During my PhD, I created a multipage online survey based on GoogleMaps. The aim of the survey was to understand:

1. What parks were used in Portsmouth
2. What the general population knew about flood risk in Portsmouth
3. What was most important for them to protect against flood risk
4. Which park were the most important to the population (and how much were they willing to pay for them)
5. What future they wanted for one of the parks.

I coded the interface using the GoogleMaps API. GIS was the central tool of the survey, which I used to collect geolocalised information either:

• on specific locations (step 2 and 5, the questions involved adding metadata to geolocated polygons),
• created geodata by respondents (step 3, adding points to indicate places at risk),
• as an information tool (step 4 and 6, where each land use is shown).

First step

The first step of the survey asks the participants to indicate where they live. They can do it by clicking on the map or by indicating their postcode. It uses a simple ‘listener’ added to the mouse click in order to record the users’ position while they click on the map on the position of their rough address. The marker can be dragged afterwards.

Second step: Open space use

During this step, the participant is invited to indicate which open spaces he/she uses and for what purposes.

The map on the page shows the usual GoogleMap map, adding a ‘kml’ layer of the open spaces in Portsmouth (the same layer that was created for the map paper survey). When the participant clicks on the park, an ‘infowindow’ appears. Participants can enter information into text boxes that are sent to the panel on the right of their screen. They can delete the open spaces if they need.

Once the participant is satisfied with the open spaces and they click on ‘save’, data is sent into the database.

Third step: Flood zones

The third step of the survey records the places that participants think are at risk of flooding.

The map ‘listens’ to the event ‘click’, so for each click on the map a marker is added onto the displayed map. Once a marker has been added, it stays on the map. The participant can add more points and indicate when it is finished. All the points indicated by the participants are stored in the database with their longitude and latitude.

The participant can indicate if he does not think that Portsmouth is at risk of flooding, or if does not know where it floods in Portsmouth.

Fourth step: Land use priorities

The participant is shown a list of land uses to indicate in which priority these land uses should be protected against flooding.

The map displayed shows the GoogleMap map and three other layers: the land uses, the schools and hospitals and the flood zones. Because of the amount of information shown, the Google Map has been reduced to minimum information (main roads and rails) and the three layers can be shown and hidden.

The participant can indicate his preferences in the panel on the right from 1 to 7. Any other number than 0 to 7 would return an error.

Fifth step: Willingness-to-pay

The fifth step investigates the willingness-to-pay for the open spaces of Portsmouth.

The map shows the open spaces in Portsmouth that are in flooding area. A click on any open space opens an ‘infowindow’ with a ‘text box’ where the participant indicates the amount he would be willing to donate for a year to protect the park against flooding. The ‘ok’ button sends the information given to the panel on the right. Each donation can be deleted before saving. This information is then saved into the database once the participants has clicked on the ‘save’ button.

Sixth step: Scenario for the FM

The last step invites the participant to indicate their preferred solution for the future of the FM.

The three possible scenarios are shown on the Googla Map as a ‘kml’ layer. An infowindow appears to explain further what the option would imply for the FM. The participant indicates his preference and save his preference.

Conclusion

I had added the first step for three reasons: first, I needed to know where the respondents lived. But I could have done it by asking only for a postcode or street. I also wanted to give them time to familiarize themselves with the map. Where you live is probably the easiest thing to find on a map, so asking this as an introduction helps the respondents to see how to navigate. And finally, this helped me to assess the respondents’ accuracy when I had the location and the postcode.