Background...
Accra, the capital of Ghana, is a city that is beleaguered by perennial flooding, with peaks during the rainy season during June and July.
In June. 2015 for example, a heavy rain, an explosion and a fire occurred at a petrol station in Accra, reportedly killing over 250 people, many of whom had gone there to take shelter from the flooding and rain. Devastating floods hit areas around Accra and claimed the lives of a number of persons after heavy rainfall from 18 June 2018. The floods occur year in, year out.
In an attempt to resolve this menace, the World Bank through its Greater Accra
Resilience and Integrated Development (GARID) Project is working to support resilient, clean, and inclusive
development in the entire Greater Accra Region.
Open Cities Accra Project...
The World Bank Group, Open Data for Resilience Initiative (ODRI) and Global Facility for Disaster Reduction and Recovery (GFDDR) initiated the Open Cities Africa Project in which Accra is a beneficiary city, hence, Open Cities Accra.
Open Cities Accra is to support the GARID project through the creation of
map data to inform participatory solid waste management, and informal
settlement upgrading, in these flood prone areas. The focus of the Open Cities Accra project is on Alogboshie, Akweteyman, Alajo and Nima.
My experience...
In July 2018 I traveled to the capital city from Kumasi to participate in the Open Cities Accra Project. I based at the offices of Mobile Web Ghana. I was eager to start so I probably arrived quite early.
The following week after I arrived, other interested volunteers were invited to an interview and subsequent selection to participate in the project. Thereafter, we underwent a week long training. We were introduced to OpenStreetMap and Geographical Information Systems in general. Enock Seth Nyamador (OSM Ghana) took us into details about OpenStreetMap and how to contribute remotely using the iD Editor and JOSM. I supported in teaching the beginners because I was well grounded in that, I had been mapping over a year by that time.
We also had David, a HOT representative from Uganda, who came to train us on the field mapping exercise. We learnt how to use Open Data Kit (ODK) Collect and Open Map Kit (OMK) to collect OpenStreetMap data offline, had experimental sessions mapping buildings (adding attributes), drains, and Points of Interests (POIs). We fine tuned the rough edges, dotted our Is and crossed our Ts and off to Alogboshie having done with remote mapping of the area.
At Alogboshie we collected attribute data on all buildings, drains, shops, schools, kiosks. Almost everything. I noticed that, as a matter of necessity, people living in the area which was mostly affected by flooding had adapted mechanisms to reduce their vulnerability to floods. Most of the buildings in that area had raised their foundations a few meters above ground level. Astonishingly, people living in wooden kiosks had raised their abodes about about 3-4 feet above ground level like Nzulezu without the water. Why don't they just abandon such a place?
We completed the field mapping in about 10 days of work. For most of the mappers their work had come to an end, mine wasn't. It was time for data cleaning to ensure quality assurance of the data we were going to upload to OpenStreetMap. The tediousness in that work? The least I said the better.
It was through the data cleaning exercise that I noticed some of the challenges we faced during the remote mapping and field mapping. They were:
1. Omission: Some of the buildings in the area were not mapped. They were either covered by a tree or they had no roof so the remote mapper couldn't notice there was a building.
2. Homogeneity: There was a lack of uniformity in the data collected and that had to do with the names of places. Different mappers could spell the name of a place differently based on their own conception of how they think it should be spelt. One mapper could also give different spelling for the same place. Typical example 'ABC Benjilo' (abc/ A.B.C Benjilo/ABC bengilo/etc) and 'Begyewahome' (you can only imagine the variations of this).
3. No name: This was also quite a challenge. You'll come across a tailor, provision store, or any other shop and it has no name. Amenities without names.
In January 2019, the same team of mappers went to Akweteyman where we only mapped POIs and the drains.
We also had David, a HOT representative from Uganda, who came to train us on the field mapping exercise. We learnt how to use Open Data Kit (ODK) Collect and Open Map Kit (OMK) to collect OpenStreetMap data offline, had experimental sessions mapping buildings (adding attributes), drains, and Points of Interests (POIs). We fine tuned the rough edges, dotted our Is and crossed our Ts and off to Alogboshie having done with remote mapping of the area.
 |
| David (HOT) making his presentation |
At Alogboshie we collected attribute data on all buildings, drains, shops, schools, kiosks. Almost everything. I noticed that, as a matter of necessity, people living in the area which was mostly affected by flooding had adapted mechanisms to reduce their vulnerability to floods. Most of the buildings in that area had raised their foundations a few meters above ground level. Astonishingly, people living in wooden kiosks had raised their abodes about about 3-4 feet above ground level like Nzulezu without the water. Why don't they just abandon such a place?
We completed the field mapping in about 10 days of work. For most of the mappers their work had come to an end, mine wasn't. It was time for data cleaning to ensure quality assurance of the data we were going to upload to OpenStreetMap. The tediousness in that work? The least I said the better.
It was through the data cleaning exercise that I noticed some of the challenges we faced during the remote mapping and field mapping. They were:
1. Omission: Some of the buildings in the area were not mapped. They were either covered by a tree or they had no roof so the remote mapper couldn't notice there was a building.
2. Homogeneity: There was a lack of uniformity in the data collected and that had to do with the names of places. Different mappers could spell the name of a place differently based on their own conception of how they think it should be spelt. One mapper could also give different spelling for the same place. Typical example 'ABC Benjilo' (abc/ A.B.C Benjilo/ABC bengilo/etc) and 'Begyewahome' (you can only imagine the variations of this).
3. No name: This was also quite a challenge. You'll come across a tailor, provision store, or any other shop and it has no name. Amenities without names.
In January 2019, the same team of mappers went to Akweteyman where we only mapped POIs and the drains.
The Open Cities Accra project so far has been a tremendous adventure and a great learning experience which has increased my capacity in the creation of geographic data. It has also given me the opportunity to meet amazing people and made new friendships out of the lot. - Samuel Darkwah Manu
 |
| The Open Cities Accra Team |
Good! Keep it up
ReplyDeleteSuper adventure and eye opening
ReplyDeleteGreat job