Digitizing North East Unguja Island and Stone Town on Zanzibar

Between August 8th and September 5th, Spatial Collective held a series of workshops at the Commission for Lands (COLA) on Zanzibar. The aims of the workshops were to train and supervise the digitization of the UAV imagery in order to create a series of detailed spatial data layers while at the same time build capacity of the Commission for Lands (COLA) and the State University of Zanzibar (SUZA) staff in QGIS.

Specifically, the objectives of the workshops were:

  1. Leverage the emerging data communities in Zanzibar and transfer knowledge and skills to local government officials and university staff, specifically the skills of project management and execution of digitization of drone imagery.
  2. Produce a building footprint dataset of the pilot areas that demonstrate high quality agreed upon by the stakeholders and that follows the coding convention put forth by the Client.

To achieve objective 1, the following workshops were held with 12 participants (10 from COLA and 2 from SUZA):

Week 1 (August 14 – August 18):

  • Introduction into project background, expectations, technology used, expected outputs and timelines
  • Workshop on QGIS basics
  • Workshop on data taxonomy
  • Workshop on digitizing rules
  • Workshop on quality control
  • Setting up the digitizing environment in QGIS
  • Start of digitizing both Stone Town and North East areas
  • Start of manual error detection
  • Fixing of errors

Week 2 (August 21 – August 25):

  • All the buildings in the pilot areas digitized
  • Write a script for automatic allocation of Building Reference Numbers
  • Workshop on error types and error detection
  • Workshop on checking layer for error detection
  • Workshop on creating a database through QGIS
  • Expansion of digitizing outside the pilot areas

Week 3 (August 28 – September 1):

  • Cleaned the pilot area dataset based on feedback from stakeholders
  • Workshop on setting up the Spatialite database
  • Workshop on managing the database (downloading, uploading, saving)
  • Lecture on characteristics of spatial data
  • Workshop on data cleaning
  • Workshop on Edge Matching
  • Preparation of tailored manuals and guidelines

Week 4 (September 4 – September 6):

  • Final handover of data and manuals
Some of the participants of the workshops

To achieve objective 2, the following datasets were produced:

  • All the structures within the 19 grids of the North Eastern pilot area were digitized
  • All the structures within Stone Town were digitized
  • Altogether 18,911 buildings were digitized (16,338 in NE Unguja and 2,573 in Stone Town)
  • The following attributes were assigned to each building:
    • Building condition (Complete, Incomplete, Foundation)
    • Problematic Capture for buildings which attribution or digitization was problematic
    • Change-set (date)
    • Building Area (square meters)
    • Unique Building Reference Numbers assigned to each building
Pilot areas digitized
Stone Town with unique Building Reference Numbers assigned

Digitizing the Outputs of the Zanzibar Mapping Initiative (Photo Blog)

Spatial Collective was hired to help the Revolutionary Government of Zanzibar and the participating local entities to collect and verify geospatial data by utilizing rectified Unmanned Aerial Vehicle (UAV) imagery provided by the Zanzibar Mapping Initiative. The Zanzibar Mapping Initiative is a collaboration of Commission of Science and Technology, Commission for Lands and the World Bank. Specifically, the aim of this assignment is to train and supervise the digitization of the UAV imagery, build the capacity of local staff and students, and produce digitized roof-print of buildings as well as road and street networks suitable for further work by the Commission for Lands and the government.

 

 

Types of Crime and Support Systems in Informal Settlements

Which types of crime are most prevalent in Soweto-Kayole and where? Which support systems for victims of crime exist within the settlement?

These are some of the questions we wanted to answer during a mapping study to support the World Bank and Kenya Informal Settlement Improvement Project (KISIP) in generating settlement specific designs aimed at linking perceptions of safety to infrastructural upgrading in informal settlements. This is the third in a series of datasets and visualizations that Spatial Collective created for this initiative. Other outputs can be found in our previous blog posts here, here, here, here, and here.

During the exercise, some 97 participants (52 women and 45 men) helped us construct and design a map showing types of crime and support systems within Soweto-Kayole.

