Harmonization of Political and Administrative Boundaries in Kenya Using GIS : A Case Study of Mavoko Constituency

The objective of this research was to harmonize political and administrative boundaries in Mavoko Constituency, Machakos County, Kenya, using geospatial technology and public consultation in order to address the numerous disputes that exist. Specifically, the project focused on analyzing the existing electoral and administrative boundaries in the area of study by identifying all existing gaps and overlaps, rectifying the identified gaps and overlaps and created a web based geo-portal of the harmonized sub location boundaries to enable sharing. This was achieved through first putting the various datasets on the same Spatial Reference System and comparing them in order to identify and rectify them using various geospatial operations, public consultation and field work. The result was a single harmonized layer for the sub location data set for the study area. The project finally developed a customized GIS web based portal for sharing the harmonized Sub location boundary. This was uploaded on an open source platform, which has the capability for offline use in the field. The project worked on a hypothesis that if the sub location boundaries adopted by the various Government and Public organizations and individuals doing any kind of mapping in Kenya was a single harmonized data set, the emerging issues on boundary disputes between Counties, Constituencies and County Assembly Wards (CAW’s) would not occur. This is because there would be authentic, consisted and authoritative geo-information which maintains a harmonized, consistent, reusable and readily available data. The development of the geo portal ensured that the results of the research are easily accessible on www.msc2015boundariesmavoko.info.ke/mavoko.


Introduction
This project reviewed administrative boundaries at the sub location level, electoral boundaries at the three levels of {County Assembly Ward (CAW), Constituency and County} and the enumeration area boundaries used by the Kenya National Bureau of Statistics (KNBS) in doing their 2009 population and housing census.The purpose of this review was to determine the specific inconsistencies in the datasets, and find ways of harmonizing them in order to solve the numerous discrepancies that exist.Kenya undertook the last boundaries delimitation in 2012 in a process that involved collection and analysis of the 2009 Kenya National Population and Housing Census, existing administrative boundaries and input from the public through public participation.This was in an effort to achieve population quota as defined in the constitution.The population quota is obtained in accordance with chapter seven article 89 section (12) of the constitution of Kenya, by dividing the national population by the number of constituencies or wards as applicable, into which Kenya is divided [1].
The main objective of the research was to harmonize political and administrative boundaries in Mavoko using geospatial techniques and public consultation in order to address the numerous disputes that exist.In specific the research sought to analyze the existing electoral and administrative boundaries in the area of study by identifying all existing gaps and overlaps then rectify the identified gaps and overlaps between the electoral and administrative boundaries.Finally a web-based geo-portal to enable sharing of the harmonized sub location boundary data was created.

Study Area
Mavoko Constituency is geographically located in Machakos County, Kenya as shown on Figure 1.It has four CAWs namely Syokimau/Mlolongo, Muthwani, Athi River and Kinanie with a total population of 139,502 persons and an approximate area of 843.20 Km 2 (National Census, 2009) [2].Mavoko Constituency borders four counties of Nairobi, Kajiado, Kiambu and Machakos in which it belongs.This makes it a perfect study area because it addresses similar problems that exist in other parts of the country.Due to its proximity to Nairobi, it has both rural and urban areas set up.The land tenure systems in the area are varied with both freehold and leasehold tenure systems which have both general and fixed boundaries.

Methods and Materials
The project identified both spatial and non-spatial data acquired for use.The area, put on a common spatial reference system and compared.Errors were checked; minor errors identified whose threshold was above the maximum allowed.Major gaps and overlaps were identified and these required to be aligned.
All the errors were solved case by case manually in the appropriate manner.The result of all these was a harmonized Sub-Location boundary for the study area.

Datasets and Data Sources
The research used various data sets including the Constitution of Kenya 2010, the District and Provinces Act, 1992, the National Government Co-ordination Act, 2013, digital Electoral maps from IEBC, digital Administrative and 1: 50,000 Topographic maps from Survey of Kenya (SOK), digital Census maps from Kenya Bureau of Statistics.Table 1 shows the various data sets and their origin metadata.

Data Preparation and Analysis
The first task was to overlay the various spatial datasets involved in the alignment and put them into a common Spatial Reference System.Cleaning and removal of unnecessary geometries and matching the corresponding objects in the datasets was also done [3].

Spatial Reference System
The two boundary data sets were in different reference system, thus there was need to harmonize them in to a common one for them to be integrated for  accurate viewing and analysis.In Arc Toolbox data management tools, projections and transformations the data were projected to geographic coordinate system WGS84 as shown in Figure 3.This was adopted in order to enable web mapping as most online platforms use this coordinate system.The IEBC data was in Arc 1960, thus it had to be projected to WGS84.

Geo Portal Design
Once the harmonized boundary was achieved, a web-based GIS portal for it was

Geo Portal Design Methodology
Once the harmonized boundary was achieved, a web-based GIS portal for it was designed and the resultant shape-file imported.After the database creation, it was linked to Qgis.This enabled the administrator to add features directly to the database.This was done by using PostGis in Qgis, database, DBmanager, and then connected to a localhost in postgres by creating a new creation in the steps shown in the methodology flow chart 2 in Figure 4.

Server Service and Spatial Database Creation
The research project used a single PostgreSQL server service to store the data.PostgreSQL 9.4 was downloaded and installed from the web site [5], in the default c/program files/postgres and port 5432.PostGIS add-on was added by using the postgreSQL application stack builder.To check whether postGIS was installed clicking the 10th icon on the pdAdmin toolbar.Mavoko database was created on the PostgreSQL server by converting the shape files to their corresponding tables in a PostgreSQL database using the import shape file plugin in PostGIS extension.This was done by right clicking databases, new database in the PgAdmin III.extension.

