OR/15/032 St. Vincent and the Grenadines: A case study

Mee K, Duncan M J. 2015. Increasing resilience to natural hazards through crowd-sourcing in St. Vincent and the Grenadines. British Geological Survey Internal Report, OR/15/32.

STEP 1: Identification of key partners

Several key stakeholders were contacted in connection with the app (Figure 3), including representatives of the National Emergency Management Organisation (NEMO), the University of West Indies Seismic Research Centre (SRC) based in Trinidad and Tobago and the University of West Indies Disaster Risk Reduction Centre (DRRC) in Jamaica. Monitoring of volcanic, earthquake and tsunami hazards in the Eastern Caribbean is conducted by SRC. SRC collaborates with a small local unit called the Soufrière Monitoring Unit (which operates from the Ministry of Agriculture in Kingstown). Key insight into the appropriateness and need for an app for monitoring environmental change, volcanic unrest and hazards in St Vincent was provided by Dr Richard Robertson (Director of SRC) and Dr Barbara Carby (Director of DRRC), who has extensive DRR experience in the country:

“I think there is a place for crowd sourcing, it gives potentially wide coverage and encourages public participation, fosters an interest in DRM and builds awareness for the importance of research” (Dr Barbara Carby, email correspondence 2015).

In St Vincent, both seismicity and ground deformation (possible means of anticipating volcanic eruptions) are continuously monitored by SRC. These data are analysed in real-time by scientists at SRC and are used, in combination with any other available information, to inform decisions regarding alert levels at the volcano. SRC also produces hazard maps and are involved in a number of public engagement activities, focusing on engaging with school children and, more recently, institutions and governments that can enact policy to ready Caribbean nations to deal with natural hazards (Trinidad Daily Express):

“In an environment of slow-moving policy changes, having such intimate knowledge of risk and the potential destruction of a geologic event can take a toll on some scientists. According to [Dr Richard] Robertson, it can either 'frustrate or motivate'. For the UWI-SRC Director — who admits that he reflexively identifies the exits when he’s in any building — knowing the risks makes the outreach function of the SRC an even higher priority.” (Trinidad Daily Express).

Collectively, SRC scientific staff has experience of volcano-seismic crises in several islands over the past three decades (SRC contribution to Brown et al., 2015b). Over half the present staff (60%) have experience of volcanic eruptions (SRC contribution to Brown et al., 2015b). They also monitor and advise on earthquakes and tsunami across the region. There are 35 full-time staff (19 scientific, 8 technical and 8 support staff) made up of seismologists, volcanologists and geologists, amongst many others (SRC contribution to Brown et al., 2015b). There exists, therefore, substantial regional knowledge and expertise, but also a huge demand on capacity, especially during volcanic unrest and hazardous events: SRC currently monitor active volcanoes on 8 territories (a total of 21 volcanoes) of the Lesser Antilles (SRC, 2011) where 12 of the islands have one or more active volcanoes (see Lindsay et al., 2005).

Challenges facing SRC

Scientists at SRC believe that by gaining a greater understanding of natural hazards and enhancing the centre’s operational facility, they can provide much needed information that can be used in the development of policy for land use, building codes, volcanic contingency plans and other means of coping with the threats inherent in the environment:

“We also need to make a greater effort to bridge the gap between science and policy and so move towards a state where the results of research by scientists are incorporated more seamlessly into the policy and plans for development plans in the Caribbean.” (Dr Richard Robertson, Trinidad Daily Express)

File:OR15032fig3.jpg

Figure 3 Key organisations involved in volcano monitoring and DRR in St. Vincent and the Grenadines and the wider Caribbean.

SRC have the resources to respond to a developing situation at a volcano in the region (SRC contribution to Brown et al., 2015b), but a challenge is the timely installation of often costly additional monitoring equipment. Crowdsourcing could assist SRC in the identification of possible increases in volcanic unrest. Collected at a comparatively low cost, crowdsourcing (‘big data’) could be used to justify installation of additional monitoring equipment and/or personnel. As demonstrated in other volcanic environments, during times of heightened activity at the volcano, public observations can support/confirm interpretation of instrumental observations (Stone et al. 2014). Dr Robertson states that observations of environmental changes, geophysical hazards, hydro-meteorological hazards and impacts would all be of value because scientists and emergency managers cannot be everywhere at once ( Appendix 6 and Appendix 7).