During the focus group discussions, an interesting discussion emerged, namely, what is a crime under the law as opposed to what si an activity that is not illegal but it increases the sense of insecurity in people; and whether we should map both. We found that community’s perceptions of safety, or of what constitutes a crime, relate to both categories. Sometimes they would indicate crimes under the Kenyan law (for example murder, robbery or physical violence), other times they would point out activities that are not necessarily illegal but are perceived to increase the risk of insecurity (for example gambling or drinking). To us, it was important to address both categories – ‘true’ and so-called ‘perceived’ crimes – because of their impact on community’s mobility, economic development, people’s daily choices and access to opportunities.

The typology of crime and violence identified and defined by community members included snatching, theft, mugging, house break-in, child abuse, and gender-based violence (especially rape). Other issues such as land grabbing, gambling, and substance abuse were identified as key, yet indirect, factors contributing to crime and violence affecting the community. For instance, young women said they avoid areas where gambling takes place since these are places where idle young men congregate and they feel exposed to harassment.

Types of crime and support systems represented geographically
Types of crime and support systems presented geographically

Participants were surprised by the number of existing amenities mapped in Soweto-Kayole, which indicates that some support systems for survivors of violence do exist in the settlement. The list of existing amenities included police stations, security lights, a hospital, social halls, rehabilitation centers, religious institutions, schools, and non-governmental organizations. Nevertheless, community members noted that some of these facilities are inadequately equipped to offer anything more than basic services to victims of crime. For example, victims of sexual assault do not have a support or a referral system, and a large number of private – often informal – clinics are unlicensed, expensive, and poorly equipped. They also noted that reporting on crime and violence in the area is relatively low due to the mistrust in authorities. Young men often said they avoid the police because of harassment.

We avoid walking in large groups because of police harassment.

a youth participating in the focus group discussion

However, there was a general consensus among the participants that all forms of crime should be reported to the police. In addition, the Chief’s Camp, the social hall (Ward’s representative’s office), children’s home, and a hospital were all indicated as alternative amenities offering support.

Types of crime and support systems within Soweto-Kayole
Types of crime and support systems within Soweto-Kayole

Mapping Safe and Unsafe Areas in Informal Settlements

In our efforts to determine safe and unsafe areas of Soweto Kayole – as perceived by the residents, we first developed a base map of the area. Around twenty participants from the area mapped several hundred locations depicting amenities related to safety and security or crime. Some of the points collected were: a police station and chief’s camp, security lights, hospitals, clinics, pharmacies, chemists, social halls and rehabilitation centres, religious institutions, schools and children’s homes. This base map was further enhanced by the building extraction of the area created by Mathare youth.

To further map out perceptions of safety, we held a series of focus group discussions with fifty-two women and forty-five men. Through a well-structured interview process, they drew on the base map areas that they perceived as safe (green) or areas that were unsafe (red).

The final result is two detailed and neighbourhood specific maps depicting Safe and Unsafe Areas as Perceived by Men and Women in Kayole Soweto (detail below).

This project was done with the support of The World Bank and the Kenya Informal Settlement Improvement Project (KISIP).

MAP 2 copy

Men’s and Women’s Perceptions of Safety Related to Mobility in Informal Settlement

How are spaces in informal settlements traversed differently according to gender? Which paths are the most travelled in a community, when and why? Does gender influence perceptions of safety when it comes to movements within the informal settlement?

These are some of the questions we wanted to answer during participatory mapping to support Kenya Informal Settlement Improvement Project (KISIP) in Soweto Kayole. During the exercise, some 60 participants (35 men and 25 women) were asked to identify two locations – apart from their home – that they frequent most often. The aim of this exercise was to understand the perceptions of safety of men and women while moving through the settlement; whether in their movements to and from these locations, participants felt safe or unsafe, and why they felt a certain way. We named the method the triangle method of mobility.

dscf7473-1
The Triangle Method of Mobility

The 35 men interviewed drew 110 segments while 25 women drew 68 segments. Each segment represents a path between two locations that individual visits every day. The red colour indicates that a person felt unsafe walking along the path, while the blue colour indicates that a person felt safe walking between two locations in the settlement.

