/01/2011 15‐Jan
17/01/2011 16‐Jan
17‐Jan
(or unspecified).
18/01/2011
18‐Jan
19/01/2011
19‐Jan
20/01/2011
20‐Jan
21/01/2011 21‐Jan
22/01/2011 22‐Jan
releases on a daily basis.
23/01/2011 23‐Jan
24/01/2011 24‐Jan
25/01/2011 25‐Jan
Major warnings
26/01/2011
26‐Jan
27‐Jan
27/01/2011
SES Message Out
28‐Jan
28/01/2011
29‐Jan
29/01/2011 30‐Jan
30/01/2011 31‐Jan
31/01/2011 1‐Feb
Warnings
1/02/2011 2‐Feb
2/02/2011 3‐Feb
3/02/2011
4‐Feb
5‐Feb
4/02/2011
6‐Feb
5/02/2011
7‐Feb
Per day
6/02/2011 8‐Feb
Information Releases
7/02/2011 9‐Feb
8/02/2011 10‐Feb
9/02/2011 11‐Feb
10/02/2011 12‐Feb
13‐Feb
Community/Advices
11/02/2011
Number of SES Releases
14‐Feb
12/02/2011
15‐Feb
13/02/2011
16‐Feb
14/02/2011 17‐Feb
15/02/2011 18‐Feb
16/02/2011 19‐Feb
Nature of SES Releases Over Time
Bulletins
20‐Feb
7. Profile of the Event from SES
17/02/2011
18/02/2011 21‐Feb
19/02/2011
22‐Feb
23‐Feb
20/02/2011
24‐Feb
Misc
21/02/2011
25‐Feb
Total of 653 releases over the period 10 Jan to 4 March
22/02/2011 26‐Feb
23/02/2011 27‐Feb
24/02/2011 28‐Feb
25/02/2011 1‐Mar
26/02/2011 2‐Mar
27/02/2011 3‐Mar
4‐Mar
the course of the events. The following graph charts the number of information
28/02/2011
Information releases in relation to the floods for warnings major/minor flooding,
1/03/2011 The nature of the message shifted over time in accordance with the floods as they
2/03/2011
unfolded. The main message though, is ‘warnings’, either major or moderate/minor
12
evacuations, updates, bulletins, emergency alerts, community news were issued over
15. 7.2 The Response to the Floods in Social Media
The total number of comments on Victorian Floods collated over the review period
was 12,405.
The social media comment peaks and drops away in line with the course of the flood
events. When something is happening there are comments, then it drops away. We
also observe this much more acutely in other social media tracking where an event
will be very short lived in social media.
The social media comment is not necessarily directly informed by the SES information
releases. Social media commentary was slow to pick up on the Victorian Floods in
January (due possibly to media focus on the Queensland Floods) and was ahead of
SES information releases in February.
Comments per Day Compared to
1000
Number of SES Releases 70
895 (All Sources)
60
800
50
654
600 40
400 30
20
200
146 10
0 0
10/1/11
11/1/11
12/1/11
13/1/11
14/1/11
15/1/11
16/1/11
17/1/11
18/1/11
19/1/11
20/1/11
21/1/11
22/1/11
23/1/11
24/1/11
25/1/11
26/1/11
27/1/11
28/1/11
29/1/11
30/1/11
31/1/11
10/2/11
11/2/11
12/2/11
13/2/11
14/2/11
15/2/11
16/2/11
17/2/11
18/2/11
19/2/11
20/2/11
21/2/11
22/2/11
23/2/11
24/2/11
25/2/11
26/2/11
27/2/11
28/2/11
1/2/11
2/2/11
3/2/11
4/2/11
5/2/11
6/2/11
7/2/11
8/2/11
9/2/11
1/3/11
2/3/11
3/3/11
4/3/11
SES Releases Total
The different nature and location of two flooding events may in part account for the
different profile. The February event was in Melbourne with a higher density of
people, and was a sudden flooding event. The peak includes people involved in the
event. The January event was in sparsely populated areas and was not sudden.
13
16. The majority of comment collated is generated in Twitter ‐ Blogs have a surprisingly
high profile, but blogs from news sources underpin these results (to some extent).
Source of Comment
Communication Channel
Twitter 6975
News 2490
Blogosphere 1959
Facebook 722
Forums 197
Video 62
0 1000 2000 3000 4000 5000 6000 7000 8000
Number of comments
The following analysis shows Twitter comment is event sensitive rising and falling
quickly – this is also evident in other categories Alliance Strategic Research tracks.
Blogs and news are more persistent sources of commentary. Facebook peaked in the
Melbourne event, but did not feature strongly in the regional floods.
700
Channel Over Time
600
500
400
300
200
100
0
1/02/2011
2/02/2011
3/02/2011
4/02/2011
5/02/2011
6/02/2011
7/02/2011
8/02/2011
9/02/2011
1/03/2011
2/03/2011
3/03/2011
4/03/2011
10/01/2011
11/01/2011
12/01/2011
13/01/2011
14/01/2011
15/01/2011
16/01/2011
17/01/2011
18/01/2011
19/01/2011
20/01/2011
21/01/2011
22/01/2011
23/01/2011
24/01/2011
25/01/2011
26/01/2011
27/01/2011
28/01/2011
29/01/2011
30/01/2011
31/01/2011
10/02/2011
11/02/2011
12/02/2011
13/02/2011
14/02/2011
15/02/2011
16/02/2011
17/02/2011
18/02/2011
19/02/2011
20/02/2011
21/02/2011
22/02/2011
23/02/2011
24/02/2011
25/02/2011
26/02/2011
27/02/2011
28/02/2011
twitter News Blogosphere facebook Forums Video
Unlike Twitter not all Facebook comments are public, so this chart cannot represent
the full extent of Facebook commentary which is unknowable.