Communities have an existing role in monitoring the volcano:

“Currently, the community have a role in monitoring partly through local staff in some territories, but otherwise through informal interactions in response to an elevation in activity levels.” (SRC contribution to Brown et al., 2015b)

The app could enhance these existing relationships by providing a means of systematically and quickly recording this information and encouraging people to share their observations without SRC personnel needing to be available in country. It should also encourage the role and ownership of communities in monitoring their own environment (e.g. Stone et al. 2014).

Gaps in governance and disaster management

Any initiative for building resilience is only as successful as the underlying governance. In SVG, legislation for disaster management has existed for over sixty years (see Appendix 7). Since the implementation of the Hyogo Framework for Action in the Caribbean, there has been progress towards achieving, at least partially, some of the priorities for action although challenges still remain (GFDRR, 2010; Table 1) . In the Caribbean as a whole, limited resources, lack of political will, inconsistent enforcement of laws and the need for a shift in culture from a focus on hurricane preparedness to comprehensive all-hazards disaster risk management approach, are seen as the biggest obstacles to success (Carby, 2011). Many of these shortfalls have also been recognised in SVG (Table 2).

SRC agree, that in order to reduce volcanic risk within SVG, there is a need to implement effective land use planning regulations to reduce vulnerability and exposure of the population and infrastructure (SRC contribution to Brown et al., 2015b) . A major problem is that the long return periods of volcanic crises undermine response and mitigation efforts (SRC contribution to Brown et al., 2015b). The app is therefore designed to monitor not only volcanic unrest, but also other hazards that are often common in volcanic environments, particularly in the tropics, and that can interact with the volcano. Ensuring the monitoring of these additional hazards will be achieved by emphasising that the app monitors environmental change. Describing the app in these terms may assist in maintaining the awareness of the threat of volcanic eruptions during periods of inactivity.

STEP 2: Natural hazards and disasters name

Small island developing states (SIDS) face disproportionately high risks (UNISDR, 2015). They are particularly vulnerable to volcanic hazards with disproportionately high numbers of fatalities, probably due to the proximity of the population to the volcano and evacuation difficulties (Auker et al. 2013).

SVG has suffered a number of disasters over the past decades. Up to 1680 people were killed and 144 injured in an eruption of La Soufrière in 1902, and in 2013 there were 9 fatalities and over 500 affected by floods and landslides. Since 1990 most fatalities have been caused by floods and landslides and most economic losses are caused by cyclones (UNISDR, 2015).

Robertson (2005) described eruption scenarios at Soufrière St. Vincent and presents integrated hazard maps for effusive dome-forming and explosive eruptions. The maps show much of the north of the island as Very High Hazard, with decreasing hazard moving southwards. For the purposes of making national comparisons, a country volcanic threat profile for St Vincent was compiled by the ‘Global Volcano Model network’ in 2015 (Brown et al., 2015b, Appendix 8). A flashflood hazard map and landslide susceptibility map for SVG have recently been produced by the EO-RISC project (see Step 5; Appendix 4).

Table 1 Successes made towards achieving the HFA Priorities in Saint Vincent and the Grenadines

HFA Priorities	Successes in Saint Vincent and the Grenadines
(1) Ensure that disaster risk reduction is a national and local priority with a strong institutional basis for implementation	New legislative act established in 2006 NEMO established as a result of legislative act, with full-time staff dedicated to disaster response and planning National Emergency Plan developed and operational Line ministries are taking responsibilities for their functional areas during a disaster
(2) Identify, assess and monitor disaster risks and enhance early warning	Volcano hazard maps have been produced and are regularly revised Seismic monitoring is in place across the Eastern Caribbean, including the island of St. Vincent and Soufriere St. Vincent volcano. Local monitoring is carried out by the Soufriere monitoring unit Meteorological hazards are monitored in SVG and early warning systems in place Good communication network in place for disseminating information and early warning to meteorological hazards
(3) Use knowledge, innovation and education to build a culture of safety and resilience at all levels	NEMO sponsors regular public risk-awareness programmes and the PM issues an annual public address at the onset of hurricane season Volcanic hazard awareness has been boosted through public and stakeholder events (e.g. CaASHI workshop), STREVA Introduction of disaster preparedness in the educational system
(4) Reduce the underlying risk factors	Government-sponsored development and larger infrastructure projects generally include resilient design DRR planning has been drafted for the tourism sector and there are plans to strengthen risk mapping
(5) Strengthen disaster preparedness for effective response at all levels	NEMO has significantly raised public awareness to natural disasters SVG is a member of CDEMA Tourism sector is largely insured by commercial underwriters and is therefore relatively well-protected 40 community disaster groups have been established and received training in damage assessment, shelter managements, relief supplies management, first aid etc. Vulnerability assessments for the health sector infrastructure have been carried out, with some recommendations already implemented

