Plan COD-AB

Planning

Working on common operational datasets (CODs) can be a time-consuming activity, developing a clear plan with a detailed goal can help everyone focus their limited time and resources. Before undertaking any COD related activity first develop a clear goal (SMART - specific, measurable, achievable, relevant and time-bound). Once clear goals and objectives are identified with partners (IM Network) work can begin on the identification of sources and datasets.

IM Network - Partners

Relationships are a key part of the COD cycle and just as important and the datasets. IM Network partners should work together to develop their plan for all CODs including prioritisation of datasets. Partners are also vital resources in understanding the intricacies of various datasets and government agencies.

Identification of datasets

The most authoritative source for the COD-administrative boundaries (AB) is the national government. UN agencies and NGOs are allowed to operate in any country only with the permission of the government therefore if, at all possible, the national government should be the source of administrative boundaries, names and codes.

The governmental agencies/ministries responsible for Administrative Boundaries will vary in each country (e.g., Council for Development and Reconstruction, Population Census Office, Agency of Land Affairs, Construction, Geodesy, and Cartography Office, Bureau of Statistics, etc.) In some countries, several governmental agencies may claim authority and provincial governments may have different boundaries than at the federal level.  Keep in mind that different agencies might have different versions of administrative boundaries for different purposes (e.g., elections, emergency response region, economic zones, or for general cartographic purposes at varying scales); research should be done to understand the various formats and their purpose to identify the most appropriate COD-AB.

List from SALB for points on contact within a country will be available Q4 2017


StepStep DetailsTips
1. Identify Partners and develop goalsDetermine if appropriate partners are partners are part of your IM Network

Work with IM Network partners to determine your collective goals.
  • Why do you need this dataset?
  • What will it be used for?
  • Who will use it?
  • What is the purpose of the dataset (cartographic, database, online data collection)?
A possible goal for administrative boundary:
All organisations responding to the humanitarian emergency are using the same AB down dataset to level x by xx date
  • Ask existing partners if others should be added to the IM Network. For AB the appropriate government department technical staff could be invited to an IM Network meeting.
  • Many partners will have different goals for the dataset. Make sure all partners understand the needs of others.
  • Start by asking what partners are currently using and why?
  • Understand the implications of changing AB.
  • Having a functioning IM Network prior to the development of goals and objectives would be ideal.
  • Consider a specific working group to deal with core datasets like Administrative boundaries
2. Develop objectives and strategies with partnersWork with IM Network partners to determine objectives and strategies to reach your goal(s).
Possible objectives for administrative boundaries:
a. Ensure that the most appropriate AB dataset is available to all actors. 

b. Persuade all entities responding in this country to use the same administrative boundaries,names and codes.
  • Use your IM Network to determine the most appropriate objectives and strategies.
  • With an agreed goal develop objectives and strategies. This does not need to be detailed effort but should help frame the investigate of sources and datasets.
3. Identify sources Start with your IM Network, ask which agencies should be contacted.

Prepare a table with detailed metadata for every dataset 

Share this table with your IM Network

Ask IM Network to contribute metadata and contact information of the Administrative Boundaries in the table if they are using different sources of data

Review metadata and current knowledge of datasets to see if a short list can be created
  • Consider this an investigation, there will be multiple data sources for admin boundaries, be clear on your ultimate goal while investigating these sources.
  • During your investigation acquire as much metadata as possible and keep a data trail.
  • The government department responsible should always be the first choice. 
  • Ask data sources to participate in the IM Network (e.g. explain data needs, provide updates, process, build relationships)
  • If there are multiple sources available compile metadata and contact information about all available administrative boundaries in a table for easy analysis  (Example of table) (use a google spreadsheet or Dropbox for more privacynot to be public)
    • The "best" dataset is useless if it cannot be shared with all the partners.
    • A "dirty" dataset that can be shared can be cleaned and brought up to minimum standards.
    • Not ideal, but multiple admin boundary datasets can be stitched together if necessary.


