Roads access analysisTechnical material for Leave no one behind report

Catherine Simonet, Senior Research Officer, ODI

December 2016

Data sources Road network :

Center for International Earth Science Information Network - CIESIN - Columbia University, and Information Technology Outreach Services - ITOS - University of Georgia. 2013. Global Roads Open Access Data Set, Version 1 (gROADSv1). Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC). Accessed May 2016.

Village and settlements: This is a point coverage showing the villages in Kenya according to Almanac Characterization tool (ACT) database ILRI

Population : 2008 and 2014 DHS databases for Kenya Reference report : Kenya National Bureau of Statistics, Ministry of Health/Kenya, National AIDS Control Council/Kenya, Kenya Medical Research Institute, National Council for Population and Development/Kenya, and ICF International. 2015. Kenya Demographic and Health Survey 2014. Rockville, MD, USA: Kenya National Bureau of Statistics. Raw data: http://www.dhsprogram.com/Data/

PresenterPresentation NotesData sources shoulbe cited as mentioned in the slide

http://www.dhsprogram.com/Data/

Road Network(primary data)

The Road database provide information on

- road categories- surface types - on road lengths

Road Category LENGTH_KM CumulativedistributionHighway 3480.27 5.30%Primary 2910.02 9.72%Secondary 15603.88 33.47%Tertiary 43506.15 99.67%Local 95.33 99.81%Unspecified 121.58 100.00%Grand Total 65717.23

Road Surface LENGTH_KM CumulativedistributionPaved 6655.31 10.13%Graved 25832.98 49.44%Dirt/Sand 29979.41 95.06%Unspecified 3249.54 100.00%

!

!!

!

!

NakuruKisumu

Mombasa

NAIROBI

Eldoret

Road CategoriesHighway

Secondary

Tertiary

Tertiary and Local

Unclassified

PresenterPresentation NotesThe Road Network database provides information on the types and the categories of the road.

Here are the the categories of each variables : Road Classes : 1 "Highway" 2 "Primary" 3 "Secondary" 4 "Tertiary" 5 "Local/Urban" 6 "Trail" 0 "Unspecified" Road Surfaces :1 "Paved" 2 "Graved" 3 "Dirt/Sand" 4 "Steel" 5 "Wood" 6 "Grass" 0 "Unspecified"

Village Locations(primary data)

Information on- Lat/Lon of village locations- Name of villages/settlements- County/District location

PresenterPresentation NotesEach dot represents one village settlement

Distance calculation (1) VillagesWe identified the nearest roads for each village and the distance to this roads using a buffer analysis.

We reproduced the analysis by roads category and surface classes and by county

Key hypothesis: the centroid of the village settlement is considered as the village location. The

heterogeneity of access to road within a village/sublocation is not taking into account.

The villages are not weighted by population so this analysis only look at physical settlements/ administrative units access to roads.

Cumulative share of settlements by near distance (in meters)

0.2

.4.6

.81

Shar

e vi

llage

(pct

)

0 20000 40000 60000NEAR_DIST (in meters)

50% of the administrative settlements of Kenya are far from less than 1km to a road

1100m

0.5

An unequal distribution at the county level

Variable Nbvillages MeanStd. Dev. Min Max

Near distance (meters) 43956 3294 5800 0 62450

National distribution

PresenterPresentation NotesIn yellow is the national average

Please note that these results can be biased by the unequal number of administrative settlements by county. In county with more locations the variance of distance could be more important

An unequal access quality (1)0

.2.4

.6.8

1Sh

are

of S

ettle

men

ts in

%

0 100000 200000 300000Near distance in meters

Surface Type I Surface Type IISurface Type III

S f I i d d S f II i d d Cl III i di t/ d

Distance of settlements to Roads

Surface type I is paved roads, Surface II is graved roads, Surface III is dirt/sand and unspecified surface type.

Only 20% of the settlements are far from less than 10km to improved surface roads

PresenterPresentation NotesSurface type I is paved roads, Surface II is graved roads, Surface III is dirt/sand and unspecified surface type.

