How is the biodiversity of a river related to its properties? And which parameters determine the structure of a river network and its amount of water? A new high-resolution map of global river systems and their characteristics is now helping to clarify this. It provides a basis for more detailed research into what constitutes riverine habitats and how they are related. This is also important for aquatic ecology.
Rivers are the lifelines of the earthly land areas. Because they bring water to dry regions, transport nutrients and offer aquatic dwellers ways for their spread and migration. But what influences the size and shape of watercourses and their catchment areas? Why are rivers more common in some areas than others? And how do climate and the surrounding landscape influence the biodiversity of animals and plants in the water?
From topography to river structure
A new world map of the earth's river systems now provides more information about these questions. Although it is not the first of its kind, it is the one with the highest resolution and the most detailed. "Unlike the previous datasets, Hydrography90m also depicts smaller and smallest branches of watercourses," reports co-author Sami Domisch from the Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB). This level of accuracy can be seen from the name of the new map: the smallest unit is 90 meters in length. Since small rivers make up the largest share of the global river network at around 70 percent, they are particularly important for river-bound biodiversity.
To create the model on which the map is based, the team evaluated a topographical terrain model of the earth based on satellite data. This landscape data forms the basis for the model, because where there are cuts in the landscape with certain properties, there is also the potential for a watercourse to flow. "At first we don't know where a river actually flows," explains Domisch. Additional data is used to determine whether the rivers actually carry water – all year round or at times. These include, for example, precipitation, temperature, land use, soil conditions and slope gradient. In the model, these parameters are related to the water volumes measured by 30,000 measuring stations worldwide.
Answers also for water ecology
It took the team two and a half years and a supercomputer to create the map based on the Hydrography90m dataset. "We work with machine learning, which means that with each additional data set, our model can recognize better and better which parameter values are related to which amounts of water," explains lead author Giuseppe Amatulli from Yale University. As a result, the model can show the probable outflow, i.e. the amount of water in the river, even where there is no measurement data on water quantities. This data and other parameters can now be determined for a total of 726 million river sections worldwide.
In the future, the model can provide answers to crucial questions of aquatic ecology. This is possible primarily because Hydrography90m records the catchment areas of river sections in very small detail and already provides environmental data for these five-hectare catchment areas. This can then be used to characterize the occurrence of species communities, for example with which climate data or with which slope inclination they are connected. From this, conclusions can be drawn about the biogeography of these habitats – where they occur and what factors shape them.
Source: Leibniz Institute for Freshwater Ecology and Inland Fisheries (IGB); Article: Earth System Science Data, doi: 10.5194/essd-14-4525-2022