Slum identification in urban settlements is a crucial step in the process of formulation of pro-poor policies. However, the use of conventional methods for slum detection such as field surveys can be time-consuming and costly. This paper explores the possibility of implementing a low-cost standardized method for slum detection. We use spectral, texture and structural features extracted from very high spatial resolution imagery as input data and evaluate the capability of three machine learning algorithms (Logistic Regression, Support Vector Machine and Random Forest) to classify urban areas as slum or no-slum. Using data from Buenos Aires (Argentina), Medellin (Colombia) and Recife (Brazil), we found that Support Vector Machine with radial basis kernel delivers the best performance (with F2-scores over 0.81). We also found that singularities within cities preclude the use of a unified classification model.

This paper presents a general framework for modeling the growth of three important variables for cities: population distribution, binary urban footprint, and urban footprint in color. The framework models the population distribution as a spatiotemporal regression problem using machine learning, and it obtains the binary urban footprint from the population distribution through a binary classifier plus a temporal correction for existing urban regions. The framework estimates the urban footprint in color from its previous value, as well as from past and current values of the binary urban footprint using a semantic inpainting algorithm. By combining this framework with free data from the Landsat archive and the Global Human Settlement Layer framework, interested users can get approximate growth predictions of any city in the world. These predictions can be improved with the inclusion in the framework of additional spatially distributed input variables over time subject to availability. Unlike widely used growth models based on cellular automata, there are two main advantages of using the proposed machine learning-based framework. Firstly, it does not require to define rules a priori because the model learns the dynamics of growth directly from the historical data. Secondly, it is very easy to train new machine learning models using different explanatory input variables to assess their impact. As a proof of concept, we tested the framework in Valledupar and Rionegro, two Latin American cities located in Colombia with different geomorphological characteristics, and found that the model predictions were in close agreement with the ground-truth based on performance metrics, such as the root-mean-square error, zero-mean normalized cross-correlation, Pearson’s correlation coefficient for continuous variables, and a few others for discrete variables such as the intersection over union, accuracy, and the $f_1$ metric. In summary, our framework for modeling urban growth is flexible, allows sensitivity analyses, and can help policymakers worldwide to assess different what-if scenarios during the planning cycle of sustainable and resilient cities.

A large proportion of Africa’s infrastructure is yet to be built. Where and how these new buildings are constructed matters since today’s decisions will last for decades. The resulting morphology of cities has lasting implications for a city’s energy needs. Estimating and projecting these needs has always been challenging in Africa due to the lack of data. Yet, given the sweeping urbanization expected in Africa over the next three decades, this obstacle must be overcome to guide cities toward a trajectory of sustainability and resilience. Based on the location and surface of nearly 200 million buildings on the continent, we estimate the interbuilding distance of almost six thousand cities. Buildings’ footprint data enable the construction of urban form indicators to compare African cities’ elongation, sprawl, and emptiness. We establish the BASE model, where the mean distance between buildings is a functional relation to the number of Buildings and their average Area, as well as the Sprawl and the Elongation of its spatial arrangement. The mean distance between structures in cities—our proxy for its energy demands related to mobility—grows faster than the square root of its population, resulting from the combined impact of a sublinear growth in the number of buildings and a sublinear increase in building size and sprawl. We estimate that when a city doubles its population, it triples its energy demand from transport.

High-resolution population mapping is of high relevance for developing and implementing tailored actions in several fields: From decision making in crisis management to urban planning. Earth Observation has considerably contributed to the development of methods for disaggregating population figures with higher resolution data into fine-grained population maps. However, which method is most suitable on the basis of the available data, and how the spatial units and accuracy metrics affect the validation process is not fully known. We aim to provide recommendations to researches that attempt to produce high-resolution population maps using remote sensing and geospatial information in heterogeneous urban landscapes. For this purpose, we performed a comprehensive experimental research on population disaggregation methods with thirty-six different scenarios. We combined five different top-down methods (from basic to complex, i.e., binary and categorical dasymetric, statistical, and binary and categorical hybrid approaches) on different subsets of data with diverse resolutions and degrees of availability (poor, average and rich). Then, the resulting population maps were systematically validated with a two-fold approach using six accuracy metrics. We found that when only using remotely sensed data the combination of statistical and dasymetric methods provide better results, while highly-resolved data require simpler methods. Besides, the use of at least three relative accuracy metrics is highly encouraged since the validation depends on level and method. We also analysed the behaviour of relative errors and how they are affected by the heterogeneity of the urban landscape. We hope that our recommendations save additional efforts and time in future population mapping.