Cloud detection is an important preprocessing step for the precise application of optical satellite imagery. In this paper, we propose a deep learning based cloud detection method named multi-scale convolutional feature fusion (MSCFF) for remote sensing images of different sensors. In the network architecture of MSCFF, the symmetric encoder-decoder module, which provides both local and global context by densifying feature maps with trainable convolutional filter banks, is utilized to extract multi-scale and high-level spatial features. The feature maps of multiple scales are then up-sampled and concatenated, and a novel multi-scale feature fusion module is designed to fuse the features of different scales for the output. The two output feature maps of the network are cloud and cloud shadow maps, which are in turn fed to binary classifiers outside the model to obtain the final cloud and cloud shadow mask. The MSCFF method was validated on hundreds of globally distributed optical satellite images, with spatial resolutions ranging from 0.5 to 50 m, including Landsat-5/7/8, Gaofen-1/2/4, Sentinel-2, Ziyuan-3, CBERS-04, Huanjing-1, and collected high-resolution images exported from Google Earth. The experimental results show that MSCFF achieves a higher accuracy than the 2 traditional rule-based cloud detection methods and the state-of-the-art deep learning models, especially in bright surface covered areas. The effectiveness of MSCFF means that it has great promise for the practical application of cloud detection for multiple types of medium and highresolution remote sensing images. Our established global high-resolution cloud detection validation dataset has been made available online
This paper presents a novel method to estimate active layer thickness (ALT) over permafrost based on InSAR (Interferometric Synthetic Aperture Radar) observation and the heat transfer model of soils. The time lags between the periodic feature of InSAR-observed surface deformation over permafrost and the meteorologically recorded temperatures are assumed to be the time intervals that the temperature maximum to diffuse from the ground surface downward to the bottom of the active layer. By exploiting the time lags and the one-dimensional heat transfer model of soils, we estimate the ALTs. Using the frozen soil region in southern Qinghai-Tibet Plateau (QTP) as examples, we provided a conceptual demonstration of the estimation of the InSAR pixel-wise ALTs. In the case study, the ALTs are ranging from 1.02 to 3.14 m and with an average of 1.95 m. The results are compatible with those sparse ALT observations/estimations by traditional methods, while with extraordinary high spatial resolution at pixel level (~40 meter). The presented method is simple, and can potentially be used for deriving high-resolution ALTs in other remote areas similar to QTP, where only sparse observations are available now.
ABSTRACT. An anomalously slight glacier mass gain during 2000 to the 2010s has recently been reported in the Karakoram region. However, to date, no investigations of the region-wide glacier mass balance in the Karakoram prior to 2000 have been reported, leaving a knowledge gap for assessing glacier responses to climate change. We calculated elevation and mass change using DEMs generated from KH-9 images acquired during 1973-1980 and the 1 arc-second SRTM DEM. We find a slight mass loss of −0.09 ± 0.03 m w.e. a −1 (12 366 km 2 ) for 1973-2000, which is less negative than the global average rate for 1971-2009 (−0.31 ± 0.19 m w.e. a −1 ). Within the Karakoram, the glacier change patterns are spatially and temporally heterogeneous. In particular, a nearly stable state in the central Karakoram (−0.04 ± 0.05 m w.e. a −1 during the period implies that the Karakoram anomaly dates back to the 1970s. Combined with the previous studies, we conclude that the Karakoram glaciers as a whole were in a nearly balanced state during the 1970s to the 2010s.
The long‐term effects of increased temperatures on sediment fluxes in cold regions remain poorly investigated. Here, we examined the multidecadal changes in runoff and sediment fluxes in the Tuotuohe River, a headwater river of the Yangtze River on the Tibetan Plateau (TP). The sediment fluxes and runoff increased at rates of 0.03 ± 0.01 Mt/yr (5.9 ± 1.9%/yr) and 0.025 ± 0.007 × km3/yr (3.5 ± 1.0%/yr) from 1985 to 2016, with net increases of 135% and 78% from 1985–1997 to 1998–2016, respectively. The increases are primarily due to warming temperature (+1.44°C) and intensified glacier‐snow‐permafrost melting, with enhanced precipitation (+30%) as a secondary cause. Sediment fluxes are much more susceptible to climate warming than runoff in this undisturbed cold environment. The substantially increased sediment fluxes from the headwater region could threaten the numerous constructed reservoirs and influence the aquatic ecosystems of the TP and its marginal areas.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.