Table 2 Barriers to success in achieving HFA priorities in Saint Vincent and the Grenadines

HFA Priorities	Barriers to success
(1) Ensure that disaster risk reduction is a national and local priority with a strong institutional basis for implementation	DRR through policy and planning still in early development and not yet mandated as a development objective
(2) Identify, assess and monitor disaster risks and enhance early warning	Hazard mapping not so advanced in other areas with limited GIS capability across some organisations
(3) Use knowledge, innovation and education to build a culture of safety and resilience at all levels	Most efforts are focussed on preparedness and response, less so on risk reduction through effective planning and risk avoidance strategies
(4) Reduce the underlying risk factors	Resilient design not employed on a widespread basis Land use planning is not currently a factor for DRR in SVG
(5) Strengthen disaster preparedness for effective response at all levels	SVG still relies heavily on outside assistance in the event of a major disaster Whilst the tourism sector may be well insured other sectors, such as agriculture, transport and housing remain relatively vulnerable

SRC monitors regional earthquakes and has an alert level system for the volcano, there is also a regional tsunami and coastal hazards early warning system (CARIBE EWS). An app and associated tools can contribute significantly to the collation and sharing of information (including maps) on all hazards and environmental change not just during disasters but continuously. There are several Facebook and Twitter accounts (SRC and NEMO) that could provide useful information to myVolcano users, as well as useful information that people might post or tweet that could be ‘pinned’ to the map interface in the app.

The app can also support SRC and NEMO in managing community anxiety through provision of information. For example, in mid-February 2005, widespread sulphurous odours and haze on the island of St. Vincent and as far as the Grenadines (50-75 km S) led some people to conclude that the smells reflected increased output of volcanic gases from the volcano, there were fears that an eruption was imminent. SRC worked with the head of the local volcano monitoring unit, to first investigate the reports and later to quell fears that an eruption was imminent.

STEP 3: Characterise mobile phone and internet use

Paramount to the success of any citizen science initiative is establishing that (a) a ‘crowd’ exists and (b) there is sufficient access to the necessary technology (see Pocock et al.’s [2014] decision framework for citizen science). This step is necessary in understanding whether this infrastructure exists and, therefore, whether the use of an app would be appropriate. It also highlights any potential development needs (i.e. does the app need to be developed on another platform?).

Appendix 9 contains statistics compiled by the National Telecommunications Regulatory Commission (NTRC) of Saint Vincent and the Grenadines. In 2014, the number of fixed line telephone subscribers (19 176) accounted for only 17.5% of the population, whereas the number of mobile phone subscribers (108 965) is equivalent to 99.6% of the population (some people have two phones). The percentage of users accessing the internet increased from < 20% in 2007 to > 60% in 2014. The dominant mobile phone platform according to the International Business Times (2013) is Blackberry, whilst Dr Robertson of SRC believes Android is most popular (Dr. Robertson, pers. comm.). Dr Robertson also stated that Facebook is very popular and both SRC and NEMO have well established Facebook pages through which they disseminate information and advice ( Appendix 7).

The app is currently developed only for iOS devices (iPhone, iPad, IPod) although an Android version is being planned. Most modern Blackberry systems can use Android apps, since few people develop apps specifically for Blackberry. However, until an Android version is fully operational, a web version of the app has been developed. This is relatively cheap and quick to update and ensures that anyone with access to the internet is still able to contribute their observations.

STEP 4: Recruit citizen scientists and networks

Motivating participants requires that we provide them with clear reasons as to why their participation is so important (Pocock et al., 2014). The motivation for engagement may relate to different socio-economic and cultural factors and will require understanding the country context where the app is being promoted. However, we can take lessons from existing studies. At Tungurahua volcano, the motivations of volunteer observers (vigías) are an important component of the network’s success — all feel a sense of duty or moral obligation and want to help reduce risk to their family and community (Stone et al., 2014). Vigías stated that the voluntary nature of the role is very important to them. These volunteers were recruited through existing relationships with Civil Defence and scientists owing to the fact that monitoring equipment was located in their farmland. During the 2010 and 2011 eruptions in Iceland, UK citizens could relate to the hazard because it threatened something many of them do on a weekly, monthly or yearly basis — fly. Key entry points therefore include in-country contacts, networks and cultural interests and perceptions of risk.

