Spatial interpolation is a traditional geostatistical operation that aims at predicting the attribute values of unobserved locations given a sample of data defined on point supports. However, the continuity and heterogeneity underlying spatial data are too complex to be approximated by classic statistical models. Deep learning models, especially the idea of conditional generative adversarial networks (CGANs), provide us with a perspective for formalizing spatial interpolation as a conditional generative task. In this article, we design a novel deep learning architecture named conditional encoder-decoder generative adversarial neural networks (CEDGANs) for spatial interpolation, therein combining the encoder-decoder structure with adversarial learning to capture deep representations of sampled spatial data and their interactions with local structural patterns. A case study on elevations in China demonstrates the ability of our model to achieve outstanding interpolation results compared to benchmark methods. Further experiments uncover the learned spatial knowledge in the model’s hidden layers and test the potential to generalize our adversarial interpolation idea across domains. This work is an endeavor to investigate deep spatial knowledge using artificial intelligence. The proposed model can benefit practical scenarios and enlighten future research in various geographical applications related to spatial prediction.

Inferring the unknown properties of a place relies on both its observed attributes and the characteristics of the places to which it is connected. Because place characteristics are unstructured and the metrics for place connections can be diverse, it is challenging to incorporate them in a spatial prediction task where the results could be affected by how the neighborhoods are delineated and where the true relevance among places is hard to identify. To bridge the gap, we introduce graph convolutional neural networks (GCNNs) to model places as a graph, where each place is formalized as a node, place characteristics are encoded as node features, and place connections are represented as the edges. GCNNs capture the knowledge of the relevant geographic context by optimizing the weights among graph neural network layers. A case study was designed in the Beijing metropolitan area to predict the unobserved place characteristics based on the observed properties and specific place connections. A series of comparative experiments was conducted to highlight the influence of different place connection measures on the prediction accuracy and to evaluate the predictability across different characteristic dimensions. This research enlightens the promising future of GCNNs in formalizing places for geographic knowledge representation and reasoning.

Mixed use has been extensively applied as an urban planning principle and hinders the study of single urban functions. To address this problem, it is worth decomposing the mixed use. Inspired by the concept of spectral unmixing in remote sensing applications, this paper proposes a framework for mixed-use decomposition based on big geo-data. Mixed-use decomposition in terms of human activities differs from traditional land use research, and it is more reasonable to infer the actual urban function of land. The framework consists of four steps, namely temporal activity signature extraction, urban function base curve extraction, mixed-use decomposition, and result validation. First, the temporal activity signatures (TASs) of each zone are extracted as the proxy of human activity patterns. Second, the diurnal TASs of routine activities are extracted as urban function base curves (i.e. endmembers). Third, a linear decomposition model is used to decompose the mixed use and obtain multiple results (urban function composition, dynamic activity proportions, and the mixing index). Finally, result validation strategies are concluded. This framework offers method extensibility and has few requirements for the input data. It is validated by means of a case study of Beijing, based on a social media check-in dataset.

The spatial concentration of human activity is a crucial indication of socio-economic vitality. Accurately mapping activity volumes is fundamental to support regional sustainable development. Current approaches rely on mobile positioning data, which record information about human daily activity but are inaccessible in most cities due to privacy and data sharing concerns. Alternative methods are needed to provide more generalized predictions on extensive areas while maintaining low cost. This study demonstrates how remote sensing imagery can be used through an end-to-end deep learning framework for reliable estimates of human activity volumes. The neighbour effect, representing the inherent nature of spatial autocorrelation in the volumes, is incorporated to improve the network. The proposed model exhibits strong predictive power and demonstrates great explainability of physical environment on variations of activity volumes. Landscape interpretations based on hierarchical features provide both object-based and region-based insights into the co-evolvement of landscape and human activity. Our findings indicate the possibility of extensively predicting activity volumes, especially in areas with limited access to mobile data, and provide support for the promising framework to better comprehend broad aspects of human society from observable physical environments.

Big geo-data are often aggregated according to spatio-temporal units for analyzing human activities and urban environments. Many applications categorize such data into groups and compare the characteristics across groups. The intergroup differences vary with spatio-temporal units, and the essential is to identify the spatio-temporal units with apparently different data characteristics. However, spatio-temporal dependence, data variety, and the complexity of tasks impede an effective unit assessment. Inspired by the applications to extract critical image components based on explainable artificial intelligence (XAI), we propose a spatio-temporal layer-wise relevance propagation method to assess spatio-temporal units as a general solution. The method organizes input data into an extensible three-dimensional tensor form. We provide two means of labeling the spatio-temporal tensor data for typical geographical applications, using temporally or spatially relevant information. Neural network training proceeds to extract the global and local characteristics of data for corresponding analytical tasks. Then the method propagates classification results backward into units as obtained task-specific importance. A case study with taxi trajectory data in Beijing validates the method. The results prove that the proposed method can evaluate the task-specific importance of spatio-temporal units with dependence. This study also attempts to discover task-related knowledge using XAI.