数据短缺条件下基于概率密度函数优化的LX河生态水位核定研究

Research on Ecological Water Level Determination of LX River Based on Probability Density Function Optimization under Data Shortage Conditions

摘要: 针对中小河流水位监测数据短缺导致生态水位核定不确定性突出的问题，本文以LX河为研究对象，提出了一种基于概率密度函数的年保证率设定法增强方法。以MZ闸为控制断面，基于2018~2021年有限日均水位数据，采用皮尔逊Ⅲ型、Weibull、广义极值、Burr及t位置尺度分布等5种函数进行拟合优度比选，识别最优分布函数以扩充样本代表性，进而以75%年保证率推求生态水位。同时利用曲线相关法建立水位-NDVI响应关系进行对比分析。结果表明，Burr分布拟合优度最佳(0.9814)，据此计算生态水位为28.10 m，较曲线相关法(R² < 0.7)更具可靠性，最终核定LX河生态水位为28.10 m。研究可为水位监测数据短缺情况下中小河流生态水位的计算提供方法借鉴。

Abstract: Addressing the significant uncertainty in ecological water level determination due to the scarcity of water level monitoring data for small and medium-sized rivers, this paper takes the LX River as the research object and proposes an enhanced method based on the annual guarantee rate setting method using probability density function. Using the MZ sluice gate as the control section, and based on limited daily average water level data from 2018 to 2021, five functions—Pearson Type III, Weibull, generalized extreme value, Burr, and t-location scale distribution—were used for goodness-of-fit comparison. The optimal distribution function was identified to expand the representativeness of the sample, and then the ecological water level was estimated with a 75% annual guarantee rate. Simultaneously, the water level-NDVI response relationship was established using the curve correlation method for comparative analysis. The results showed that the Burr distribution had the best fit (0.9814), and the calculated ecological water level was 28.10 m, which was more reliable than the curve correlation method (R² < 0.7). The final ecological water level of the LX River was determined to be 28.10 m. This study can provide a methodological reference for calculating the ecological water level of small and medium-sized rivers when water level monitoring data is scarce.