Suggested Partners for Administrative Boundary COD

EntityReasons
Government
  • Only authoritative source for AB data in a country

  • Ideally, the AB used by the humanitarian community is the same as the AB used by the government agencies responding to the crisis. However, this is not always possible due to a range of issues. If the government datasets cannot be used consider the alternatives below.

UN Agencies, INGOs, NGOs
  • GIS officers and IM staff that have been working in the country are often the most knowledgeable on the AB situation in the country (which government agencies are responsible and which are willing to share)

  • May already have contacts with the government

  • May have resources to help process data

Universities or research institutions
  • May have knowledge on AB situation in the country

  • May have capacity to help process data
    Note: OCHA did this successfully in Mali. OCHA partnered with a University that had a GIS department which was able to provide useful support resources. This case study will be included in the TSP as a good partnership example.

ITOS
  • Can help process data (clean, standardize, create multiple format, provide evaluations and web service)

  • The datasets ITOS is working on and the web services they are developing will be made available on HDX

Technical NGOs (GIS Corps, MapAction, OSM)
  • Can help process during a response

  • To be discussed in the TSP

IMWG at global level (compiled list)
  • May already have contacts with the government

  • May already have data

  • May have resources to help process data

  • Field IMOs should be aware they can contact the global level and who at the global level they can contact.



Alternatives to government sources

When considering alternative sources to government administrative boundaries below are a few questions that should be asked. Case studies of COD-AB related to this can be found in resources. 

  1. If government data is not available (or shareable), what are operational (including development) partners using?  
  2. Be aware that using a non-government source could impact the relations with the government?
  3. Will exposing an alternate source push the government to share data?
  4. Is the government likely to release data if an emergency does occur, requiring the community to quickly analyze, clean, and share a new dataset? (contact information could be a good start rather than identifying a new dataset)?
  5. Why is this source authoritative?
  6. Why is the best available dataset?
Somewhat Authoritative AdvantagesDisadvantages Purpose and Process
Global Administrative Unit Layers (GAUL) Endorsed by national governmentsNot as regularly updated


Not as easy to access as GADM

 
Global Administrative Areas (GADM)More spatially accurate than GAUL

Updated regularly

Doesn't include the historical or conflicting overlap documentation that GAUL does


Not officially endorsed by national governments


Data source not always clear

 MapAction uses these when CODs not available
Second Administrative Level Boundaries (SALB) Provides admin level 1 and 2Not updated since 2014. New version to be released Q3 2017 
WHO polio health boundaries
(to be requested)
Provides  Adm1, 2 

Health-related boundaries for the entire world.

Most comprehensive and up to date dataset.

Captures historical changes in a single dataset right from 2000 for most of the countries.

WHO is the legal entity and not sure how this one can be shared outside WHO.  Maintained and updated every month. 

UNCS Level 0National boundaries edge matched

UN member state approved
only available as a web mapping service


admin 1 and 2 will be available later this year.

 Official UN international boundary
 U.S. Department of State - Humanitarian Information Unit Possible global large scale international boundary source 
Available as API
 Need more information about data sources used 
Not Authoritative
(use with caution)
AdvantagesDisadvantages Purpose and process
 Open Street MapData is open and can easily be downloadedOriginal sources not always known 
 Google Maps Freely availableShould only be used as a reference Reference

Resources


Case Study
In some situations alternative sources may be required, for instance when: government data is not be available, if there is no data for an area of the country (e.g., see Nigeria example where E-Health data was combined with government data for the Adm3 COD-AB for Borno state); the government is not able to share for various reasons and a COD-AB is required quickly, etc. 

How to deal with non-coherent national datasets coming from government - or how to find the right person in the government? 
SALB will probably have contact info for every country, but really only the IMOs are in the best position. Additional advice: cross-reference such information when there are multiple versions. If there are multiple choices (sources), choose the one that can be joined to the population statistics. Emphasis is on OPERATIONAL datasets.

ROAP - Bangladesh (See Janet or Tom)
Colombia  and Mali (partner with universities)