A new study suggests that the commonly used split sample approach in hydrology, where time series are divided into a part for model calibration and a part for model validation, should be abandoned.
Marc F. P. Bierkens
Editor, Water Resources Research
Posted inEditors' Highlights
Combining Deep Learning Methods with Process-based Models
Using turbulent heat fluxes as an example, a new study shows that exchange of information between process-based models and deep learning methods may lead to improved predictions.
Posted inEditors' Highlights
Gauging Ungauged River Basins with Smart Remote Sensing
A clever combination of hydrologic modelling and discharge estimates from the Landsat satellite provides good discharge estimates throughout the Missouri river basin.
Posted inEditors' Highlights
The Ocean-Land Connection of Droughts
Around 16 percent of large-scale droughts over land originate above the ocean and these types of droughts are more extensive and severe than droughts that originate over land.
Posted inEditors' Highlights
Machine Learning Predicts Subsidence from Groundwater Pumping
Machine learning and data on aquifer type, sediment thickness, and proxies for irrigation water use has been used to produce the most comprehensive map of land subsidence in the western U.S. to date.
Posted inEditors' Highlights
Arsenic Pollution in Bangladesh is Catching Up with Deeper Wells
Inhabitants of Bangladesh have deepened drinking water wells to avoid extracting arsenic-rich groundwater from shallow aquifers, but these may not be free from pollution either.
Posted inEditors' Highlights
Reconstructing Natural Streamflow at Unprecedented Resolution
A new research effort has mapped 35 years of naturalized streamflow for 2.94 million river reaches worldwide: an invaluable dataset for hydrology, biogeochemistry, ecology, and remote sensing.
Posted inEditors' Highlights
Groundwater Drawn Downward After Kumamoto Quake
A unique set of high-frequency groundwater level monitoring reveals a loss of approximately ten million cubic meters of groundwater after a major earthquake.
Posted inEditors' Highlights
Calibrating Hydrological Models by Satellite
Hydrological models are usually calibrated using observations of streamflow, but a new method uses remotely sensed land surface temperature for this purpose.
Posted inEditors' Highlights
A City’s Challenge of Dealing with Sea Level Rise
A well-developed case study in Ho-Chi Min City, Vietnam, exemplifies how other mega-cities located on deltas could face the major challenge of adapting to rising sea-level.