3 December 2021 in Earth System Science Data
The absence of a compiled large-scale catchment characteristics dataset is a key obstacle limiting the development of large-sample hydrology research in China. We introduce the first large-scale catchment attribute dataset in China. We compiled diverse data sources, including soil, land cover, climate, topography, and geology, to develop the dataset. The dataset also includes catchment-scale 31-year meteorological time series from 1990 to 2020 for each basin. Potential evapotranspiration time series based on Penman's equation are derived for each basin. The 4911 catchments included in the dataset cover all of China. We introduced several new indicators that describe the catchment geography and the underlying surface differently from previously proposed datasets. The resulting dataset has a total of 125 catchment attributes and includes a separate HydroMLYR (hydrology dataset for machine learning in the Yellow River Basin) dataset containing standardized weekly averaged streamflow for 102 basins in the Yellow River Basin. The standardized streamflow data should be able to support machine learning hydrology research in the Yellow River Basin. The dataset is freely available at https://doi.org/10.5281/zenodo.5729444 (Zhen et al., 2021). In addition, the accompanying code used to generate the dataset is freely available at https://github.com/haozhen315/CCAM-China-Catchment-Attributes-and-Meteorology-dataset (last access: 26 November 2021) and supports the generation of catchment characteristics for any custom basin boundaries. Compiled data for the 4911 basins covering all of China and the open-source code should be able to support the study of any selected basins rather than being limited to only a few basins.
13 November 2023 in IEEE Transactions on Geoscience and Remote Sensing
Super-resolution mapping (SRM) is an effective way to alleviate the mixed pixel problem of remotely sensed imagery, by transforming the coarse-resolution fraction image originated from spectral unmixing into a fine-resolution land-cover map. Deep learning has been widely used in SRM since it has a powerful ability to represent the complex heterogeneous spatial distribution patterns of land-cover patches. However, the accuracy of existing deep learning-based SRM models is compromised by the fact that the fraction images used in SRM always contain errors. In this paper, we propose an end-to-end convolutional neural network (CNN)-based fraction error eliminating SRM (DeepNESRM) method to overcome the negative effect of fraction errors in SRM. In DeepNESRM, to better learn the complex nonlinear relationship between the actual coarse-resolution fraction image and the fine-resolution land-cover map by the CNN, a practical error simulation method that considers the characteristics of fraction errors is introduced to produce training samples. In addition, a multi-level feature fusion CNN is adopted to eliminate fraction errors and simultaneously implement SRM. Experiments using Sentinel-2 and Landsat 8 images were conducted to test the performance of the proposed method. Two conventional SRM methods, namely the pixel swapping method and the spatial dependence and L2 norm combined SRM (L2_SRM) method, and also a stacked very deep CNN based SRM (VDSRM) method, were used as the comparison methods. The results show that DeepNESRM can deal with fraction errors and preserve the spatial detail information, and achieve higher average overall accuracies than the other methods.
13 September 2024 in Water Resources Research
Accurate monitoring of river discharge variations is essential for managing floods and droughts and understanding the response of global river systems to climate change. Remote sensing of discharge (RSQ) offers a timely and efficient alternative for widespread monitoring, particularly in ungauged areas. Current methods often struggle with accuracy, especially when estimating the width of narrow rivers from medium‐resolution images. We first observe that, although estimating the width variation of narrow rivers can be challenging from medium‐resolution satellite imagery, river discharge still correlates with river surface color or reflectance. However, existing methods can only correlate river surface reflectance with discharge in gauged rivers. Here, we introduce a novel method employing an advanced Transformer architecture to map river discharge variations directly from time‐series reflectance imagery. Our model, trained on quality‐checked data from 2,036 discharge gauges, outperforms existing methods in discharge estimation accuracy and is less affected by the need for precise river width estimation. The proposed model yields positive Kling‐Gupta Efficiency (KGE) in 68.6% of ungauged rivers, a substantial improvement over the BAM and geoBAM methods, which show positive KGEs in only 28.4% and 33.1% of rivers, respectively. Notably, this performance is achieved despite two‐thirds of the rivers being less than 100 m wide, a range where traditional RSQ methods typically struggle, and the RSQ performance does not show degradation for braided rivers. Our approach suggests a significant shift toward more efficient, extensive, and adaptable space‐based river discharge monitoring.
15 February 2024 in Environmental Research Letters
Solar energy, as an environmentally sustainable power source, is gaining increasing popularity worldwide, driving a surge in the number of solar photovoltaic (PV) plants. China, which has a prominent role in this domain, requires continuous updates to its PV plant data for spatiotemporal analyses. However, there remains an absence of a comprehensive and timely dataset of PV plants across China, leaving PV installation dates and other crucial attributes for comprehensive analyses underexplored. This study leverages Sentinel-2 data as a primary source to propose an optimized deep learning approach for PV plant extraction in China. Statistical analyses of PV plant attributes, including its installation date, size, site slope, and site land cover, were implemented from multiple data sources. Comprehensive analyses were conducted to unravel their spatiotemporal development patterns in the 21st century. The results indicate that as of 2023, China boasts 4347 PV plants, collectively spanning 4146 km2, which are predominantly concentrated in Northwest and North China. 2016 and 2017 marked substantial growth in China’s PV plants, while other years exhibited stability. These plants exhibit the distinct spatial characteristics of installing smaller PV plants on flat terrain covered by vegetation or barren land. Over time, a notable trend in the installation of China’s PV plants has been the increasing preference for establishing larger ones in smooth terrain, with a focus on preventing damage to natural resources. The results reveal China’s optimization of PV plant site selection and construction strategies, aligning with global environmental goals and sustainable energy practices.
3 March 2025 in IEEE Transactions on Geoscience and Remote Sensing
Medium spatial resolution remote sensing images are widely used for cropland mapping. In areas where cropland is fragmented, however, the limitation of the spatial resolution may lead to inaccurate or even impossible mapping of small croplands. Super-resolution mapping is an effective method to address this issue by transforming coarse-resolution fraction images, derived from spectral unmixing, into fine-resolution land cover maps. In practical applications, a crucial obstacle of this approach is the difficulty in collecting training samples for spectral unmixing and super-resolution mapping. To address this problem, this paper proposed a novel super-resolution cropland mapping approach by simulating spectral and spatial training samples. Specially, a mixture spectral simulation method was used to generate training samples for the regression unmixing model to estimate cropland fraction images. A multi-level feature fusion U-NET model was proposed for super-resolution cropland mapping and was trained with simulated training samples considering fraction errors. The proposed method was tested in the Jianghan Plain, China, by generating 2.5m cropland maps from the 10m Sentinel-2 images. The results show that the proposed method can accurately extract more small and linear land cover features, preserve the spatial structure of the boundaries and achieve higher accuracy than other cropland mapping methods. This method overcomes the dependency on actual samples collection in traditional methods, better utilizes spectral and spatial features in remote sensing data, and reduces the impact of spectral unmixing errors on the final fine-resolution cropland maps.