OR/15/032 Appendix 6 Correspondence
| 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. |
Correspondence with Dr Barbara Carby, Head UWI Disaster Risk Reduction Centre:
Dear Katy
Please see my thoughts in text below — I am away from Jamaica on official travel at the moment.
From: Mee, Katy [[1]]
Sent: Tuesday, May 12, 2015 7:18 AM
To: Carby, Barbara E
Cc: Loughlin, Susan C.; Duncan, Melanie J.
Subject: Crowd-sourcing app for monitoring natural hazards
Dear Barbara,
I am a GIS specialist at the British Geological Survey (BGS), working with the STREVA project. My colleagues at BGS (Sue Loughlin and Melanie Duncan, copied in) and I are currently also working on a small pilot project (funded by NERC-ESRC-DfID) to explore the potential of ‘Big Data’ to support disaster and climate resilience in St. Vincent. This is part of a new Science for Humanitarian Emergencies and Resilience (SHEAR) programme.
We’re interested in two aspects of Big Data in particular: crowd sourcing and earth observation (satellite) data. We would be very grateful indeed if you could take the time to briefly answer a couple of questions. First some background:
BGS has developed an app for crowd-sourcing observations of natural hazards. It was initially designed as a tool for collecting observations about volcanic ash and gases in a distal setting (i.e. the types of things we might see/smell in the UK from Icelandic eruptions). The app enables people to upload descriptions and photos, geo-located by their phone’s GPS, as well as showing users how to collect volcanic ash samples. The app could be developed further specifically for St Vincent. It enables the public to collect and share useful observations and such information might be useful for a range of organisations (e.g. NEMO, SRC). At a national scale, the app can also direct people to NEMO and other official organisations for advice and information. Richie Robertson also suggested, for example, that people use the app to take photos of things ‘out of the ordinary’, environmental changes that they might see or feel such as landslides, or felt earthquakes that might precede a volcanic eruption etc. It could be very valuable for recording impacts. We’re interested in exploring exactly how such crowd-sourcing could increase resilience.
We’re also interested in the very large number of projects that have aimed to increase resilience in the region using earth observation. For example, we were involved in setting up a platform to provide earth observation products to volcano observatories but sustaining the platform has been a challenge. BGS alone has been involved in a number of projects in St Vincent over the years funded by for example World Bank, EC, ESA, NERC, ESRC and others.
We appreciate that these are very big topics! Given your experience, particularly from a regional point-of-view, we’d be very interested in your brief views on the following:
1. What are the biggest challenges for St. Vincent in terms of increasing resilience to natural hazards?
A change in culture perhaps — thinking risk reduction instead of preparedness — and a concomitant change in planning horizons to medium to long term, where long term may now have be fifty — 100 years in order to accommodate CC Ensuring coherence and synergy among govt policies which impact risk reduction and CCA
2. Do you think there is a place for crowd-sourcing through apps like ours — how could it help and what are the main limitations of such tools?
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 The main problem would e validation and quality assurance
3. Are there, or have there been any other crowd-sourcing initiatives in the region that you’re aware of?
We are about to try it for a project out of DRRC
4. What are the challenges in using earth observation (satellite) data and products?
Don't do much remote sensing myself — the problem I suspect would be resources if the data/products are not open access. There may also be problems of downscaling in some cases
5. Do you have any other comments on these topics?
How can international research investment be more effective, supporting decision makers and really building resilience?
Should be solution-oriented — thus decision makers should be involved in problem definition and research design, related to specific national needs but with potential regional benefits, should build national/regional capacity, should complement previous work if at all possible, should encourage standardisation of methodologies
Many thanks
Katy Mee
Meeting notes from conversation with Dr Richard Robertson, UWI Seismic Research Centre, Trinidad and Tobago:
(Our questions in black, notes from Dr Robertson’s responses in red)
What data might be of use in St. Vincent and for whom?
- Data collected by people during eruptions is key'
- Ash database — can compare the public’s collection with SRC point of validation
- Pyroclastic Density Currents (PDCs) — record locations of them
- Data on people and impacts — of use to emergency management teams (call this ‘other’ data and information)
- Collecting info like this would help NEMO etc.
- Collect ash samples for SRC
- Focus on unrest — small changes in environment before an eruption – felt earthquakes/earthquake swarms, storms over the volcano (look for storm clouds)
- Link to hydro-meteorological hazards — landslides are under-measured
- Observatory at Belmont on west side of volcano — only staffed when an eruption is imminent and during an eruption — can only see one side of the volcano so having people take photos, observations on the other side would be useful
- Where flows (lahar, PDCs) might go
Who will collect the data?