Men identified only 10% of the segments as unsafe, while women identified 34% segments as unsafe. The two circles represent all of the segments, or paths, of men and women combined.

circles
Men feel a much greater sense of safety in the community than women

Overall, all men — young and old — move much more freely within Kayole Soweto. Men of all age groups reported very few areas of insecurity in the settlement and were thus much more mobile than women. Young men, in particular, identified few unsafe areas, which implies a low perception of insecurity of this age group. However, they stated that they constantly change their routes as a conscious strategy to enhance personal security and to avoid police who often targets youth under a pretense of being associated with criminal activities. Additionally, older men added that to enhance individual safety it is better to use the main road rather than side streets.

Overall, women travelled within and outside the settlement less than men. Women of all ages reported planning their routes from one point to another in advance to avoid areas perceived as unsafe. Women mobility patterns are dynamic. They change their routes constantly as a security precaution to avoid areas perceived as insecure. They also reported that to enhance their security they walk where the streetlights are and along the main roads. Young women identified their neighbourhoods and their homes as the places they felt most safe.

Mobility of Men and Women in Informal Settlement Related to Seeking Opportunities

How do people navigate the places in which they live? Does gender influence mobility and in what way? How do men and women access opportunities within and outside the slum?

These are some of the questions we wanted to answer during participatory mapping to support Kenya Informal Settlement Improvement Project (KISIP) in Soweto Kayole. During the exercise, some 97 participants (52 women and 45 men) responded to a series of questions designed to help understand gender-specific patterns of movements in and out of Soweto, as well as gender-specific perceptions of the safety of movements within the informal settlement.

Participants were asked to first identify locations of their households and locations of work. The aim of this exercise was to understand how much men versus women leave the settlement to seek opportunities outside.

We found that men travel outside the settlement for work more than twice as often as women, in fact, 40% of men indicated that they went to work outside the boundaries of the informal settlement compared with only 20% of women (see the infographic below).

menwomen
Men travel outside the settlement for work more than twice as often as women

Linking Perceptions of Safety to Infrastructural Upgrading in Informal Settlements

Spatial Collective was hired in the spring of 2016 to develop a Geographic Information Systems-based participatory mapping approach to support Kenya Informal Settlement Improvement Project (KISIP) in developing settlement specific designs aimed at linking perceptions of safety to infrastructural upgrading in informal settlements. The project was implemented in Kayole Soweto, in collaboration with The World Bank and the Government of Kenya Ministry of Lands, Housing, and Urban Development.

Kayole Soweto is a government-sanctioned settlement located in Savannah location of Embakasi East constituency in Nairobi City County. The settlement is further divided into 10 administrative zones. Soweto is a settlement consisting of permanent and semi-permanent structures which house an estimated population of 90,000 people in approximately 22,400 households.

dscf3577-1
Kayole Soweto

As a part of the process, a mapping study took place between May 23 and 30 2016, aimed at linking perceptions of safety and insecurity to infrastructure development in order to inform the future of infrastructural upgrading in informal settlements. Focus group discussions, GPS mapping and participatory map-drawing exercises were conducted with residents of Kayole Soweto. Participants of the focus group discussions included youth, male and female, ages 18-25; reformed youth*, male and female, ages 18-25; and men and women, ages 40 and above. GPS mapping was done by 10 male and 10 female community members representing all ten zones** of the settlement.

dscf3904-1-1
Some of the participants and facilitators

Some of the questions that the study wanted to answer were: How do people navigate the places in which they live? Does gender influence mobility and in what way? How do men and women access opportunities within and outside the slum? How are spaces in informal settlements traversed differently according to gender? Which paths are the most traveled in a community, when and why? Does gender influence perceptions of safety when it comes to movements within the informal settlement? What types of crime are most prevalent in Soweto Kayole and where? Which support systems for victims of crime exist within the settlement?

