Evaluating the Ability to Use Contextual Features Derived from Multi-Scale Satellite Imagery to Map Spatial Patterns of Urban Attributes and Population Distributions
With an increasing global population, accurate and timely population counts are essential for urban planning and disaster management. Previous research using contextual features, using mainly very-high-spatial-resolution imagery (<2 m spatial resolution) at subnational to city scales, has found s...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-10-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/13/19/3962 |
| _version_ | 1797515816813461504 |
|---|---|
| author | Steven Chao Ryan Engstrom Michael Mann Adane Bedada |
| author_facet | Steven Chao Ryan Engstrom Michael Mann Adane Bedada |
| author_sort | Steven Chao |
| collection | DOAJ |
| description | With an increasing global population, accurate and timely population counts are essential for urban planning and disaster management. Previous research using contextual features, using mainly very-high-spatial-resolution imagery (<2 m spatial resolution) at subnational to city scales, has found strong correlations with population and poverty. Contextual features can be defined as the statistical quantification of edge patterns, pixel groups, gaps, textures, and the raw spectral signatures calculated over groups of pixels or neighborhoods. While they correlated with population and poverty, which components of the human-modified landscape were captured by the contextual features have not been investigated. Additionally, previous research has focused on more costly, less frequently acquired very-high-spatial-resolution imagery. Therefore, contextual features from both very-high-spatial-resolution imagery and lower-spatial-resolution Sentinel-2 (10 m pixels) imagery in Sri Lanka, Belize, and Accra, Ghana were calculated, and those outputs were correlated with OpenStreetMap building and road metrics. These relationships were compared to determine what components of the human-modified landscape the features capture, and how spatial resolution and location impact the predictive power of these relationships. The results suggest that contextual features can map urban attributes well, with out-of-sample <i>R</i><sup>2</sup> values up to 93%. Moreover, the degradation of spatial resolution did not significantly reduce the results, and for some urban attributes, the results actually improved. Based on these results, the ability of the lower resolution Sentinel-2 data to predict the population density of the smallest census units available was then assessed. The findings indicate that Sentinel-2 contextual features explained up to 84% of the out-of-sample variation for population density. |
| first_indexed | 2024-03-10T06:52:36Z |
| format | Article |
| id | doaj.art-f06d92453c5242e5a2e6dbcc2096633b |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T06:52:36Z |
| publishDate | 2021-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-f06d92453c5242e5a2e6dbcc2096633b2023-11-22T16:43:25ZengMDPI AGRemote Sensing2072-42922021-10-011319396210.3390/rs13193962Evaluating the Ability to Use Contextual Features Derived from Multi-Scale Satellite Imagery to Map Spatial Patterns of Urban Attributes and Population DistributionsSteven Chao0Ryan Engstrom1Michael Mann2Adane Bedada3Department of Geography, The George Washington University, Washington, DC 20052, USADepartment of Geography, The George Washington University, Washington, DC 20052, USADepartment of Geography, The George Washington University, Washington, DC 20052, USADepartment of Geography, The George Washington University, Washington, DC 20052, USAWith an increasing global population, accurate and timely population counts are essential for urban planning and disaster management. Previous research using contextual features, using mainly very-high-spatial-resolution imagery (<2 m spatial resolution) at subnational to city scales, has found strong correlations with population and poverty. Contextual features can be defined as the statistical quantification of edge patterns, pixel groups, gaps, textures, and the raw spectral signatures calculated over groups of pixels or neighborhoods. While they correlated with population and poverty, which components of the human-modified landscape were captured by the contextual features have not been investigated. Additionally, previous research has focused on more costly, less frequently acquired very-high-spatial-resolution imagery. Therefore, contextual features from both very-high-spatial-resolution imagery and lower-spatial-resolution Sentinel-2 (10 m pixels) imagery in Sri Lanka, Belize, and Accra, Ghana were calculated, and those outputs were correlated with OpenStreetMap building and road metrics. These relationships were compared to determine what components of the human-modified landscape the features capture, and how spatial resolution and location impact the predictive power of these relationships. The results suggest that contextual features can map urban attributes well, with out-of-sample <i>R</i><sup>2</sup> values up to 93%. Moreover, the degradation of spatial resolution did not significantly reduce the results, and for some urban attributes, the results actually improved. Based on these results, the ability of the lower resolution Sentinel-2 data to predict the population density of the smallest census units available was then assessed. The findings indicate that Sentinel-2 contextual features explained up to 84% of the out-of-sample variation for population density.https://www.mdpi.com/2072-4292/13/19/3962machine learningcontextual featurespopulationurban attributesmodelingspatial resolution |
| spellingShingle | Steven Chao Ryan Engstrom Michael Mann Adane Bedada Evaluating the Ability to Use Contextual Features Derived from Multi-Scale Satellite Imagery to Map Spatial Patterns of Urban Attributes and Population Distributions Remote Sensing machine learning contextual features population urban attributes modeling spatial resolution |
| title | Evaluating the Ability to Use Contextual Features Derived from Multi-Scale Satellite Imagery to Map Spatial Patterns of Urban Attributes and Population Distributions |
| title_full | Evaluating the Ability to Use Contextual Features Derived from Multi-Scale Satellite Imagery to Map Spatial Patterns of Urban Attributes and Population Distributions |
| title_fullStr | Evaluating the Ability to Use Contextual Features Derived from Multi-Scale Satellite Imagery to Map Spatial Patterns of Urban Attributes and Population Distributions |
| title_full_unstemmed | Evaluating the Ability to Use Contextual Features Derived from Multi-Scale Satellite Imagery to Map Spatial Patterns of Urban Attributes and Population Distributions |
| title_short | Evaluating the Ability to Use Contextual Features Derived from Multi-Scale Satellite Imagery to Map Spatial Patterns of Urban Attributes and Population Distributions |
| title_sort | evaluating the ability to use contextual features derived from multi scale satellite imagery to map spatial patterns of urban attributes and population distributions |
| topic | machine learning contextual features population urban attributes modeling spatial resolution |
| url | https://www.mdpi.com/2072-4292/13/19/3962 |
| work_keys_str_mv | AT stevenchao evaluatingtheabilitytousecontextualfeaturesderivedfrommultiscalesatelliteimagerytomapspatialpatternsofurbanattributesandpopulationdistributions AT ryanengstrom evaluatingtheabilitytousecontextualfeaturesderivedfrommultiscalesatelliteimagerytomapspatialpatternsofurbanattributesandpopulationdistributions AT michaelmann evaluatingtheabilitytousecontextualfeaturesderivedfrommultiscalesatelliteimagerytomapspatialpatternsofurbanattributesandpopulationdistributions AT adanebedada evaluatingtheabilitytousecontextualfeaturesderivedfrommultiscalesatelliteimagerytomapspatialpatternsofurbanattributesandpopulationdistributions |