Humanitarian OSM Team/Core Impact Area Datasets , Use cases & Data Quality Metrics
Jump to navigation
Jump to search
Core Impact Area Datasets & Use cases
Data use case | Impact Area | Examples | Sample services | Datasets | What's important for this use case? |
---|---|---|---|---|---|
Urban planning: Urban landscapes are mapped for informed urban planning |
Sustainable Communities | Public transportation planning Administrative navigation for census (place names) Urban planning/zonning analysis.(residential vs commercial, green space, etc) Population/demographic estimations for service delivery Education service delively planning (school placement) Existing road conditions, road planning Informal Settlements - locations and inclusion |
Populated places | Names of subdivisions within communities and urban areas - wards, neighborhoods, cartiers, etc. | |
Buildings | Overall coverage, and change over time | ||||
Roads | Coverage of paved / well maintained roads | ||||
Public transport | Types, routes, and accessibility | ||||
Administrative boundaries | |||||
Access to services: How to access and improve the availability of basic services |
Accessibility of: education, health, clean water, sanitation Functioning of: [public] transport systems, solid waste disposal, drainage |
Education facilites | |||
Water Points | |||||
Sanitation | |||||
Railways | |||||
Logistics planning for aid delivery and evacuation(s). | Disasters & Climate Resilience | Road datasets are used for planning during relief material delivery and evacuation services during disaster response. Distance (road length & type) can be used to calculate time and fuel for evacuation. |
https://innovation.wfp.org/project/humanitarian-topographic-atlas | Roads | Access constraints - where can certain types of vehicles pass? Need to know 'surface' (paved or not, all weather or seasonal), secondary 'width'. |
Awareness of locations of population centers, and routing towards those | Place Names | ||||
Locate impacted areas and damaged infrastructure | Identification of population at risk or affected. Extent of disaster and infrastructure (shelter, services, etc) impacted. | Buildings | |||
Risk analyses to identify vulnerable infrastructure and population | Buildings, roads, waterways, land use etc create risk maps - help identify pre-disaster work and prioritize recovery Fire hazard analysis Flood hazard analysis |
Waterways | |||
Disaster preparedness and mitigation - infrastructure, coping capacity, etc | Flood resilience Evacuation centers At risk infrastructure & mitigation |
Waterways | |||
Evacuation Centers | |||||
Buildings & highways | |||||
Increase resiliency by understanding socio-economic climate impacts | Drought Livelihoods |
||||
Accessibility of health facilities | Public Health | Spatial distribution of health facilities is critical in decision making while planning on where to allocate and build facilities Catchment areas and barriers - Open Routing Services (ORS) use OSM roads to calculate and create isochrones indicating the locations of people who can access the health facility in a specified time. Speciality tag is used to tell which health facilities peovide specific specialities like antenatal care and vaccination services. |
https://openrouteservice.org/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9113998/ |
Health Facilities |
|
Roads |
|||||
Healthcare programming logistics planning & surveillance (vaccination delivery, malarial campaigns, patient origin tracing, etc). | Patient origins and contact tracing, Operational planning for antenatal care, and vaccination services. Place names are used for tracing the origins of the patients in case they need to refer or evacuate them using ambulance services |
https://www.hotosm.org/updates/piloting-tanzanias-first-patient-origin-tracking-system/ | Place Names |
||
Camp service availability analysis | Displacement & Safe Migration |
Refugee/IDP site locations Markets and cash based assistance (Cellular) networks Available public services like health facilities, sanitation facilities, water points, that will be used by the refugees |
Refugee sites | ||
Health Facilities | |||||
Markets | |||||
Water Points | |||||
Shelter/Buildings | |||||
General Social services (also in urban/non-camp settings) | Food, shelters, etc open to migrants - border & disaster adjacent regions? Livelihoods Legal status & services |
Shelter/Buildings Wash/sanitation facilities |
|||
Gender | |||||
Gender | Teenage pregnancy healthcare | Health Facilities | |||
Gender-based violence & FGM | Shelter/Buildings | ||||
Accessibility of HIV Facilities | Social Facilities |
Data Quality Metrics
Dataset | Data representation | Data Quality category | # | Measurement | Metric | Metric type | |
---|---|---|---|---|---|---|---|
