This Global Dated Landslide Database (GDLDB) is part of the project WeMonitor (Weakly Supervised Deep Learning Models for Detecting and Monitoring Spatio-Temporal Anomalies in Optical and Radar Satellite Time Series), funded by the Helmholtz Imaging Platform. The aim is to develop a deep learning model that uses satellite image time series from Sentinel1/2 to automatically monitor changes caused, for example, by landslides, deforestation, large fires, dam failures, or the emergence of waste dumps. To train such a model, a reference dataset is required that shows the area and date of the changes as precise as possible. To allow for a generic and transferable model, the reference data also needs to cover the diversity of the process to be detected. Thus, the aim of the GDLDB is to comprise landslides of different sizes, shapes, and types, occurring at different seasons and in different regions with varying natural conditions and different triggering mechanisms such as rainfall and earthquake-induced landslides.
To build the GDLDB, available local and regional landslide inventories from around the world are combined into one coherent database by verifying their location and date of occurrence with high-resolution remote sensing data. The selection criteria for the source inventories are the definition of the landslide location as polygons, at least a rough indication of the landslide origin date, and that the landslides occurred during the Sentinel-2 data availability from 2016 onwards. A total of 16 individual inventories are included (Table 1), one each from the USA, Dominica, Italy, Zimbabwe, southern India, Nepal, China, Papua New Guinea, and New Zealand, and two each from Kyrgyzstan, Japan, and the Philippines. In addition, a global inventory was added, including a small number of landslides from the USA, Peru, Chile, Europe, Pakistan, Nepal, India, and Taiwan, and a larger number of landslides from Indonesia.
From each inventory, approximately 100 landslides were randomly selected to ensure an unbiased selection of landslides in terms of shape, size, and location. The original source inventories are produced using a variety of methods, including manual mapping in airborne data with ground verification and automatic identification in satellite remote sensing data. As a result, the mapping quality of the inventories varies greatly. In cases where landslides could not be verified by us using available optical remote sensing data (e.g. Sentinel-2, Planet Scope, and data available in Google Earth) new polygons are selected until the number of approximately 100 landslides is reached. In some inventories, the number of 100 landslides could not be guaranteed, due to a lack of suitable landslides (e.g., small size, incorrect classification) or the total number of landslides in the selected inventory was less than 100. For inventories with a lot of small landslides, that were difficult or impossible to observe, a size threshold of 1000m2 was introduced.
This dataset is supplementary to the article of Scherler et al. (submitted), in which the global distribution of supraglacial debris cover is mapped and analyzed. For mapping supraglacial debris cover, we combined glacier outlines from the Randolph Glacier Inventory (RGI) version 6.0 (RGI consortium, 2017) with remote sensing-based ice and snow identification. Areas that belong to glaciers but that are neither ice nor snow were classified as debris cover. This dataset contains the outlines of the mapped debris-covered glaciers areas, stored in shapefiles (.shp).For creating this dataset, we used optical satellite data from Landsat 8 (for the time period 2013-2017), and from Sentinel-2A/B (2015-2017). For the ice and snow identification, we used three different algorithms: a red to short-wavelength infrared (swir) band ratio (RATIO; Hall et al., 1988), the normalized difference snow index (NDSI; Dozier, 1989), and linear spectral unmixing-derived fractional debris cover (FDC; e.g., Keshava and Mustard, 2002). For a detailed description of the debris-cover mapping and an analysis of the data, please see Scherler et al. (2019) to which these data are supplementary material.This dataset includes debris cover outlines based on either Landsat 8 (LS8; 30-m resolution) or Sentinel 2 (S2; 10-m resolution), and the three algorithms RATIO, NDSI, FDC. In total, there exist six different zip-files that each contain 19 shapefiles. The structure of the shapefiles follows that of the RGI version 6.0 (RGI consortium, 2017), with one shapefile for each RGI region. The original RGI shapefiles provide each glacier as one entry (feature) and include a variety of ancillary information, such as area, slope, aspect (RGI Consortium 2017a, Technical Note p. 12ff). Because the debris-cover outlines are based on the RGI v6.0 glacier outlines, all fields of the original shapefiles, which refer to the glacier, are retained, and expanded with four new fields:- DC_Area: Debris-covered area in m². Note that this unit for area is different from the unit used for reporting the glacier area (km²).- DC_BgnDate: Start of the time period from which satellite imagery was used to map debris cover.- DC_EndDate: End of the time period from which satellite imagery was used to map debris cover.- DC_CTSmean: Mean number of observations (CTS = COUNTS) per pixel and glacier. This number is derived from the number of available satellite images for the respective time period, reduced by filtering pixels due to cloud and snow cover.The dataset has a global extent and covers all of the glaciers in the RGI v. 6.0, but it exhibits poor coverage in the RGI region Subantarctic and Antarctic, where the debris cover extents are based on very few observations.
Against the backdrop of global change, both in terms of climate and demography, there is an increasing need for monitoring global water cycle. The publicly available global database is very limited in its spatial and temporal coverage worldwide. Moreover, the acquisition of in situ data and their delivery to the database are on the decline since the late 1970s be it for economical, political or other reasons.
Given the insufficient monitoring from in situ gauge networks, and without any outlook of improvement, spaceborne approaches are currently being investigated. Satellite-based Earth observation with its global coverage and homogeneous accuracy has been demonstrated to be a potential alternative to in situ measurements. The Institute of Geodesy (GIS), within the Faculty of Aerospace Engineering and Geodesy at University of Stuttgart has a long-standing expertise, both theoretically and practically, in dynamic satellite geodesy. In recent years, GIS initiated and participated in studies and projects on application of spaceborne geodetic sensors for hydrological studies. HydroSat provides the results of these studies and projects, in which spaceborne geodetic sensors are used to estimate Surface water extent from satellite imagery Water level from satellite altimetry Water Storage Anomaly from satellite gravimetry River discharge from satellite altimetry, imagery or gravimetry
Orbital products describe positions and velocities of satellites, be it the Global Navigation Satellite System (GNSS) satellites or Low Earth Orbiter (LEO) satellites. These orbital products can be divided into the fastest available ones, the Near Realtime Orbits (NRT, Zitat), which are mostly available within 15 to 60 minutes delay, followed by Rapid Science Orbit (RSO, Zitat) products with a latency of two days and finally the Precise Science Orbit (PSO) which, with a latency of up to a few weeks or longer in the case of reprocessing campaigns, are the most delayed. The absolute positional accuracy increases from NRT to PSO.
This dataset compiles the PSO products for various LEO missions and GNSS constellation in sp3 format.
GNSS Constellation:
- GPS
LEO Satellites:
- ENVISAT
- Jason-1
- Jason-2
- Jason-3
- Sentinel-3A
- Sentinel-3B
- Sentinel-6A
- TOPEX
Each solution follows specific requirements and parametrizations which are named in the respective processing metric table.