Generating multi-temporal landslide inventories through a general deep transfer learning strategy using HR EO data.

Mapping of landslides over space has seen an increasing attention and good results in the last decade. While current methods are chiefly applied to generate event-inventories, whereas multi-temporal (MT) inventories are rare, even using manual landslide mapping. Here, we present an innovative deep learning strategy which employs transfer learning that allows for the Attention Deep Supervision Multi-Scale U-Net model to be adapted for landslide detection tasks in new areas. The method also provides the flexibility of re-training a pretrained model to detect both rainfall- and earthquake-triggered landslides on new target areas. For the mapping, we used archived Planet Lab remote sensing images spanning a period between 2009 till 2021 with spatial resolution of 3-5 m to systematically generate MT landslide inventories. When we examined all cases, our approach provided an average F1 score of 0.8 indicating that we successfully identified the spatiotemporal occurrences of landslides. To examine the size distribution of mapped landslides we compared the frequency-area distributions of predicted co-seismic landslides with manually mapped products from the literature. Results showed a good match between calculated power-law exponents where the difference ranges between 0.04 and 0.21. Overall, this study showed that the proposed algorithm could be applied to large areas to generate polygon-based MT landslide inventories.

Chamoli disaster: pronounced changes in water quality and flood plains using Sentinel data.

The Himalayan rivers are vulnerable to devastating flooding caused by landslides and outbreak of glacial lakes. On 7 February 2021, a deadly disaster occurred near the Rishi Ganga Hydropower Plant in the Rishi Ganga River, killing more than 100 people. During the event, a large volume of debris and broken glacial fragments flooded the Rishi Ganga River and washed away the Rishi Ganga Hydropower plant ongoing project. This study presents the impact of the Chamoli disaster on the water quality of Rishi Ganga River in upstream near Tapovan and Ganga River in downstream near Haridwar through remote sensing data. Five points have been used at different locations across the two study areas and three different indices were used such as Normalized difference water index (NDWI), Normalized difference turbidity Index (NDTI), and Normalized difference chlorophyll index (NDCI), to analyze changes in water quality. Spectral signatures and backscattering coefficients derived from Sentinel-2 Optical and Sentinel-1 Synthetic-aperture radar (SAR) data were also compared to study the changes in water quality. It was evident from the water quality indices and spectral signatures that the flood plains changed significantly. Using spectral signatures and different indices, the water level in the Chilla dam canal near Haridwar was found to decreased after the Chamoli disaster event as the flood gates were closed to stop the deposit of sediments in the canal. Results suggest changes in water quality parameters (turbidity, chlorophyll concentration, NDWI) at the five locations near the deadly site and far away at Haridwar along the Ganga River. This study is a preliminary qualitative analysis showing changes in river flood plain and water quality after the Chamoli disaster.