HOT and KLL set out on site visits across Kathmandu to assess the homogenous zones identified for mapping in OpenStreetMap. These seven zones have been identified as rural, residential, dense residential, urban, industrial, informal, high urban and new industrial. KLL will remotely digitise all the building footprints within the homogenous zones, before collecting detailed attribute information on the ground for a select sample of these buildings.
Kathmandu is one of the two cities that is the focus for testing and validating the exposure datasets created by ImageCat as part of the Monitoring Exposure Through Earth Observation Routines: METEOR project. The Humanitarian OpenStreetMap Team made a trip over to Nepal last week to meet with Kathmandu Living Labs (KLL), who are partnering on the METEOR project and will be collecting data in the city. Following the second biggest festival in the country, Tihar, also known as the festival for lights, the team set out on site visits across Kathmandu to assess the homogenous zones identified for mapping in OpenStreetMap.
The homogenous zones are areas of the city that have been classified to have general building patterns. For Kathmandu, there are 7 homogeneous zones which have been identified as rural, residential, dense residential, urban, industrial, informal, high urban and new industrial. The plan is that KLL will remotely digitise all the building footprints within the homogenous zones, before collecting detailed attribute information on the ground for a select sample of these buildings.
The sampling strategy looks to select 40 randomised points in each of the homogenous zone types. Once the points have been selected, 10 buildings surrounding each point are then selected and surveyed following the simplified version of the GED4ALL taxonomy. This information will help to validate the exposure datasets created by ImageCat for Nepal. Similarly, the same mapping process is being carried out in Dar es Salaam by Ramani Huria to assist with the validation of the exposure datasets created through earth observations for Tanzania.
A total of six sites were visited across Kathmandu. The first stop was to the Thapathali area located in an identified informal zone, which can be seen just across the river. Buildings in this area are single level and largely made up of thin sheets of materials such as plastic, with flat roofs made out of thin sheets of metal held down by bricks. The second stop was to Chaysal in a dense residential zone where many of the buildings are multiple level, made of a combination of brick and cement. The third visit was to Bhaktapur in an industrial zone, the fourth to Jhamsikhel in a high urban zone, fifth visit was to Sundhara which was a combination of both a high urban and new industrial zone, while the sixth and final stop for the day was to the Mahabouddha and Ashan areas which fall in an urban homogenous zone.
The field visits allowed the team to review the GED4ALL building attributes that will be part of the data collection on the ground. Most of them can be surveyed externally, such as the number of building levels, building condition, height of the ground floor, as well as the roof shape and material. Collecting details of the building capacity and age can be a little trickier, with the former requiring some estimation based on the building size or in the case of a hospital for example, will require the surveyor to speak with a member of staff. A proposed method for determining the building age is to assess satellite imagery capture on a certain date, for example from 2000 and identify whether or not the building in question was present or not.
Determining the lateral load resisting of a building and the material used will prove the most difficult, and will likely require additional expertise. KLL are looking to collaborate with the National Society of Earthquake Technology in Nepal, so that the surveyors can receive training from their structural engineers to help better identify the correct systems and materials used.
Thank you to KLL for their hospitality in Kathmandu and sharing some of the beautiful Nepalese culture during this quick, but very insightful trip to the capital. It is by far, no easy feat to remotely digitise some of the building footprints in the dense urban parts of the city, even with the support of current, high resolution imagery. We also anticipate that the data collection of attribute information on the ground will equally have its difficulties, but the OpenStreetMap community is great at collaborating, coming together and problem solving to ensure that our obstacles are overcome.