After about a month of data collection, the following four maps were designed:

  • Men’s and women’s perceptions of safety related to mobility in Kayole Soweto
  • Safe and unsafe areas as perceived by men and women in Kayole Soweto
  • Types of crimes and support systems as perceived by men and women in Kayole Soweto
  • Improvements as proposed by men and women in Kayole Soweto

These represent an overlay of information obtained using a mixed-method approach including a) stakeholder consultations and participatory project planning; b) building extraction by youth using aerial imagery; c) GPS data collection by community members; d) participatory map drawing through focus group discussions with community members; e) and finally, map creation.

We will present each map separately and subsequent blog post.

* Reformed youth’ is a term coined by Kayole citizens and it represents youth whose experiences range from substance abuse and sexual work to homelessness. Participants were identified through a local rehabilitation program.

** Planned and unplanned settlements in Nairobi are usually further divided into ‘Villages’ or ‘Zones’ which represent the smallest administrative areas. Kayole Soweto is divided into 10 zones.

The Power of Community Mapping and Community Data

In June 2016, ten youth from Mathare digitized more than a thousand structures in Kayole Soweto, creating a building footprint of the area. At the same time, twenty community members from Soweto, mapped several hundreds of amenities including hospitals, clinics, pharmacies, schools, security lights, religious institutions, bars, government offices, etc. In just one week, under the mentorship of Spatial Collective, both teams created a base map of Kayole Soweto which was later used during focus group discussions addressing perceptions of safety of people in the area. More about that project coming up soon in the upcoming blogs. This initial mapping showcased the power of community mapping and community data collection and its potential to fill in the gap of missing data in development.

Press on the image below!

gif

 

Community Land Mapping – Post Fieldwork Data Management and Map Creation

The final step in community land mapping in Tana River County was to design the maps of the two targeted communities. Previous steps are described here, here and here.

To complete the two maps we used data collected from the field, including GPS files of points and tracks, two drawings made by community members of their community land, notes, and photographs. Additionally, we used DigitalGlobe’s satellite imagery to digitize relevant land features.

Specifically, the final step can be broken down into the following activities:

  • Review of data collected through fieldwork
  • Selection and purchase of relevant satellite imagery: DigitalGlobe
  • Digitization of natural features identified by the communities
  • Design of the two maps using QuantumGIS

These were also the final activities of the training process that we offered to Kenya Land Alliance and Namati on how to collect and manage community land data.

Natural resources and amenities that were either gathered in the field or digitized in post-processing and included on the maps were: rivers, brooks and streams; forests; swamps; ocean; sand dunes; grazing areas; farm lands; village locations; schools; hospitals and dispensaries; mosques and churches; water points; roads; air strip; hotels; bridges; and cattle dips.

Finally, maps that resemble Topo Sheets in Kenya were designed. We chose this style because it offers a sense of familiarity to the citizens and government agencies to whom they will be presented.

To simply indicate the amount of data gathered by our field teams, we compared it to openly available OpenStreetMap data in the area.

Comparison between existing points in OSM and points collected during field work
Comparison between existing points in OSM (left) and points collected during fieldwork (right)

Number of points in OSM: 6

Number of points collected during fieldwork: 126

Comparison between existing rivers in OSM and data generated through and post fieldwork
Comparison between existing rivers in OSM (left) and data generated through and post-fieldwork (right)

Length of rivers in OSM: 91 km

Length of rivers digitized: 381 km

Comparison between existing roads in OSM and data generated through and post fieldwork
Comparison between existing roads in OSM (left) and data generated through and post-fieldwork (right)

Length of roads in OSM: 103 km

Length of roads digitized: 110 km

Comparison between existing data in OSM (left) and data generated through and post fieldwork (right)
Comparison between existing data in OSM (left) and data generated through and post-fieldwork (right)

Areas in OSM: 17 km2

Areas digitized: 590 km2

Images below compare what was first input into the mapping process – community’s drawings of their land – and what was the final output of the process – final georeferenced maps.

Map drawn by the community members
This map was drawn by the community members
Same area as above georeferenced
Same area as above – georeferenced