Roads | Data type: area or way Primary key: highway=motorway | primary | secondary | tertiary | unclassified | residential Secondary keys: surface=* lanes=<int> name=<str> |
Completeness | Roads-1 | Number of kilometers of roads as compared to buildings in the same area. |
Proposal: “percentage of buildings within <x> meters of a maintained road (highway=unclassified or higher)” (direct link to SDG indicator 9.1.1 - https://unstats.un.org/sdgs/metadata/ ) |
Data gap indicator | |
Roads-2 | Number of major roads (motorway, primary, secondary, tertiary) with an end node. | Proposal: “Major roads (motorway, primary, secondary, tertiary) that suddenly transition or end: 1, transition to a classification of unclassified or lower; 2, that end and don’t connect to another highway; 3, orphaned nodes”. | Direct mapper feedback/Data gap indicator | ||||
Roads-3 | Comparative coverage vs third party datasets | Proposal: Use the Kontur analysis & request more raw data insights from them | Data gap indicator | ||||
Roads-4 | Number of nodes per segment | ||||||
Semantic Accuracy | Roads-5 | How long does the highway change a tag without a junction? | |||||
Roads-6 | Segmented roads with opposite directions - that are tagged as `oneway` | ||||||
Roads-7 | Valid value for `highway` tag | number of highways with bad tag values | Direct mapper feedback | ||||
Roads-8 | Roads with a `surface` tag | Data gap indicator | |||||
Waterways | Data type: way Primary key: waterway=river | stream | canal | ditch | drain |
Positional Accuracy | Waterways-1 | Comparing water flow and altitude - water flows to low lands | |||
Waterways-2 | Coarse tracing. Define what's a reasonable metric? - Comparing waterways geometries/shapes with other datasets - Long distance between nodes - and sharp angles - (Too) Long segments |
||||||
Waterways-3 | Segmented waterways (with the same value for `waterway`) with opposite directions | ||||||
Completeness | Waterways-5 | Comparative study from hydrological models | |||||
Buildings | Data type: area Primary key: building=yes (TM/remote) building=<value list> Additional data: building:levels building:material addr:housenumber=* addr:street=<str> |
Positional Accuracy | Buildings-1 | Count of buildings overlapping other features | Number of bad geometries (overlaps) | - Direct mapper feedback | |
Buildings-2 | Flagging unrealistic intersections like waterways and railway line crossing buildings | Number of bad geometries (overlaps) | - Direct mapper feedback | ||||
Buildings-3 | Measure OSM history for geometry adjustment. | ||||||
Buildings-4 | Unsquared tracing on squared buildings | ||||||
Buildings-5 | Advanced auto-derived offset from strava heatmap | ||||||
Buildings-6 | Percentage of OSM data deviation against GPS traces | ||||||
Completeness | Buildings-7 | ||||||
Semantic Accuracy | Buildings-8 | Invalid values - such as building=building or building=no | |||||
Water Points Water management |
Data type: node Primary key: amenity=drinking_water | water_point man_made=water_well | water_tap | borehole Secondary keys: operational_status=* |
Semantic Accuracy | Waterpoints-1 | Minimum tag requirements for the water points data models | Number of features with incomplete/required tags | - Direct mapper feedback | |
Completeness | Waterpoints-2 | Density of water points in relation to buildings/population | # of water points per 1,000 inhabitants (adjust scale - perhaps 10k or 100k yields better outputs?) Heatmap. Inhabitants calculated from bulidings * a country's average household size. Group by adm2 or adm3 boundaries? |
- Data gap indicator | |||
Sanitation | Data type: node Primary key: amenity=toilets Secondary keys: fee=* operator=* access=* |
Positional Accuracy | Sanitation-1 | Density of sanitation facilities in urban areas | - Data gap indicator | ||
Completeness | Sanitation-2 | Health facilities with no sanitation within close proximity (x meters) | Number of health facilities with no sanitation facilities with in xx meters | - Data gap indicator | |||
Semantic Accuracy | Sanitation-3 | Minimum tag requirements for saninitation data models | |||||
Place Names | Data type: node, area, relation Primary key: place=* Additional data: name=* |
Semantic Accuracy | Place-1 | Use of local languages in tagging Place names vs offical administrative names |
|||
Place-2 | Place=* must be accompanied by a `name` tag | Number of features with incomplete tags | - Direct mapper feedback | ||||
Positional Accuracy | Place-3 | ||||||
Completeness | Place-4 | Comparative gap analysis between data from government agencies that are mandated to map boundaries. | Differences in admin units from OSM data and authoritative data | - Data gap indicator | |||
Place-5 | Spatial distribution of place names and data points | ||||||
Place-6 | Map labelling information for place names | ||||||
