Written in collaboration with Dr. Ali Ameli

Most of the world’s flowing waters remain unmeasured, posing one of the biggest challenges to water security worldwide. This lack of data makes it difficult to forecast floods, protect water quality and sustain aquatic ecosystems. A new study led by UBC researchers establishes the first globally scalable framework for the functional classification of catchments, grouping them by how consistently rainfall is converted into runoff.

Published in Nature Water, the study mapped how rainfall becomes streamflow–the movement of water from the land to water bodies–across more than 80,000 catchments–areas of land where rainfall and other water drain into a common outlet. The research was led by Dr. Ali Ameli, Associate Professor in UBC’s Department of Earth, Ocean and Atmospheric Sciences, with Hamed Sharif, Researcher in the department, and Dr. Jeffrey McDonnell, Professor at the University of Saskatchewan and the University of Birmingham. 

The global picture was striking. “For most of the planet, knowing how much rain fell tells you surprisingly little about how much water will reach the river,” said Dr. Ameli. Catchments classified as functionally “complex” drain 87% of the ungauged land area assessed in the study, some 121 million of 139 million km². In these catchments, the same amount of rain can produce very different amounts of streamflow from one storm to the next, making river responses to storms difficult to anticipate.

Western British Columbia, however, stands out as one of the clearest exceptions to this complex global pattern. Along BC’s steep, wet Coast Mountains, the researchers found a broad band of catchments that behave much more consistently: the amount of rain falling during a storm is closely related to the amount of runoff reaching the river. These catchments were classified as functionally “simple”. “This consistency is rare globally and was one of the most striking surprises of the study,” said Dr. Ameli.

Why these catchments behave so consistently comes down to the rhythm of their rainfall. Across the study, the strongest single discriminator between the functional types was rainfall persistence, the fraction of each season carrying non-negligible rain. Simple catchments receive rain frequently and stay persistently wet, so successive storms meet similar antecedent conditions. Complex catchments receive rain only intermittently and swing between wet and dry states.

Image

Two catchments at opposite ends of the classification. Left: the Rio Itapecuru at Rosário, in Maranhão, northern Brazil. The catchment is classified as complex; its mean slope is about 3 degrees with no persistent rainfall falls in the dormant season. Credit: Johan Fredriksson, CC BY-SA, via Wikimedia Commons. Right: Barkley Sound near Bamfield, on the west coast of Vancouver Island, within the broad band of simple catchments identified along the Coast Mountains. The catchment here has a mean slope of about 21 degrees, and receives non-negligible rain in every week of the dormant (non-growing) season. Credit: UBC.

By classifying and locating these different functional types of catchments, the study provides a global blueprint for prioritizing streamflow monitoring and selecting flood and water-resource models. Relatively simple catchments may not require highly complex models, whereas the complex catchments that dominate much of the world require models capable of representing changing hydrologic behavior. A supplementary modeling experiment makes the point concretely: a lumped, parsimonious conceptual model reproduced simple catchments well but failed in complex ones, a structural shortcoming that no amount of calibration could overcome.

The team has also released an interactive global map of the results, together with a web application that returns season-specific functional classes for any catchment, either from a boundary polygon the user uploads or from an outlet selected on the map.
 

Read more:

Why most rivers don’t respond to storms the same way twice

A global classification of hydrologic functional diversity in gauged and ungauged catchments