Search

Glacial meltwater directly influences glacier dynamics. However, in the case of debris-covered glaciers, the drivers of glacier velocity and the influence of supraglacial lakes have not yet been sufficiently analysed and understood. We present a spatio-temporal analysis of key glacier characteristics for Baltoro Glacier in the Karakoram from October 2016 to September 2022 based on Earth observation data and climate parameters extracted from the High Asia Refined analysis (HAR) data set. For the glacier variables, we used surface velocity, supraglacial lake extent, melt of snow and ice, and proglacial run-off index. For climate variables, we focused on air temperature and precipitation. The surface velocity of Baltoro Glacier was characterized by a spring speed-up, summer peak, and fall speed-up with a relative increase in summer of 0.2–0.3 m d -1 (75 %–100 %) in relation to winter velocities, triggered by the onset of or an increase in basal sliding. Snow and ice melt have the largest impact on the spring speed-up, summer velocity peak, and the transition from inefficient to efficient subglacial drainage. The melt covered up to 64 % (353 km 2) of the entirety (debris-covered and debris-free) of Baltoro Glacier and reached up to 4700 m a.s.l. during the first melt peak and up to 5600 m a.s.l. during summer. The temporal delay between the initial peak of seasonal melt and the first relative velocity maximum decreases downglacier. Drainage from supraglacial lakes (3.6–5.9 km 2) contributed to the fall speed-up, which showed a 0.1–0.2 m d -1 (20 %–30 %) lower magnitude compared to the summer velocity peak. Most of the run-off can be attributed to the melt of snow and ice. However, from mid-June onward, the lakes play an increasing role, even though their contribution is estimated to be only about half of that of the melt. The observed increase in summer air temperatures leads to a greater extent of melt, as well as to a rise in the number and total area of supraglacial lakes. This tendency is expected to intensify in a future warming climate. [ABSTRACT FROM AUTHOR]

Landslides, resulting from disturbances in slope equilibrium, pose a significant threat to landscapes, infrastructure, and human life. Triggered by factors such as intense precipitation, seismic activities, or volcanic eruptions, these events can cause extensive damage and endanger nearby communities. A comprehensive understanding of landslide characteristics, including spatio-temporal patterns, dimensions, and morphology, is vital for effective landslide disaster management. Existing remote sensing approaches mostly use either optical or synthetic aperture radar sensors. Integrating information from both these types of sensors promises greater accuracy for identifying and locating landslides. This study proposes a novel approach, the ML-LaDeCORsat (Machine Learning-based coseismic Landslide Detection using Combined Optical and Radar Satellite Imagery), that integrates freely available Sentinel-1, Palsar-2, and Sentinel-2 imagery data in Google Earth Engine (GEE). The approach also integrates relevant spectral indices and suitable bands used in a machine learning-based classification of coseismic landslides. The approach includes a robust and reproducible training and validation strategy and allows one to choose between five classifiers (CART, Random Forest, GTB, SVM, and Naive Bayes). Using landslides from four different earthquake case studies, we demonstrate the superiority of our approach over existing solutions in coseismic landslide identification and localization, providing a GTB-based detection accuracy of 87–92%. ML-LaDeCORsat can be adapted to other landslide events (GEE script is provided). Transfer learning experiments proved that our model can be applied to other coseismic landslide events without the need for additional training data. Our novel approach therefore facilitates quick and reliable identification of coseismic landslides, highlighting its potential to contribute towards more effective disaster management. [ABSTRACT FROM AUTHOR]