Refugee sites | Data type: point, area Primary key: amenity=refugee_site Secondary keys: name=<str> |
Completeness | Refugee-1 | Shelter - compare to UNHCR data & estimated IDP & refugee populations | |||
Refugee-2 | Coverage and completeness of (formal) refugee and IDP camps (compare to UNHCR camp sites - https://im.unhcr.org/apps/sitemapping/#/ ) |
Completeness of sites with `amenity=refugee_site` - including name tag | |||||
Semantic Accuracy | Refugee-3 | Minimum tag requirements | Number of features with incomplete tags | - Direct mapper feedback | |||
Health Facilities | Data type: node, area Primary key: amenity=clinic | doctors | hospital | pharmacy Secondary keys: name=<str> healthcare=<str> emergency=yes | no opening_hours=* phone=* |
Completeness | Health-1 | Percentage of representation in comparison datasets provided by the government or other agencies like UN | |||
Health-2 | Health sites coverage in populated areas (estimated catchment area & density per population?) | # of health facilities per 10,000 inhabitants (adjust scale - perhaps 100k or 1m yields better outputs?) Heatmap. Inhabitants calculated from buildings * a country's average household size. Group by adm2 or adm3 boundaries? |
- Data gap indicator | ||||
Semantic Accuracy | Health-3 | Comparative measure of minimum attributes using data models | Number of features with incomplete tags Number of features with bad tag values |
- Direct mapper feedback | |||
Health-4 | Percentage of hospitals and clinics with a `name` | Number of health facilities with and without required name tag expressed as percentages | - Data gap indicator | ||||
Completeness | Health-5 | ||||||
Education facilites | Data type: node, area (to denote school grounds) Primary key: amenity=school | kindergarten | college | university Secondary keys: isced:level=<int> name=<str> operator=<str> landuse=education (for an area) |
Completeness | Education-1 | Comparing education facilities with population | Number of administrative units with populations missing the required number of education facilities | - Data gap indicator | |
Semantic Accuracy | |||||||
Positional Accuracy | Education-2 | Deviation of "amenity=school" from "building=school/classroom" | |||||
Waste disposal Solid waste management sites |
Data type: node, area (landuse=landfill) Primary key: amenity=waste_disposal|waste_basket|recycling|waste_transfer_station landuse=landfill Secondary keys: waste=<str> access=<str> |
Completeness | waste-1 | Minimum tag requirements for waste_disposals | Number of features (nodes) with incomplete/required tags | - Direct mapper feedback | |
waste-2 | Bad value (tagging waste=* with values off the data model) | Number of features (nodes) with bad values | - Direct mapper feedback | ||||
Administrative boundaries | Data type: area (relational) Primary key: admin_level=2:10 name=<str> Secondary keys: boundary=<str> |
Completness | Admin-1 | Comparative analysis of admin boundaries in OSM with other authoritative boundaries | Differences in number of statistical unitis | - Data gap indicator | |
Admin-2 | Comparative analysis of admin boundaries in OSM with other authoritative boundaries | Differences in the geometries and shapes of authoritative boundaries | - Data gap indicator | ||||
Public Transport | Data type: way Primary key: public_transport = platform | stop_position | station| stop_areas route=<str> highway=bus_stop Additional keys: name=<str> |
Incompleteness | Public_transport-1 | Minimum tag requirements for public_transport | Number of features with incomplete tags | - Direct mapper feedback | |
Logical consistency | Public_transport-2 | route=* matching the logical location of the features | Mis-matching tags for the logical features | - Direct mapper feedback | |||
Financial Services & Markets | Data type: node, area (market buildings/grounds) Primary key: amenity=marketplace Secondary keys: name=<str> opening_hours=* |
Semantic Accuracy | Markets-1 | Minimum tag requirements for marketplaces | Number of features with incomplete tags | - Direct mapper feedback | |
Landuse | Data type: node, area Primary key: landuse=* Other keys: amenity=* leisure=* |
Positional Accuracy | Landuse-1 | Unrealistic overlaps like institutional overlapping industrial, cemetary overlapping residential | Bad geometries | - Direct mapper feedback | |
Semantic Accuracy | Landuse-2 | Wrong tagging | Number of features with bad values | - Direct mapper feedback | |||
Agency use - Entry - Registration - Reception - Government office - NGO office - Outreach -Admin |
Data type: node Primary key: office=* Secondary keys: name=<str> |
Semantic Accuracy | Office-1 | Wrong tagging | Number of features with bad values | - Direct mapper feedback | |
Semantic Accuracy | Office-2 | Minimum tag requirements for offices | Number of features with incomplete tags | - Direct mapper feedback |