Whilst MyVolcano would be freely available to anyone with access to the internet, Dr Robertson suggested that the app should target key individuals within communities who could be relied upon to regularly upload information. NEMO supports 20 community disaster groups that meet to assess and plan for the hazards specific to their community (Lowe, 2010). It is therefore envisaged that NEMO could assist with the identification of key individuals as well as highlighting the presence of the app for collecting observations and finding information. Likewise, building on SRC’s existing public engagement activities should allow for the identification of key individuals and groups.

Both NEMO and SRC are already involved in many community outreach and schools projects. The BGS has also successfully engaged with schools for several years as part of the Schools’ Seismology Project and has a network of schools across the British Isles who are equipped to collect ash samples when the next eruption occurs. In SVG, disaster preparedness is included in the educational curriculum (GFDRR, 2010) so there is a possibility of producing educational resources around the app to complement these targets. Schools would only require one computer with internet access to enable regular observations to be made and can be tasked with looking for specific changes depending on their location and hazard context. Staff and pupils can also be provided with additional instructions and guidelines on how to make and record observations, whereas such a level of detail on the app might be off-putting for casual users.

In addition to targeting specific users, the app could be marketed and promoted (through SRC and NEMO) across SVG and beyond to encourage individual users to download and contribute observations through the app. A simple way to elevate the presence of the app is to link through social media. myVolcano already has its own Twitter account which can easily retweet information coming from SRC and NEMO to help engage people with the app.

STEP 5: Identify other ‘big data’ and research initiatives name

‘Big data’ can be used in a number of different contexts and with different objectives. Several ‘big data’ initiatives that specifically aim to increase resilience and support decision-making are active in SVG and the wider Caribbean region. According to Dr Robertson, research projects initiated outside the region can provide new knowledge that can contribute to DRR however, there is a need for greater linkages between some of these and existing regional institutions involved in similar research. International research projects provide opportunities for linkages with external agencies who may be better resourced and so are able to innovate and advance existing knowledge and understanding of regional systems. Dr Barbara Carby (DRRC) suggests that research should be solution-oriented and related to specific national needs but with potential regional benefits ( Appendix 6). Thus, decision-makers should be involved in problem definition and research design, so that research builds national/regional capacity, complements previous work and encourages standardisation of methodologies where appropriate.

The SVG proposal was conceived partly because many projects were active on one small island but were not interacting in any way and because key in-country stakeholders and experts (e.g. SRC) were unaware of some projects ( Appendix 3). International research investment can appear rather fragmented (see comments by Barbara Carby — Appendix 6) and rarely, if ever, connected or linked. The app can make connections between these independent initiatives that may fail to recognise all stakeholders.

Earth observation

Earth observation (EO) data products come in many forms including images, maps, time series plots, videos etc. Satellite EO products are highly complementary to the traditional ground-based and airborne data and information collected by volcano monitoring institutions. The EO needs of monitoring institutions are challenging to meet particularly in terms of spatial and temporal resolution). Often products that make the biggest difference are simple — for example those that detect change (temperature and topography), topographic mapping and visual images. Surono et al. (2012) documents the first time co-production of knowledge from ground (seismic, geodetic and gas observations) and space (near-real-time satellite radar imagery) was achieved to save lives and reduce economic losses during the 2010 eruption of Merapi in Indonesia (facilitated by the International Disaster Charter). A current EO project, EO-RISC (Earth Observation for Risk Information Services in the Caribbean) is an European Space Agency (ESA) ‘eoworld’ funded project to demonstrate the benefits of satellite earth observation (EO) technology ( Appendix 4). The project has produced an improved land cover map, a flash flood hazard map, a landslide inventory map and a landslide susceptibility map for SVG (http://charim.net/stvincent/maps). These maps have been delivered to NEMO and other government departments in SVG.

Few EO projects begin with an analysis of user needs; however, the BGS have experience of a ‘user-led’ EO service that developed the required services and an interface for simple interaction with data products.

