There is a growing recognition of the worsening problems of accessibility in Africa's major urban centres and city-regions. Yet, research assessing differential accessibility in these contexts is limited, partly because of the lack of population and land-use data at granular spatial resolutions. In this paper, we contribute to addressing the existing challenges by presenting a geospatial modelling approach that combines historical urban physical development data extracted from Landsat satellite images, travel time data from Openrouteservice and Google Maps API services, and relevant transport infrastructure data. We apply the approach to analyse place-based (100 m²) accessibility in the Greater Kumasi city-region of Ghana. The analysis reveals a strong co-evolutionary relationship between physical development and geographical accessibility, with an estimated 83% of the emergent physical development located within half-a-kilometre distance of a primary road. First-and-last-mile accessibility in terms of walking time to public transport stations and public transport service routes are quantified. We reveal huge first-and-last-mile accessibility deficits, with an estimated 14% and 33% of the city-region's built-up area having a 5-minute and 10-minute walking time reachability to a public transport station, as a critical node of accessibility. Differential accessibility by car and public transport is analysed, focusing on the city-region's most important functional centre (i.e. CBD). The results show that the CBD is within 15 min and nearly 30 min reach by car and public transport respectively, for only 1% of the city-region's built-up area. This implies that overall accessibility is poor regardless of mode choice, due to congestion. The findings underscore the need for integrative urban development and accessibility planning in the city-region for equitable transport and mobility outcomes.

人们越来越认识到非洲主要城市中心和城市区域的无障碍问题日益严重。然而，在这些情况下评估差异可达性的研究是有限的，部分原因是缺乏细粒度空间分辨率的人口和土地利用数据。在本文中，我们通过提出一种地理空间建​​模方法来应对现有挑战，该方法结合了从 Landsat 卫星图像提取的历史城市物理发展数据、来自 Openrouteservice 和 Google Maps API 服务的旅行时间数据以及相关的交通基础设施数据。我们应用该方法分析基于地点 (100 m ²) 加纳大库马西城市地区的可达性。分析表明，自然发展与地理可达性之间存在很强的共同进化关系，估计有 83% 的新兴自然发展位于距离主要道路半公里范围内。就步行到公共交通站点和公共交通服务路线的时间而言，第一英里和最后一英里的可达性被量化。我们发现第一英里和最后一英里的可达性存在巨大的缺陷，估计该城市区域的建成区中有 14% 和 33% 的步行时间可达公共交通站，作为可访问性的关键节点。分析了汽车和公共交通的不同可达性，重点是城市区域最重要的功能中心（即 CBD）。结果表明，CBD距离分别在15分钟和近30分钟车程内和公共交通，仅占城市区域建成区的1%。这意味着由于拥堵，无论模式选择如何，整体可达性都很差。调查结果强调需要在城市区域进行综合城市发展和可达性规划，以实现公平的交通和流动性成果。