Rural Access Indicator (RAI) Results

Data Sources

• Population: WorldPop global per country for year 2019
• Roads: MapBox OSM QA Tiles
• Urban Areas: GRUMP Urban Extent Polygons v1.01
• Country Borders: Natural Earth Admin 0 - Countries v4.01

Release Artifacts

RAI v3 GeoJson FeatureCollection (rai-v3.0.1.geojson.zip)

Includes summary statistics for each country in the analysis as a GeoJson FeatureCollection

This geojson file was converted to a vector tile layer with layer name data available at:

https://un-sdg.s3.amazonaws.com/tiles/cardno-world-summary/{z}/{x}/{y}.pbf

Each feature includes all of the same properties as the geojson file.

Forgotten Population Raster Layer

An image tile layer is available for each country in the analysis at:

https://un-sdg.s3.amazonaws.com/tiles/cardno/{countryCode}/forgotten-pop-global/{z}/{x}/{y}.png

Each tile layer uses the same color ramp with the same breaks, defined here.

Here is the Nepal layer as an example: https://tilejson.io/g/43323e5629561a734dc8aadd4a5bcea8/view

OSM Roads Vector Tile Layer

A vector tile layer is available for each country in the analysis at:

https://un-sdg.s3.amazonaws.com/tiles/cardno/{countryCode}/roads/{z}/{x}/{y}.mvt

It is valid for zoom levels 6 - 10 with a layer name roads.

• At zoom 10, all roads from the MapBox QA tiles are displayed
• At zoom 8 - 9, only the roads used in the mask are displayed
• At zoom 6 - 7, only roads with the highway tag values: primary, secondary, trunk

Each feature has the following properties:

• highway (String): Value of the OSM highway tag
• surface (String): Value of the OSM surface tag
• isIncluded (Boolean): true if this feature was used in the mask, otherwise false

Malawi, Myanmar and Nepal Alternative RAI Approach

For these three countries, we ran an alternative analysis using a refined RAI approach. Statistics for these countries for this alternate approach are available below as rai-v3-alternate.geojson.zip.

The countryCodes are:

• MWI: Malawi
• MMR: Myanmar
• NPL: Nepal

A forgotten population image tile layer for this calculation is available at:

https://un-sdg.s3.amazonaws.com/tiles/cardno/{countryCode}/country-data/{z}/{x}/{y}.png

Each tile layer uses the same color ramp with the same breaks, defined here.

A vector tile layer of the roads used in this calculation is available at:

https://un-sdg.s3.amazonaws.com/tiles/cardno/{countryCode}/country-data-roads/{z}/{x}/{y}.pbf

These layers are valid for zoom levels 0 - 11 with a layer name data.

Filters

OSM Roads are included in the mask if:

• OSM highway tag is one of primary, secondary, tertiary, trunk

OSM Roads are included in the roads vector tile layer at max resolution if:

• OSM highway tag is NOT one of path, steps, bridleway, footway, track (road is not motor road)
• OSM surface tag is one of paved, asphalt, concrete, concrete:lanes, concrete:plates, paving_stones,
  sett, unhewn_cobblestone, cobblestone, metal, wood

Methodology

• Population is read from WorldPop in original projection
• Roads are filtered using above tags
• Road buffers are buffered in UTM projection to 2km
• Road buffers are rasterized to masks
• GRUMP polygons are rasterized to masks
• Population is counted as zonal summary over WorldPop using above masks
• Results attached to country boundaries from Natural Earth Admin 0 dataset

All weather roads only

Data Sources

• Population: WorldPop global per country for year 2019
• Roads: MapBox OSM QA Tiles
• Urban Areas: GRUMP Urban Extent Polygons v1.01
• Country Borders: Natural Earth Admin 0 - Countries v4.01

Filters

sdg-911-v2.geojson

OSM Roads selected if:

• OSM highway tag is NOT one of path, steps, bridleway, footway, track (road is not motor road)
• OSM surface tag is one of paved, asphalt, concrete, concrete:lanes, concrete:plates, paving_stones,
  sett, unhewn_cobblestone, cobblestone, metal, wood

sdg-911-all-roads.geojson

OSM Roads selected if:

• OSM highway tag is NOT one of path, steps, bridleway, footway, track (road is not motor road)

This output is included for apples-to-apples comparison.

Methodology

• Population is read from WorldPop in original projection
• Roads are filtered using above tags
• Road buffers are buffered in UTM projection to 2km
• Road buffers are rasterized to masks
• GRUMP polygons are rasterized to masks
• Population is counted as zonal summary over WorldPop using above masks
• Results attached to country boundaries from Natural Earth Admin 0 dataset