Unfortunately, the project could not be sustained much beyond the research project duration (EVOSS) ( Appendix 4). We wish to develop an app that is sustainable, useful throughout the ‘risk management cycle’, and user-designed and led. The service would connect those on the ground with open access data products, operational EO resources and existing platforms that are most likely to build resilience and reduce risk.

Crowd sourcing

The Caribbean DEWETRA Platform created by the Caribbean Institute of Meteorology and Hydrology (CIMH) and partners is essentially a data fusion platform that manages and displays hydro-met hazard, exposure and impact data from various sources in a real-time, geo-spatial environment in support of improved decision making ( Appendix 4). CIMH used the platform to capture impacts during an assessment of the floods in SVG in December 2013 (Dr Shawn Boyce, 2015, pers. comm. — see Appendix 6).

SRC has an existing ‘Did you feel it?’ template on its website for citizens to report their experiences. The University of the West Indies Disaster Risk Reduction Centre is also about to trial a crowd-sourcing project (Dr. Carby, 2015 pers. comm.). We aim to connect with all these initiatives which operate at a regional scale.

Due to time constraints, one aspect of crowd-sourcing that has not been fully explored during this scoping study is how the App could link to social media. There are several Facebook and Twitter accounts (SRC and NEMO) that could provide useful information to myVolcano users, as well as useful information people might post or tweet that could be ‘pinned’ to the map interface in the app. The BGS web tool ‘GeoSocial: Aurora’ is a crowd-sourcing tool that scours Twitter for tweets related to aurora sightings and maps them on an interactive map. The database archives tweets and therefore can be used for the purposes of evaluation. There is potential to have a similar function within myVolcano that displays tweets with certain hashtags on the interactive myVolcano map. The app could also create/suggest hashtags for people to use in their tweets so that the app is picking up the most relevant information. Whilst SRC should own the app, BGS can provide this data management support through the existing ‘GeoSocial: Aurora’ initiative with agreement from SRC and NEMO.

Open GIS

The World Bank’s Latin America and Caribbean Region (LCR) DRM team helped to implement open source, collaborative, geospatial, data-management platforms in nine countries, including SVG. This means that a wide range of practitioners (e.g. ministries of works and physical planning, academics and development agencies) have access to free GIS software (such as QGIS) and free data so that they can create their own geospatial resources, such as hazard maps.

Joining the dots

The app can act as an important link between these disparate data types and projects. The app can provide links to different projects and sources of data to ensure users are aware of different initiatives operating in SVG and the region. Data can also be served directly to another platform as a web service, which means up-to-date data from one platform is automatically displayed in another.

STEP 6: Agree app specifications

Following discussions with key stakeholders (e.g. SRC, UWI Disaster Risk Reduction Centre) and the app developers, suggested additions/changes can be made to the app. These changes include additional information added to the map interface (e.g. sign-posting to monitoring and emergency management organisations relevant to SVG), data collection (what is collected and how), data and information dissemination and the underlying technology.

Table 3 lists the specific features relevant to SVG that can be incorporated into the app. These features include links to the websites of SRC, NEMO and CDEMA, incorporation of volcanic hazard and other maps, links to existing crowd-sourcing tools and other projects operating the area, including videos created by the STREVA project about the 1979 eruption of La Soufrière and volcanic hazard associated with the volcano (Figure 4). Social media, such as Facebook and Twitter, can also be linked into the app so people sharing information in the app can also share information directly to their own social media sites.

Table 3 Suggested adaptations to the myVolcano app following consultation with UWI Seismic Research Centre.

App feature	Action	Solution
Map interface
Functionality	Set to country/region	Either by choosing a pre-defined area (SVG, Caribbean) or zooming to a preferred extent and saving the view
GIS layers	Hazards information	Turn on and off hazard maps (e.g. integrated volcanic hazard map from SRC, landslide susceptibility maps from CHARIM?).
		Links to information about different hazards in the region
	Observatory information	Click on any volcano/country in Eastern Caribbean and brings up SRC’s information (SRC website, Facebook, Twitter)
	Scientific advisories	Click on a volcano and any official scientific advisories (from the observatory) can be accessed
	National emergency management information	Click on any country and the corresponding emergency management organisation details are shown (e.g. NEMO for SVG — website, Facebook)
	Regional emergency management information	Click on any region and the corresponding emergency management organisation details are shown (e.g. CDEMA for Caribbean)
	Active projects in the region	Click on Soufrière St Vincent for links to the STREVA videos, CHARIM project, DEWETRA project
Data collection
Simplify layout	Remove focus from ash collection to multi-hazards	Hide ash measurements from main view and give user option between taking a photo and written observation. If choosing ‘observation’, brings up a list of all different hazards from user to select from. Depending which they choose will depend what options they see (e.g. earthquake brings up SRC’s form, ash brings up ash measurements, landslide brings up landslide measurements, etc.)
Disseminate information
Push notifications	Ability to send messages to users via the app	Warnings, advice, alerts, scientific advisories, encouragement to collect data if something has happened
Technology
Platform	Android version	In planning
	Web tool	Continue to develop the web version to support the iOS app until an Android version has been developed

