Climate change and eutrophication have significant impacts on aquatic ecosystems, with phytoplankton playing a critical role as primary producers. This study investigates the relationships between phytoplankton communities and environmental and meteorological factors in Wuliangsuhai Lake, a representative ecosystem in an arid and cold region of China. Using data from 25 sampling sites (April–September 2023), a generalized additive model (GAM) was employed to analyze water quality and laboratory measurements. The results showed a bimodal distribution of phytoplankton density, with peaks in July (5.33 × 10⁶ cells/L) and August (14.90 × 10⁶ cells/L). Green algae dominated in spring, while cyanobacteria became dominant in summer. GAM analysis revealed that temperature (TEMP) was the primary driver, explaining 20.7% of the deviance. When TEMP was examined together with other factors, the explanatory ability of the model was significantly enhanced, and finally, the model explained 57.10% of the deviance. GAM analysis also revealed that different algae species responded differently to environmental factors, with the cyanophyta-dominant species Pseudoalgae being more sensitive to TEMP and pH, whereas the chlorophyta-dominant species Chlorella was more affected by wind speed and salinity. Cyanobacteria dominance, exacerbated by warming, increases the risks of harmful algae blooms and ecosystem instability, highlighting the need for mitigation strategies in vulnerable freshwater ecosystems.

It is imperative to elucidate the process of evaporation in lakes, particularly those that are freshwater and are situated in middle and high latitudes. Based on one-year evaporation and high-frequency meteorological–water quality data of Lake Wuliangsuhai, this study analyzed the applicability and driving mechanism of the evaporation model. These dynamics are elucidated by the vorticity covariance method combined with the multivariate constrained evaporation Modelling method. The findings of this study revealed that (1) Lake evaporation (ET) is affected by multiple meteorological–water quality constraints, and the water quality indicators significantly related to ET are also affected by lake stratification. The coupled meteorological–water quality evaporation model can explain 93% of the evaporation change, which is 20% higher than the traditional meteorological Modelling evaporation model. (2) The nighttime ET is mainly affected by the thermal inertia lag, and the nighttime ET loss in Lake Wuliangsuhai accounts for 37.34% of the total evaporation, which cannot be ignored. (3) The actual water surface evaporation of the lake is much smaller than that measured by the pan conversion method and the regional empirical C formula method. The cumulative evaporation of Lake Wuliangsuhai from the non-freezing period to the early glacial period converted from meteorological station data is 1333.5 mm. The total evaporation in the non-freezing period is 2.77~3.68 × 10⁸ m^3, calculated by the lake area of 325 km², while the evaporation calculated by the eddy station is 1.91 × 10⁸ m³. In addition, the ET value measured by the cumulative C formula method was 424.2% higher than that of the model method and exceeded the storage capacity. Low-frequency and limited environmental index observations may lead to an overestimation of the real lake evaporation. Therefore, in situ, high-frequency meteorological–water quality monitoring and the eddy method deserve more consideration in future research on lake evaporation.