Create Extensions
To load a new extension into on the Mavoko database SQL plane, the following

PostGIS Connection
On the PostGIS shape-file and Dbf loader, a connection was done to the database then the shape files were added by importing them to PostgreSQL database tables using the shape file import manager plugin in PostGIS.

Tables
Tables which are the workhorses for the database for spatial functions meta-data were automatically created in the database's public schema.The SPATIAL_REF_ SYS table and the GEOMETRY_COLUMNS table.The SPATIAL_REF_SYS table holds EPSG IDs and descriptions of the coordinate systems used in the spatial database [7].By right clicking on the table, it was confirmed that the geometry column was present as shown below in Figure 6, thus a successful upload.
These tables contain all the attributes of the imported shape files including the coordinates and geometry that define each feature.

Boundary Harmonization
Once the datasets were acquired they were put under a common special reference system as shown in Figure 7.The research identified several minor errors as shown in Figure 8 and corrected one by one.The PDA 2012 was used to harmonize the county boundaries.The existing Harmonized Boundary Maps

County Boundaries Harmonization.
Kenya Counties were established under the PDA 2012, this act was used to replot

Harmonization of Constituency Boundaries Using Existing Municipal Council Boundary
The research realized that the two departments have different opinion on the boundary between adjacent administrative regions which make up the Muthwa-    show the boundary after rectification.

Harmonization of CAW Boundaries
The boundary between Kinanie and Muthwani CAW's had a gap, resulting into a disputed block as shown on Figure 14.In order to align the two boundary data sets, a field survey was carried out.This involved public consultation of the various stakeholders in the area, including but not limited to the area administrative officers, the school head teachers and their education counterparts, the IEBC Constituency Elections Coordinators, the County Surveyors and some residents.

A Web Application
The developed Web GIS portal is able to dispatch accurate information on time and consists of various useful GIS tools for spatial data visualization, analysis, querying and finally download data for use in mapping as base data.For the system to access the database, a Bitnami Wampstack Apache was downloaded and

Conclusions
The study led to the creation of a single sub location layer for the Mavoko Constituency.The Geo-Portal created was able to facilitate basic functionalities of search and data retrieval in simple and fast data access, data analysis and re use of the sub location data in order to fulfill various functions and so as to suit these objectives.Therefore the research used a simple interface that would allow for easy information dissemination to all.Three categories of users were envisioned: 1) Those who would want to know boundary information in relation to places.2) Those who would want to analyze the boundaries therefore download the data.3) Those who would want to add more information into the platform.

Recommendation and Future Research
The Kenya National Spatial Data Infrastructure (KNSDI) set up standards for carrying out Geo spatial work in Kenya.However very little was set in terms of alignment on boundaries by the various government departments in carrying out the various tasks.The research recommends that the IEBC, Kenya International Boundaries Office (KIBO), the Survey of Kenya (SoK), KNBS and the relevant department in state department of interior to work together and come up with a harmonized data set for the country.The methodology on boundary harmonization adopted in this research if adopted for this task can lead to a harmonized boundary in Kenya.Once the sub location data set is created, checked and authenticated then it can be used in future by the various public and private bodies or individual for any kind of mapping as base data.This will ensure harmony in the data used and avoid conflicts of spatial data hence boundary dispute experienced in many parts of the country.Some unique challenges could be experienced in areas with unique case of conflict like the international boundary in some instances and some historical issues that require negotiations

Figure 2
below shows the first part of the two part methodology undertaken in the research.

Figure 3 .
Figure 3. Boundaries after the harmonization of the spatial reference system.

Figure 4 .
Figure 4. Overview of the methodology flow chart 2.

Figure 6 .
Figure 6.Displaying the spatial database attributes of the shape files imported.

Figure 9
Figure 9 were used to ascertain the correct boundaries between constituencies.This is because most constituency boundaries follow the already defined Municipal and County Boundaries.

Figure 7 .
Figure 7. Boundaries after the harmonization of the spatial reference system.
ni and Mathatani CAW's, Kinanie and Muthwani CAW's, Syokimau and Ngelani CAW's and Athi River Township and Athi River North CAW's.To harmonize these boundaries, the research adopted the Harmonized Boundary (HB) Plan No. 22 that covers Machakos Municipality, HB 24(C) covering Mavoko Municipality and HB 24(D) covering Kangundo Town Council.These maps clearelyshowed the extent of each town, thus aking it possible to harmonize the boundaries.Figure12shows the Boundary overlaid on the HB Plans and Figure13

Figure 10 .
Figure 10.Harmonized boundary plotted using PDA Act 2012 on topographical maps.

Figure 11 .
Figure 11.Harmonized boundary plotted using PDA Act 2012 on ArcGIS online satellite image.

Figure 12 .
Figure 12.Overllapping boundary between Mavoko and Matungulu constituencies at Joska trading centre overlaid on existing HB maps.
The boundary was picked as described in the field and plotted as shown on Figure 15.

Figure 13 .
Figure 13.Harmonization of constituency boundaries using existing HB maps.

Figure 14 .
Figure 14.Disputed CAW boundary resulting into a disputed block.

Figure 15 .
Figure 15.Harmonized CAW boundary as a result of field survey.

Figure 16 .
Figure 16.The Geo-portal log in page.

Figure 17 .
Figure 17.Proposed mavoko constituency harmonized boundary on an open street base map.

Table 1 .
Datasets and data sources.