STEPS 7 and 8: Develop, pilot and evaluate

Following consultation with one of our key partners — SRC — the next step will be to continue development of the app based on the agreed user requirements and to design a suitable pilot study in-country. Technical development of the app will continue to be led by BGS and we also anticipate assisting in managing and linking the data with SRC’s systems. Training and workshops with the potential core users group will ensure data are collected consistently and that all ideas and opinions are included. Further innovation will be welcomed. For full implementation, BGS will continue to work with SRC providing technical and other support as needed. Evaluation of the app and its use will be conducted again when it is operational and has been used during a significant event.

File:OR15032fig4.jpg

Figure 4 Summary of the type of data that could be collected by the myVolcano app and the range of information it could provide to users, following consultation with UWI Seismic Research Centre, in the context of SVG and the wider Eastern Caribbean region.

Anticipating challenges

Throughout the prototype development, we have anticipated and been made aware of several challenges that need to be overcome before the crowd-sourcing approach could be successfully implemented in SVG.

Data validation and quality assurance: A major challenge anticipated by key partners (Dr Roberston and Dr Carby) is how to validate the data and provide quality assurance, particularly if SRC and NEMO are to make decisions for issuing alerts and preparing for emergencies based on these data. SRC stated that they already actively engage in quelling rumours when members of the public make them aware of changes in activity on the volcano and thus suggested that they could take on the role of data validation. Managing large volumes of data will require dedicated time and staff and this is where a partner organisation, such as BGS, could provide support for data management. This support would not be to remove ownership from SRC, but instead to support their response. Dr Robertson also emphasised that a means of ensuring data validation would be to have a reliable base of dedicated volunteers, whose recordings could be compared with SRC monitoring (see Appendix 7).

Resilience of key systems: A key process of building resilience is ensuring backups in any system. The size and proximity of the population to the volcano in SVG mean that it is necessary to explore the resilience of mobile and home internet during an eruption (or other event), particularly given the role of the app in linking users to SRC’s and NEMOs alert levels and advice. It is also necessary to anticipate how the volume of data recorded by the public may fluctuate over time. In Haiti, for instance, the volume of crowdsourcing data was overwhelming. A means of overcoming this would be for SRC to identify a threshold of information received that would trigger assistance from other agencies (e.g. BGS) for the purpose of data management.

Motivating volunteers, managing expectations and ensuring safety: Maintaining awareness of volcanic risk between eruptions is a notable problem in volcanic environments, however the means of overcoming this problem is to not separate hazards and risks but to emphasise their links. The fact that the app is designed to encourage monitoring of other hazards, e.g. hydrometerological events, under the umbrella of environmental change should help to maintain awareness. Managing expectations is also another anticipated challenge and lessons can be sought from other initiatives. For instance, a key element of our approach is the two-way exchange of information. However, as observed by Wallace et al. (2015) at the Alaska Volcano Observatory, there can be a time delay between citizens uploading information and it being displayed on the site due to the time it takes to validate the data. Their means of overcoming this was to include a disclaimer (Appendix 2). The importance of ethics and a code of conduct for users must also not be overlooked: Dr Robertson noted that the app should not encourage people to venture into dangerous areas, whilst NEMO recently emphasised on their Facebook page that images of fatalities are not appropriate for sharing. Developing such a code will be an iterative process involving key stakeholders, including SRC and NEMO.

Data sharing: An objective of the app is to link to different projects and initiatives and share data. Whilst the need for open access data is being increasingly acknowledged in building resilience and risk reduction, it continues to be a problem. In the context of the app itself, users are made aware that the information they share will be freely available, thus fostering a culture of transparency, openness and sharing.