First Assessment of the Connectivity of the River Network in the Aral Sea Basin in Central Asia: Challenges of Large Scale Connectivity Modeling in Data-Scarce Regions

Connectivity is crucial for the functioning of river corridors as it determines the natural flow and sediment regime as well as the ability of species to migrate. Thus, it is a property highly relevant for the development of riverine landscapes and their potential of ecosystem service provision. Today, the connectivity of most (large) rivers is affected by anthropogenic infrastructure such as hydropower dams. This is also true for the Aral Sea Basin in Central Asia. The importance of rivers as freshwater resource led to an intensive exploitation of water resources and to the construction of a large number of dams and thus to a fragmentation of the river network. Despite its relevance for the functioning of the river corridors, connectivity remains unexplored for this  basin. This is partly due to the fact that the large scale assessment of connectivity for such data-scarce regions is challenging. For instance, there is the need to delineate a robust and accurate river network from globally available digital elevation models (DEM) as readily available datasets like the Hydrosheds river network suffer from significant errors in this region. In this study, we present a first assessment of the connectivity of the river network in the Aral Sea Basin. In addition, we discuss the challenges associated with large scale modeling of structural connectivity of river networks in data-scarce regions and how to overcome them.

We take as a basis a channel network delineated from the 30 m Copernicus DEM along with geomorphon-based major geomorphological units to derive landscape-specific channel initiation thresholds. We use a least-cost path approach for flow routing to avoid artifacts resulting from sink filling. Multispectral satellite time series from the Landsat mission are used to remove abandoned channels and to correct the river network. Additional input are the barriers in the Aral Sea Basin. We use the dam data from Global Dam Watch and complement it by mapping from high resolution Google Earth imagery. The river network and the barrier locations are used to create a graph representation of the river network where river reaches are represented by edges and confluences as well as dam locations by nodes. This river graph is used to compute connectivity metrics such as the dendritic connectivity index, for both the whole network and at the subcatchment scale.

The results of our study deliver the first analysis of connectivity of the river network in the Aral Sea Basin. Along with the insights in this particular river basin, we present an approach which is optimized for the application in large, data-scarce study areas. Such static analysis of structural connectivity is of course a first indicator only, and further analysis is required to understand the impact of hydrological and sediment connectivity on the riverine landscapes of the river corridors of the region. Thus, rather than a final result, we see our study on river network connectivity as an important basis for assessing sediment dynamics across the network, natural flow regime and its impairment as well as river and floodplain habitat integrity.