Integrating topographic characteristics to construct lake and catchment topology on the Tibetan Plateau

The Tibetan Plateau, the Third Pole of the world. Accurate lake–catchment topological relationships are essential for hydrological modeling in endorheic systems. Traditional depression-filling methods often misclassify real lake basins as artificial depressions, creating incorrect connectivity. Many datasets also suffer from temporal mismatches between lake extents and digital elevation models, reducing reliability. To overcome this, we developed a lake-oriented algorithm that preserves terrain integrity while correctly identifying true endorheic basins. Lake boundaries are directly extracted from the Shuttle Radar Topography Mission digital elevation model circa 2000, ensuring full temporal consistency between lakes and topography. We produced the first temporally consistent lake–catchment network for the Tibetan Plateau as of 2000, containing 3985 lakes. Of these, 3743 are endorheic and 242 are exorheic. Lakes are categorized as 3764 starting, 19 middle, and 3673 terminal nodes, with 3471 serving simultaneously as both starting and terminal lakes. Our method significantly improves topological accuracy by eliminating errors from misclassified depressions and timing mismatches. This 2000 baseline offers a robust reference for tracking post-2000 hydrological changes driven by climate change and provides a transferable framework for studying water system evolution in high-altitude regions experiencing rapid environmental change. The topological relationships between lakes and their catchments serve as a fundamental basis for quantifying material and energy inputs to lakes. These relationships can be categorized into cascading connections among lakes and nested configurations within lake catchments. Developing an accurate algorithm to construct such topological structures is therefore essential for limnological and hydrological studies. However, previous approaches have been limited in their ability to capture the dynamic topological relationships within a defined temporal window. Moreover, conventional depression-filling algorithms are unable to distinguish true depressions from artificial artifacts introduced during digital elevation model (DEM) processing. To address these limitations, this study proposes a novel algorithm and a corresponding dataset, termed LCTopo, which integrates detailed topographic characteristics of lakes. The proposed method is compared with traditional algorithms and datasets to evaluate its performance. Our approach effectively avoids unrealistic configurations and incorporates a greater number of lakes when establishing nested relationships. Applying the algorithm to the Tibetan Plateau, we extracted 3985 lakes and their associated catchments, and constructed their hierarchical topological relationships. Of these, 3743 were endorheic and 242 were exorheic. Furthermore, 19 lakes were classified as middle lakes, 3673 as terminal lakes, while 3471 lakes were identified as both starting and terminal lakes. The proposed algorithm and the LCTopo dataset provide a valuable resource for hydrologists, climatologists, and geomorphologists seeking to better understand lake-catchment interactions. • Develop an algorithm to fill fake depressions and reserve true depressions. • Propose an algorithm to construct lake and catchment topology. • Develop a dataset about topology among lakes on the Tibetan Plateau.