OR/17/045 Discussion

Duncan, M, Mee, K, Hicks, A, Engwell, S, Robertson, R, Forbes, M, Ferdinand, I, Jordan, C, and Loughlin, S. 2017. Using the 'myVolcano' mobile phone app for citizen science in St. Vincent and the Grenadines: a pilot study. British Geological Survey Open Report, OR/17/045.

myVolcano was originally developed to gather transboundary observations of distal volcanic hazards (i.e. ash and gas/aerosol emissions) with the principal aim of helping scientists to constrain the movements of, and understand the characteristics of, such hazards. Whilst ash, gas and aerosol were the primary focus of the app, it was designed in such a way that it could be used to gather observations of any hazard or natural phenomena around the globe (i.e. by accepting photographs and free-text descriptions). Through the 2015 scoping study (Mee and Duncan, 2015) several adaptations to the original design of myVolcano were identified, including a more explicit inclusion of all natural hazards affecting the region and development of the app on Android. Several issues were also highlighted by the 2015 study, such as the need for local data validation and management, data sharing, resilience of key systems and how to motivate users, ensure their safety and manage their expectations. The 2017 pilot study in St. Vincent has built upon these original findings, most importantly by directly engaging with potential users of the app. Through the workshop, the key needs of users (uploaders and recipients of information) in the context of using myVolcano as a citizen science tool in St. Vincent were identified.

In general, participants were positive about the app, but would only be likely to use it pending improvements. The desire for an all-encompassing app, which could both accommodate two-way communication between citizens and officials, as well as recording a whole range of hazards, emerged. Capturing multiple hazard observations could help identify cascading hazards, impacts and timescales which, whilst complex, could contribute towards more effective and timely assessment and management of an evolving hazardous situation. Participants also expressed the need for informative and interactive communication, particularly around the concept of giving and receiving ‘response advice’ from NEMO i.e. where people should go during an emergency; what they should do during a hurricane, etc. This was discussed in both passive and active terms, ranging from raising awareness to active advice based on feedback from actual observations. Two-way exchange of information and inclusion of actionable information was emphasised throughout the discussions. Better understanding the role of existing data sharing applications (Figure 11 and Figure 14) for citizen science will help to refine this component.

Another major theme to arise from discussions was motivation, as it was deemed that myVolcano would need to be more effective for communicating than existing approaches (e.g. Facebook), be multi-purpose and remove emphasis on volcanoes. School students were identified as a particular group of users that could be targeted in the 2015 scoping study (Mee and Duncan, 2015). Whilst there was enthusiasm for myVolcano, it is clear that more will need to be done to incentivise their interest, as there remained a significant degree of uncertainty as to the extent to which they would use the app, inferred as being partly owing to the Wi-Fi connectivity problems during the workshop. As Figure 11 and Figure 14 demonstrate, participants are already making use of a number of data sharing platforms, so for myVolcano to be relevant to their needs, it would need to be all encompassing or be integrated within existing means of data sharing.

The theme of localisation also emerged during the discussion, in the context of improved functionality of the app (the inclusion of local maps and information), but also through the local management of the observations in the app. There was an emphasis on visualisation, making observations clearer (e.g. labelled by hazard and filtered), as well as discussions about data validation and real-time application of myVolcano (something the app was not originally designed for). Many of the points raised by participants were in the context of real-time, emergency situations, rather than recording day-to-day observations. This may in part be owing to the choice of scenario. Although the scenario was designed to capture the time span of an event (e.g. from pre- to post-event), it arguably lacked emphasis on the ‘day-to-day’ monitoring and observing of the multi-hazard environment. It was greatly emphasised that there should be minimal delay in visualising and being able to use the observations recorded by myVolcano, whether these are in their original form or with added interpretation. Filtering data, being able to comment on observations and the suggestion that those doing the validation should add a summary or interpretation to the observation to improve its usefulness, were all suggested.

Suggested improvements: meeting user needs and Figure 9 presented the concerns and suggested improvements participants raised, some of which have already been addressed in V2 of myVolcano (released in April 2017). In terms of the resilience of key systems, both the challenges of maintaining Wi-Fi access during the workshop and technical discussions regarding communication systems on island led to concerns about low bandwidth and operating offline. Suggestions for improved offline/low bandwidth usage were made, but there was also a mention from one group that they would have confidence that key systems would not fail during an event. A fundamental function of the app that currently doesn’t work without internet access is availability of the map bases, which are streamed via online web map services. A way of addressing this is to cache map tiles within the app which can then be called when working offline. The related concern of availability of appropriate technology, emerged from the workshop questionnaires and the school worksheets, where it was evident that Android is the dominant platform and thus myVolcano needs to be available on this platform. This has subsequently been addressed since the workshop and an Android version of myVolcano was released in April 2017.

The issue of ensuring the safety of volunteers emerged throughout both the workshop and school activities. During the workshop, Group A emphasised that myVolcano would not necessarily encourage people to get any closer to a hazard than they already do (i.e. to take a photograph). During the third stage of the scenario exercise (at the peak of the eruption), Group A felt that people would still be trying to take photographs (through myVolcano or otherwise), whereas a police officer in Group C recalled that during the 1979 eruption people were too busy running to safety at this stage of an eruption rather than taking photographs. The app would need to carefully address safety concerns to ensure that users’ safety comes before recording observations in the app, for example ensuring people are not returning to evacuated/affected areas before they have been declared safe by authorities.

The results of the workshop have provided evidence to support the five components of resilience building where citizen science is perceived to have a role (Mee and Duncan, 2015):

1. Coordination and collaboration between scientists, authorities and citizens: users want the app to create a dialogue between these key groups;
2. Decision-making by institutions and individuals: the identified role of the app as a decision making tool;
3. Anticipation of natural hazards by monitoring institutions, authorities and citizens: the discussions around verification and the 24 hour delay emphasised the need for timely, actionable information;
4. Capacity building of institutions and communities: emphasis on awareness raising as a key role of the app;
5. Co-production of knowledge: suggestion to build a commentary between uploaders and recipients of data; localisation of the app.

Although positively received by participants, the results of the scoping study indicate that myVolcano in its current form is not fully fit for purpose in St. Vincent, without a number of technological, design and operational (including local management of the app) changes and additions.