Only 20% of the settlements are far from less than 10km to improved surface roads

0.2

.4.6

.81

Shar

e of

Set

tlem

ents

(in

%)

0 50000 100000 150000 200000Near Distance (in meters)

Road Class I Road Class IIRoad Class III

An unequal access quality (2)

Class I includes highway and primary roads; Class II includes secondary roads; Class III includes tertiary local roads and unspecified classes

10 % of the settlementsare far from less than20km to primary and secondary road networks

Distribution Road access by countyvisualization to be improved

0.5

10

.51

0.5

10

.51

0.5

10

.51

0.5

1

0 20000 40000 60000 0 20000 40000 60000

0 20000 40000 60000 0 20000 40000 60000 0 20000 40000 60000 0 20000 40000 60000 0 20000 40000 60000

Baringo Bomet Bungoma Busia Embu Garissa Homa Bay

Isiolo Kajiado Kakamega Keiyo-Marakwet Kericho Kiambu Kilifi

Kirinyaga Kisii Kisumu Kitui Kwale Laikipia Lamu

Machakos Makueni Mandera Marsabit Meru Migori Mombasa

Murang'a Nairobi Nakuru Nandi Narok Nyamira Nyandarua

Nyeri Samburu Siaya Taita Taveta Tana River Tharaka Trans Nzoia

Turkana Uasin Gishu Vihiga Wajir West Pokot

Shar

e vi

llage

(in

pct)

NEAR_DIST (in meters)Graphs by COUNTY,C,20

0.5

1

0 20000 40000 60000

Turkana

Population Distribution(primary data DHS database 2014)

Information on

- Lat/Lon on cluster of sampling (1585 clusters, we removed 4 which were not georeferenced)

- 36224 Households- Region of sampling (5 regions)- Weight of each hh of the survey- Representative sample at national and regional level

PresenterPresentation NotesHere is the map for Road categories

Distance calculation (2) DHS We identified the nearest roads for each of the 400 clusters of the survey, using buffer analysis.We weight each of the household to have a national representative picture (we use the weight produced by DHS)We reproduced the analysis by roads category and surface classes and by region

Hypothesis: The centroid of the cluster is considered as the cluster location. The heterogeneity

of access to road within each cluster is not taking into account. The each cluster is weighted by population. Using DHS data, we follow DHS sampling strategy (stratification and weighting) Centroid of cluster is displaced up to 2km in urban areas and 5km in rural with 1 %

of 10km displacements. Clusters are maintain in the same regions and counties. For these reasons we strongly suggest to use with cautions all statistics using information of distance below 5km.

0

.2

.4

.6

.8

1

Shar

e of

pop

in p

ct

0 5000 10000 15000 20000 25000 30000 35000distance in meters

Source: DHS 2014 and CEISIN databases-author's calculation

Distribution of population

Cumulative share of population by near distance (in meter)

An unequal distribution at the regional level (1)

Max of Near distance (in meters)Countymaxdis

1018 - 3174

3175 - 6195

6196 - 11233

11234 - 21569

21570 - 32481

Min of Near distance (in meters)0.153509 - 1.997720

1.997721 - 12.561800

12.561801 - 39.213700

39.213701 - 79.207500

79.207501 - 159.582000

An unequal distribution at the regional level (2)

Mean of Near distance (in meters)361 - 656

657 - 941

942 - 1445

1446 - 3374

3375 - 6868

An unequal access quality (1)

Note: Dirt/Sand category encompasses other surface

0

.2

.4

.6

.8

1

Shar

e of

pop

in p

ct

0 100000 200000 300000 400000distance in meters

Dirt/Sand Roads Paved RoadsGraved Roads

Source: DHS 2014 and CEISIN databases-author's calculation

Distribution of population

An unequal access quality (2)

Note: Tertiary category encompasses others categories

0

.2

.4

.6

.8

1

Shar

e of

pop

in p

ct

0 100000 200000 300000 400000distance in meters

Dirt/Sand Roads Paved RoadsGraved Roads

Source: DHS 2014 and CEISIN databases-author's calculation

Distribution of population

0.2.4.6.81

0.2.4.6.81

0.2.4.6.81

0.2.4.6.81

0.2.4.6.81

0.2.4.6.81

0.2.4.6.81

0 50000 100000 150000 0 50000 100000 150000

0 50000 100000 150000 0 50000 100000 150000 0 50000 100000 150000 0 50000 100000 150000 0 50000 100000 150000

Baringo Bomet Bungoma Busia Elgeyo Marakwet Embu Garissa

Homa Bay Isiolo Kajiado Kakamega Kericho Kiambu Kilifi

Kirinyaga Kisii Kisumu Kitui Kwale Laikipia Lamu

Machakos Makueni Mandera Marsabit Meru Migori Mombasa

Murang'a Nairobi Nakuru Nandi Narok Nyamira Nyandarua

Nyeri Samburu Siaya Taita Taveta Tana River Tharaka-Nithi Trans-Nzoia

Turkana Uasin Gishu Vihiga Wajir West Pokot

Highway and Primary Roads Secondary RoadsTertiary Roads

Shar

e of

pop

in p

c