As the event recedes, news sources continue to publish stories (albeit in smaller
number) whereas the citizen journalist stops altogether. The public it appears is
essentially responding to the event as it affects them at the time.
14
17.
8. Nature of the Commentary
A large proportion of the social media commentary is news. News can be the
headline for press articles or TV stories or online stories. These comments are aiming
to attract readership. In this data set message spreading (of one sort or another) is a
large part of the commentary. Message spreading does not just relate to warnings.
Volume and Type of Comment
News 3553
Message Spreaders ‐ Relief/donations 1983
Message Spreaders ‐ News 1837
Spectators/Commentary 1541
Recovery/Post‐flood issues 1016
Message Spreaders ‐ Warnings 930
Message Spreaders ‐ Information 591
People in the floods 539
Indirect Commentary 177
Social Media 170
Official Posts 68
Over the course of the flooding events, not only does the scale of commentary shift,
so too does the type of commentary.
Intent of Comment Over Time
People in Floods Official Posts News M.S. Warnings Spectators M.S. Donations M.S. News M.S. Info Recovery issues
15
19. • “Massive storms just hit Melbourne! Ex TC Yasi 2500 klms away! In 11 years
never seen my street in flood! Wow!” (Twitter)
• “Just got home flood waters were lapping over the bonnet of the car scary
trip home” (Twitter )
• “Its so different when your part of a flood than watching it on news. I live near
Murray River, but i didn't expect it right outside my house haha NEVER GOING
TO FORGET TODAY!! Lol anyone want to give up there boat for me!? Haha”
(Facebook)
17
20. 8.2 Content of Commentary
The profile of the two floods in terms of comment intent is quite different, with
Melbourne floods attracting more ‘spectators’ and comments from people in flood
situations – in fact these two groups generated more messages than news services.
Type of Comment by Flood Area
News 30%
19%
M.S. Relief 17%
9%
M.S. News 15% Vic Floods Melb Floods
13%
Spectators/Commentary 11%
25%
Post‐flood issues 9%
2%
M.S. Warnings 7%
14%
M.S. Info 5%
1%
People in the flood 3%
17%
Official Posts 1%
0%
Different channels are used in quite distinct ways for different message types.
Twitter is most often used for spreading warnings and information, whereas blogs for
commentary. Facebook also profiled strongly for people involved in the floods, but
the overall number of comments was lower. Video was most often posted by people
in the floods.
Source by Intent of Comment
Message Spreader ‐ Warnings
Message Spreader ‐ Information
Social Media
News
People in the flood
Message Spreader ‐ News
Message Spreader ‐ Relief/donations
Official Posts
Spectator/Commentary
Recovery/Post‐flood issues
Indirect Commentary
Blogosphere Facebook Forums News Twitter Video
Note: Classification is normalised to allow comparison of channels used
18
21. Overall the location which attracted the greatest number of comments was
Melbourne. It appears areas with larger populations attract more comment.
Areas which were affected for longer periods such as Swan Hill also have more
comment.
Location No. of Mentions
Melbourne 1193
Swan Hill 544
Kerang 432
Horsham 234
Charlton 206
Dimboola 151
Ballarat 131
Warracknabeal 127
Rochester 103
Donald 67
Beulah 59
Echuca 57
Skipton 53
Shepparton 51
Richmond 34
The following graph charts the commentary by location over time. In terms of
locations, comment about the Melbourne floods is short lived. Swan Hill commentary
continues for some time – relating to the duration of the event.
Mentions of Locations by Date
300
250
200
150
100
50
0
1/02/2011
2/02/2011
3/02/2011
4/02/2011
5/02/2011
6/02/2011
7/02/2011
8/02/2011
9/02/2011
1/03/2011
2/03/2011
3/03/2011
4/03/2011
10/01/2011
11/01/2011
12/01/2011
13/01/2011
14/01/2011
15/01/2011
16/01/2011
17/01/2011
18/01/2011
19/01/2011
20/01/2011
21/01/2011
22/01/2011
23/01/2011
24/01/2011
25/01/2011
26/01/2011
27/01/2011
28/01/2011
29/01/2011
30/01/2011
31/01/2011
10/02/2011
11/02/2011
12/02/2011
13/02/2011
14/02/2011
15/02/2011
16/02/2011
17/02/2011
18/02/2011
19/02/2011
20/02/2011
21/02/2011
22/02/2011
23/02/2011
24/02/2011
25/02/2011
26/02/2011
27/02/2011
28/02/2011
Swan Hill Kerang Horsham
Charlton Dimboola Ballarat
Warracknabeal Rochester Melbourne
19
22. Overall comment on specific rivers is relatively low given they were the source of the
flooding in regional areas and the subject of many SES information releases.
350 322
Comments Naming
300
Rivers
250
195
200 173
141
150
100 77 68
50
0
Murray Loddon Wimmera Yarra Avoca Campaspe
The Yarra had the highest number of warnings messages spread. There was
comparatively little relating to non urban rivers.
Type of Commentary
by River Mentions
350
300 19
250
200
1
150
100 144
50 94 119
0 9
Wimmera Murray Yarra Loddon
People in Floods Official posts News MS Warnings Spectators
MS Relief MS News MS News Recovery Issues
The Yarra River flood attracted a lot of comment, but not around the flooding event
in February. The early spike in comments about the Yarra related to an incident
where two people floated down the flooded river with an inflatable doll.
20