Crowdsourcing strengthens climate resilience

The complexities of inherent bias, variation in local resource patterns and restricted perspective can make evaluating local responses to severe climate events difficult. When coupled with their low incidence, the utility of performance reporting is highly compromised. Crowdsourcing approaches can improve the evidence base when evaluating communities’ resilience to severe climate events.

To be effective in this context, crowdsourcing approaches need to meet three criteria:

1. Enlist a wide range of observer types – Observers from focus communities have the most direct stake in and exposure to the climate events under consideration. They also have a narrower view of response options and may report in a way they anticipate will elicit a beneficial response from government or donor organisations. External observers who have some level of contact with the community and a general understanding of climate change resilience can help offset these biases.

2. Collect longitudinal information – A single climate event can yield important but limited information. Collecting observations form the same set of reporters over time builds a more robust picture of climate event responses.

3. Combine observer responses with other empirical data – Sufficiently granular measurement data such as crop yields, soil erosion and migration trends provide explanatory support to subjective observer responses. If two or more of these data streams can be synchronised over time and geography, they can create robust statistical models to verify theoretical theories of adaptation.

Identifying and recruiting the right observers is as important as the data they report. Having resilience programme teams directly recruiting observers from the focus communities is one option.

Other methods include public service announcements aired on local radio stations calling for participants in activities such as a contest to predict the first rainfall of the upcoming season. Respondents are asked to send a text message to a specified number with their prediction. While not highly useful in itself, the resulting database of contacts provides a source of potential recruits who can be contacted to participate in broader observation efforts.

Other approaches entail using billboards or signage at communal sites, and teaming with civil society organisations to enlist their members as observers.

A number of relatively low-cost technologies are available to support crowdsourcing efforts. Chief among these are Short Message System (SMS) capabilities inherent in virtually all mobile communications systems. With rapidly increasing mobile penetration in even some of the most remote communities, one can establish a reporter network across large geographies at low cost. Detailed surveys, based on a question-response-question cycle, are now feasible.

With reverse charge services available in many places, it is possible to eliminate observer costs or even incentivise observers by providing phone credit for participation. Interactive Voice Response (IVR) systems use numeric keypads and guide the observer through questions using voice prompts. This technology allows for rapid data collection with easy branching for detailed responses.

Smartphone-based technologies also bring significant crowdsourcing opportunities. Applications such as WhatsApp or Viber provide forums for free discussion. Although still not ubiquitous within rural focus communities, smartphone technologies can help build robust reporting in urban and peri-urban populations.

Other key capabilities include onboard geolocation tagging (GPS and network-based), onboard cameras to document local phenomena/conditions and the ability to host customised applications to collect survey or measurement data.

With the increased coverage and decreased cost of high resolution, high revisit rate satellite imagery (having the satellite revisit the same area over time), in conjunction with observer reporting, can constitute strong evidence based on climate event datasets. Such imagery may be used to target the recruitment and engagement of observers in certain areas, or enhance the value of observations.

For instance, following a flood, event imagery can help define the flood impact area. Using this information, observers within these areas are then solicited to provide community response information.

Alternatively, imagery archives are a way to retrospectively characterise and define the magnitude of an event. Satellite imagery can also document observable changes in agriculture, drainage patterns, infrastructure conditions and settlement patterns.