GDW consensus global database
The GDW (v1) consensus global database
Here, we present a new river barrier and reservoir database developed by the Global Dam Watch (GDW) consortium that integrates, harmonizes, and augments existing global datasets to support large-scale analyses. Data curation involved extensive quality control processes to create a single, globally consistent data repository of instream barriers and reservoirs that are co-registered to a digital river network. Version 1 of the GDW database contains 41,145 barrier locations and 35,295 associated reservoir polygons with a cumulative storage capacity of 7,405 km3 and an artificial terrestrial surface water area of 302,450 km2.
The GDW consensus global database brings together the GDW foundational datasets (GOODD, GRanD and FHReD) described below, complemented by the Global River Obstruction Database (6113 barriers used) and 1,451 reservoirs from the EC Joint Research Centre’s Global Water Surface dataset, in a globally consistent way, enabling global-scale analyses. It combines barrier data with reservoir polygon data and provides a range of barrier and reservoir attribute information. The GDW database has been co-registered with the global hydrographic databases of HydroSHEDS and RiverATLAS by allocating each barrier and its associated reservoir to a raster cell of the HydroSHEDS drainage direction map. It has also been fully integrated into the global HydroLAKES lake dataset (version 1.1), i.e., each GDW reservoir is part of HydroLAKES.
The database is available for download here and for visualisation here. The associated paper is available here. Whereas the some of the foundational datasets upon which the GDW database is built (below) may continue to be developed, the GOODD and GRanD datasets are fully integrated into the GDW database and thus will be discontinued. The GDW database will be updated periodically and is distributed under a free CC BY license.
Spatial and temporal distribution of global river barriers and reservoirs in the GDW v1 database.
Global distribution of reservoirs and river barriers in GDW v1 database.
Flow diagram of the main methodological steps involved in creating, augmenting, and combining data for the creation of the GDW v1 database.
GDW v1 foundational datasets
Each of the three GDW foundational datasets used in the generation of the consensus database has been developed over many years with a particular focus and methodology.
The Global Geo-referenced Database of Dams (GOODD) maps all dams visible on Google Earth satellite imagery globally; it catalogues the geospatial coordinates of 38,660 dams (25,931 used in GDW v1). GOODD will no longer be updated outside of GDW v1.
The Global Reservoir and Dam Database (GRanD) maps the location and attribute data of 7,320 dams (an unpublished final update to 7,424 dams was used in DGW v1) greater than 15m in height or with a reservoir of more than 0.1km3. GRanD will no longer be updated outside of GDW v1.
Future Hydropower Reservoirs and Dams (FHReD) maps 3,700 dams (205 used in GDW v1) that are under construction or in advanced planning stages. Since it focuses on future dams rather than current ones, FHReD continues to be updated outside of GDW v1.
The GDW v1 database brings these three foundational datasets together and complements them with additional records of the Global River Obstruction Database and with new reservoirs that were derived from the EC Joint Research Centre’s Global Water Surface layers using machine learning techniques. The GDW database has been curated in a globally consistent way to provide a high-quality consensus product that enables global-scale analyses.