- There might be key people in a community you could work with
- Need to avoid treading on toes of civil protection
- Avoid people going into danger areas to collect data — balancing understanding with risk taking
- People walk up the volcano all the time for leisure [when it’s not erupting], people could take photos if they see something different
Data validation — who would do this?
- SRC are willing to work on this — rumour quelling
What systems do most people use?
- Most people use android and many people have more than one phone.
- Main cell providers are Digicell and LIME
Current communication of data from monitoring — is it published?
- SRC will send reports to NEMO and eventually these reports reach the web. But they only tell the public if there is unusual or concerning activity.
- Planning to put a bit of data out, but maybe not necessary, as people evacuate quickly on St Vincent.
- Probably would have regular web updates if eruption happened.
Who is the App useful for/how is it useful?
- Would need to build in something useful for the community — add education stuff.
Other initiatives — does Richie know about these projects e.g. Colm’s EO-RISC peroject to produce land-use maps:
- Not heard of this — who are they working with in St. Vincent?
Meeting notes from conversation with Dr Colm Jordan, Team Leader Earth and Planetary Observation and Monitoring, British Geological Survey.
Who are the contacts in St. Vincent that the EO-RISC project is working with?
Colm to pass on a list of contacts but certainly NEMO know about the project and are involved. Chief engineer, SVG, NEMO, Ministry of Housing, Informal Human Settlements, Lands & Surveys & Physical Planning, Ministry of Transport, Works, Urban Development and Local Development
What happens to the end products (land use maps) and who can use them?
Essentially they will be made available to anyone. Maps/all data (maps or raw data or both?) will also be available through the Geohazard Exploitation Platform (https://geohazards-tep.eo.esa.int/#!pages/initiative). Some satellite imagery can only be provided to certain users due to license restrictions.
How is the charter activated?
Anyone who is an authorised user can activate the charter so would need to be registered but essentially anyone could do it i.e. for St. Vincent could be NEMO, SRC, BGS, USGS, CDEMA, St. Vincent Government etc. A new development in the Charter is ‘universal Access’ — any national disaster management authority can now submit requests for emergency response support to the Charter. Procedures have to be followed, but the affected country does not have to be a Charter member. In the UK, authorised users are the Cabinet Office and DfID. Once activated, the Executive will nominate a Project Manager, who has received specific training. The PM can also be the Value Adder, who processes/interprets the imagery and provides products to the Authorised User. The PM is normally someone/organisation with experience in that particular disaster e.g. BGS is a PM and VA for geohazards in the UK and overseas when appropriate.
How soon could the charter be activated? Could it be activated in advance e.g. of an eruption if it was thought an eruption was imminent but hadn’t actually happened yet?
Can’t really be activated unless a disaster has happened, but this has changed slightly so might be willing to agree to activate if an eruption was imminent, for example. Worried about activating the charter pre-emptively because could be setting precedence for others to use it irresponsibly. For example, BGS are trusted but then if it got activated and nothing happened that could affect the level of trust.
How long is the charter active?
At the moment there is just a short term response (~two week from initial event) but there is an argument for providing a longer term response [e.g. volcanic eruptions can last for months/years]. The Charter does not provide data to support rehabilitation, reconstruction, prevention, preparedness or scientific research. It also does not provide maps suitable for use in the field.
What is BGS’ (Colm’s) role?
Project Managers are the link between space agencies (data providers) and whoever has activated the charter (Authorised User). Coordinates the acquisition of data. We also act as Value Adder to processes and interpret data to make maps, provide data to users.
What happens to the data afterwards?
Can’t be used after the event (see above) but in some cases it is possible to contact the space agencies individually to request non-commercial access — they may agree to continue letting you use the data but this is at their discretion.
Could you use crowd-sourcing?
Possibly — how would you ensure that you have experts looking at the maps and not picking up slides from before the earthquakes? Aside from the Charter, systems such as Tomnod use crowd sourcing. Crowd sourcing was used to identify infrastructure damage in the aftermath of the Nepal earthquakes but expertise is required to differentiate/delineate active landslides so this task was done by experts.
Other challenges with the charter?
Data volumes are huge — have 6 PCs running just to download the data [for Nepal]. The data is made available over a short period of time and there is a real urgency to create useful products. The data may not be ideal e.g. there can be significant cloud cover. Much is done on a best-endeavours basis so some flexibility is required/expected.