Enhancing understanding of lake abundance and dynamics, and the interaction between water resources and human settlements

Abstract

Freshwater resources are under increasing pressure due to intensifying climate change, population growth, and rising human water demand, highlighting the need to better understand both natural hydrological processes and human–water interactions. However, significant gaps remain in our understanding of lake systems, including (1) the lack of a consistent framework for distinguishing freshwater and saline lakes, (2) limited knowledge of intra-annual lake dynamics, particularly for small and medium-sized lakes, and (3) insufficient quantitative assessment of how human activities reshape hydrological systems. To address these challenges, this dissertation presents three independent yet disciplinarily related studies. First, in Chapter 2, a scalable framework is developed and validated to classify freshwater and saline lakes by integrating hydrological, climatological, spectral, and literature evidence, thereby overcoming the limitations of conventional approaches that rely on in situ salinity measurements. The classification synthesizes multiple sources of information, including hydrological settings (e.g., located at the drainage terminal), climatic conditions (e.g., arid and semi-arid regions), remotely sensed features (e.g., lacustrine evaporite signatures), and documented records from the literature. Using an independent regional lake survey dataset for validation, our final classification reached an overall accuracy of nearly 70% in terms of lake count and 94% in lake area, implying that the efficacy of our method tends to favor relatively large lakes. Second, leveraging high-resolution, multi-temporal water surface elevation (WSE) observations from the Surface Water and Ocean Topography (SWOT) mission, Chapter 3 systematically analyzes intra-annual variations in water level and storage for more than 30,000 lakes across the Tibetan Plateau from July 2023 to May 2025, which serves as a representative case study region. Building on the freshwater and saline lake classification, lakes are further grouped by both type (freshwater and saline) and size (small: <0.1 km²; medium: 0.1–1 km²; large: >1 km²) to examine seasonal dynamics within a unified analytical framework. The results show that the intra-annual storage variations of different lake types can be clearly captured, demonstrating the strong capability of this dataset in characterizing seasonal lake dynamics. Although temporal patterns are similar across lakes, the magnitude of variability differs substantially: WSE variability remains largely independent of lake size, whereas storage variability shows a strong scaling relationship with lake area (slope = 1.04, r = 0.88). Large lakes, despite representing only a small fraction of total lake count, contribute over 90% of regional storage variability. In addition, lake type modulates intra-annual dynamics, with saline and terminal lakes exhibiting stronger seasonal signals and earlier peak timing, likely suggesting distinct hydrological controls. Third, to quantitatively assess how floodplain urban expansion is linked with levee construction, Chapter 4 develop a multi-scale composite analysis framework leveraging a national levee database and decades of annual land cover maps. Results show that in the contiguous US, levee construction is associated with a 62% acceleration in floodplain urban expansion, outpacing that of the county (29%), highlighting a clear change in risk perception after levees being built. Regions historically lacking strong momentum for population growth while experiencing frequent floods tend to rely more strongly on levees and we suggest these areas develop a more diversified portfolio to cope with floods. Temporally, the positive levee effect is found to have weakened and then reversed since the late 1970s, reflecting the role of legislative regulations to suppress floodplain urban expansion. Our quantitative framework sheds light on how structural and non-structural measures jointly influence floodplain urban growth patterns. It also provides a viable framework to objectively assess the floodplain management strategies currently in place, which may provide useful guidance for managing flood risks towards sustainable development goals. Overall, by integrating lake classification, hydrological dynamics analysis, and human–water interaction assessment, this dissertation provides a comprehensive characterization of surface water systems from both natural and anthropogenic perspectives. Through the integration of multi-source data and the development of quantitative analytical frameworks, this work advances the understanding of lake hydrological processes and offers new insights into how human activities influence water systems, thereby contributing to improved water resource management and sustainable development.