Feb 082012

On the former webpages of TZGISUG we availed a host of data from different sources. Among others were data on protected areas, infrastructure, demography, biological diversity, digital elevation models, Quarter Degree Grid Cells and more. Is this something we should continue doing, or is this a job better done by the actual owners of the data?

Professionally I would say that the owners of any data sets, spatial or not, should provide their data with the associated meta data, be it ISO/TC211, ISO 19139 or other. Only by being provided by their rightfull owners will we know that the data has not been tampered with or otherwise doctored in any unfashionable manner.

Requests for data surface quite often on the TZGISUG mailing list. A researcher would need the updated district boundaries or other data sets. The response is more often than not that someone has a copy of the requested data set. It is forwarded by email, a link or by the use of Dropbox – and everyone is happy. Or…

Data are established by researchers, government bodies, NGOs and international bodies without proper standardization. Be it projections, legal mandate, description of origins, metadata or other – most likely something is missing. The consequences?

  • A lot of research in Tanzania, and for that sake other countries, rest on rather dodgy spatial data sets. Bad data makes for bad research.
  • Policies and plans are made based on the same data. Bad data makes for bad decisions.
  • Legal issues are of course also part of the equation. Illegal use of data could in the end of the day tangle up otherwise useful processes.

One of the challenges with the practice of sharing data on the TZGISUG website is that we are contributing to short-circuit what proper processes and institutions are available. We might also get ourselves into delicate legal problems. Or worse – redistribute “rotten” data sets.

One of the more central data sets in a country is the roads network. What’s it like in Tanzania? I guess we have all seen the data sets previously made available through this site. It’s origins are unclear. Maybe TanRoads (Tanzania National Roads Agency) would be able to provide an updated data set? Could it be that such data would be available on their website? A quick look is certainly not promising. I was only able to find PDF-versions of the roads data under their mandate. Their mandate being:

TanRoads is managing the National road network of about 33,891 km comprising 12,786 km of Trunk and 21,105 km of Regional roads. The remaining network of about 53,460 km of Urban, District and Feeder Roads is under the responsibility of the Prime Minister’s Office Regional Administration and Local Government (PMO-RALG) (Source: TanRoads).

So even if TanRoads had the data available this would only be part of the data. What about PMO-RALG? Looking at their website it was hard to find any references to their roads mandate. Neither did I have any success in finding roads data.

Is it the fault of TanRoads and/or PMO-RALG when they do not provide shapefiles or similar on their website? Probably not. It is their mandate to manage the road network. But it might still be outside their mandate to provide the general public with spatially referenced information about this network. Trying to figure these things out is demanding. Going all the way is difficult, and maybe even impossible. It is not what one would expect researchers, NGOers, government employees and others to do.

OpenStreetmap will provide some of these data. The Tanzania roads data set is growing and in a town like Arusha it is coming to a point where it in some areas is better than the Google maps data set. An other source for roads data is the Tracks4Africa Project. Its coverage is excellent in the southern parts of Africa.

Ideally there should be a functional National Spatial Data Infrastructure in place. In general it would look a bit like this:

Last time I read about this in a Tanzanian context was in an article by Jonas Johanson. In his discussion Johansson points to the following on funding issues for the NSDI initiative in Tanzania:

Funding has turned out to be a factor in the Tanzanian NSDI initiative. Remaining donor funds from the 2002 Population and Housing Census initially enabled the steering committee to hold a few meetings. When that source was depleted, the potential sources of funding of NSDI at the beginning of 2006 seem to be the involved organisations separately. Interesting research is going on today on the topic of national budgeting and SDI, where cross-agency budgeting for SDI have been presented as an option.

We are writing 2012 and I am still not aware of any alternatives to the list-based sharing of spatial data. I’m looking for it and looking forward to it.



  3 Responses to “TZGISUG website, a spatial data repository – or not”

  1. Changes: Updated the posting to reflect the excellent data available through the Tracks4Africa project.

  2. […] http://www.tzgisug.org/wp/?p=140 This entry was posted in tzgisug. Bookmark the permalink. ← Output coming up… Productive week → […]

  3. […] TZGISUG website, a spatial data repository or not […]

%d bloggers like this: