In order to promote the economic development of China’s provinces and provide references for the provinces to make effective economic decisions, it is urgent to investigate the trend of province-level economic development. In this study, DMSP/OLS data and NPP/VIIRS data were used to predict economic development. Based on the GDP data of China’s provinces from 1992 to 2016 and the nighttime light remote sensing (NTL) data of corresponding years, we forecast GDP via the linear model (LR model), ARIMA model, ARIMAX model, and SARIMA model. Models were verified against the GDP records from 2017 to 2019. The experimental results showed that the involvement of NTL as exogenous variables led to improved GDP prediction.

Buildings provide indispensable services for human well‐being, but their construction and use are responsible for a substantial fraction of societies’ resource requirements and greenhouse gas emissions. Mapping and quantifying the material stocks in buildings is a key research frontier in industrial ecology. Reliable and spatially highly resolved maps of material stocks in buildings worldwide are so far not available. Existing approaches based on nighttime light data allow large‐scale coverage, but their spatial resolution is usually ∼0.5–1 km. Other methods using light detection and ranging (LiDAR) and cadaster data achieve higher resolution and accuracy, but do not allow wall‐to‐wall mapping of large regions. Based on high‐resolution Earth Observation data combined with material intensity factors (kg per m ³ of building volume), we quantify and map material stocks in buildings at the unprecedented resolution of 90 m globally. We distinguish 18 types of materials in five types of buildings. We find that global material stocks in buildings amount to 547 (391–672) Gt, approximately half of total global societal material stocks. We find highly unequal distributions of material stocks in buildings per capita and per unit area of each country. Our results agree well with previous detailed estimates of material stocks in buildings in dedicated regions or individual cities. Improved and harmonized material intensity factors emerge as a key research area for improving the accuracy of material stock maps. Our results are available as data products with high spatial and thematic resolution to facilitate future studies; for example, of secondary resource potentials. This article met the requirements for a gold‐gold JIE data openness badge described at http://jie.click/badges .

The performance of deep learning is heavily influenced by the size of the learning samples, whose labeling process is time consuming and laborious. Deep learning algorithms typically assume that the training and prediction data are independent and uniformly distributed, which is rarely the case given the attributes and properties of different data sources. In remote sensing images, representations of urban land surfaces can vary across regions and by season, demanding rapid generalization of these surfaces in remote sensing data. In this study, we propose Meta-FSEO, a novel model for improving the performance of few-shot remote sensing scene classification in varying urban scenes. The proposed Meta-FSEO model deploys self-supervised embedding optimization for adaptive generalization in new tasks such as classifying features in new urban regions that have never been encountered during the training phase, thus balancing the requirements for feature classification tasks between multiple images collected at different times and places. We also created a loss function by weighting the contrast losses and cross-entropy losses. The proposed Meta-FSEO demonstrates a great generalization capability in remote sensing scene classification among different cities. In a five-way one-shot classification experiment with the Sentinel-1/2 Multi-Spectral (SEN12MS) dataset, the accuracy reached 63.08%. In a five-way five-shot experiment on the same dataset, the accuracy reached 74.29%. These results indicated that the proposed Meta-FSEO model outperformed both the transfer learning-based algorithm and two popular meta-learning-based methods, i.e., MAML and Meta-SGD.

Urban Agglomeration (UA) is regarded as an emerging complex urban system in China. The development of UA demands a reasonable scale structure, which can be investigated by Zipf’s law. However, few studies have been conducted to quantify the optimal scale of UA and how its development deviates from the optimal scale. With the continuous urban expansion, the problem of UAs’ scale structure has received increasing attention. In this study, we propose a method based on Zipf’s law for estimating the theoretical optimal scale of UAs in China and assessing the deviation rate from their optimal scales. Twelve typical UAs in China are selected, and their development is assessed via urban impervious surface data from 2000 to 2018. The results show that the average deviation rate of the investigated UAs decreased from 3.40% in 2000 to 2.32% in 2018, demonstrating that these UAs are on a positive evolution trajectory. Furthermore, according to the development stage, we make recommendations on “large cities vs. medium/small-sized cities and promoting vs. restraining” to each UA based on its size. The conceptual and analytical knowledge, as well as the results from this study, are expected to offer valuable insights and new references for regulating and managing UAs’ development in China.

Mapping impervious surface area (ISA) dynamics at the regional and global scales is an important task that supports the management of the urban environment and urban ecological systems. In this study, we aimed to develop a new method for ISA percentage (ISA%) mapping using Nighttime Light (NTL) and MODIS products. The proposed method consists of three major steps. First, we calculated the Enhanced Vegetation Index (EVI)-adjusted NTL index (EANTLI) and performed intra-annual and inter-annual corrections on the DMSP-OLS data. Second, based on the geographically weighted regression (GWR) model, we built a consistent NTL product from 2000 to 2019 by performing an intercalibration between DMSP-OLS and VIIRS images. Third, we adopted a GA-BP neural network model to monitor ISA% dynamics using NTL imagery, MODIS imagery, and population data. Taking the Guangdong–Hong Kong–Macao Greater Bay as the study area, our results indicate that the ISA% in our study area increased from 7.97% in 2000 to 17.11% in 2019, with a mean absolute error (MAE) of 0.0647, root mean square error (RMSE) of 0.1003, Pearson’s coefficient of 0.9613, and R² (R-squared) of 0.9239. Specifically, these results demonstrate the effectiveness of the proposed method in mapping ISA and investigating ISA dynamics using temporal features extracted from consistent NTL and MODIS products. The proposed method is feasible when generating ISA% at a large scale at high frequency, given the ease of implementation and the availability of input data sources.