Recently, the capability of deep learning-based approaches, especially deep convolutional neural networks (CNNs), has been investigated for hyperspectral remote sensing feature extraction (FE) and classification. Due to the large number of learnable parameters in convolutional filters, lots of training samples are needed in deep CNNs to avoid the overfitting problem. On the other hand, Gabor filtering can effectively extract spatial information including edges and textures, which may reduce the FE burden of the CNNs. In this letter, in order to make the most of deep CNN and Gabor filtering, a new strategy, which combines Gabor filters with convolutional filters, is proposed for hyperspectral image classification to mitigate the problem of overfitting. The obtained results reveal that the proposed model provides competitive results in terms of classification accuracy, especially when only a limited number of training samples are available.

Physical erosion and chemical weathering rates beneath glaciers are expected to increase in a warming climate with enhanced melting but are poorly constrained. We present a global dataset of cations in meltwaters of 77 glaciers, including new data from 19 Asian glaciers. Our study shows that contemporary cation denudation rates (CDRs) beneath glaciers (2174 ± 977 Σ*meq⁺ m⁻² year⁻¹) are ~3 times higher than two decades ago, up to 10 times higher than ice sheet catchments (~150-2000 Σ*meq⁺ m⁻² year⁻¹), up to 50 times higher than whole ice sheet means (~30-45 Σ*meq⁺ m⁻² year⁻¹) and ~4 times higher than major non-glacial riverine means (~500 Σ*meq⁺ m⁻² year⁻¹). Glacial CDRs are positively correlated with air temperature, suggesting glacial chemical weathering yields are likely to increase in future. Our findings highlight that chemical weathering beneath glaciers is more intense than many other terrestrial systems and may become increasingly important for regional biogeochemical cycles.