Bo Yao

Lake monitoring is essential for sustaining aquatic ecosystems, and accurate estimation/prediction of water quality parameters is crucial to this effort. Despite its importance, the performance of predictive models built on varying temporal resolutions remains underexplored systematically. This study used daily and 4 h high temporal resolution (HTR) datasets to assess the performance of multiple machine learning models—namely, Support Vector Regression (SVR), Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Long Short-Term Memory (LSTM) networks—under consistent data scales. The results indicate that dissolved oxygen (DO) exhibits pronounced sensitivity to temporal resolution, while total nitrogen (TN), total phosphorus (TP), and ammonia nitrogen (NH₃-N) show distinct, parameter-specific response patterns that align with the temporal characteristics of their underlying biogeochemical processes. This research helps to deepen the understanding of how temporal data resolution influences model performance in water quality prediction, offering valuable insights for selecting optimal data resolutions and modeling techniques to enhance lake monitoring and protection strategies.

Climate change and human activities cause lake water level (WL) fluctuations to exceed natural thresholds, with implications for the available water resources. Studies that explore WL change trends and the main driving forces that affect water level changes are essential for future lake water resource planning. This study uses the Mann–Kendall trend test method to explore the WL fluctuations trend and WL mutation in Erhai Lake (EL) during 1990–2019 and explore the main driving factors affecting water level changes, such as characteristic WL adjustments. We also use the principal component analysis to quantify the contribution of compound influencing factors to the water level change in different periods. The results showed that the WL rose at a rate of 47 mm/a during 1990–2019 but was influenced by the characteristic WL adjustment of EL in 2004 and the WL mutation in 2005. In 1990–2004, the WL showed a downtrend caused by the increase in water resource development and utilization intensity, and in 2005–2019, the WL showed an uptrend caused by the combined decrease in evaporation, outflow, and the increase in water supply for water conservancy projects. Additionally, the largest contributions of outflow to WL change were 19.34% and 21.61% in 1990–2019 and 1990–2004, respectively, while the largest contribution of cultivated area to WL change was 20.48% in 2005–2019, and it is worth noting that the largest contribution of climate change to WL change was 40.35% in 2013–2019. In the future, under the increase in outflow and evaporation and the interception of inflow, the WL will decline (Hurst exponent = 0.048). Therefore, planning for the protection and management of lakes should consider the impact of human activities, while also paying attention to the influence of climate change.

Water samples at both surface and bottom layers were taken from 102 sites in Lake Taihu to study the fluorescent components and spatial patterns of chromophoric dissolved organic matters (CDOM). Three-dimensional excitation-emission matrix data obtained from the samples were analyzed by parallel factor approach in which four humic-like and two protein-like fluorescent components (named C1–C6) were identified. The results showed that fluorescence intensities were higher in the northern and western lake regions, and notable declines of fluorescence maxima (F _max) were observed from the northwest to the center and then to the southeast of the lake. Calculated biological index (BIX) values ranged from 0.88 to 1.44 and humification index (HIX) values from 0.64 to 3.37 for all the samples. The spatial variations of BIX and HIX values suggested stronger allochthonous CDOM characteristics in Zhushan Bay and the western area and autochthonous characteristics in the southern and eastern areas. Vertically, the average F _max value of the surface samples was about 6 % less than that of the bottom samples, but noticeable variations existed among different sampling sites and components. These distribution characteristics of CDOM were mainly attributed to the spatial heterogeneity of sources and wind-induced transportation process. Interestingly, the C6 component (Ex_max/Em_max = 250/455 nm) seemed to be unique in samples from Zhushan Bay and probably resulted from the discharge of the Taige River. Therefore, it could be used as an indicator of point-source discharge and a tracer to study the fate of CDOM in the lake.