洞庭湖枯水期水位时空分异特征及其对三峡水库补水调度的响应机制

The eco-environmental problems caused by the extremely low water level of river-connected lakes in the middle reaches of the Yangtze River have drawn much attention in recent years. Increasing the release from the Three Gorges Reservoir(TGR) to raise the lake water level is most likely an effective measure to improve the eco-environment of the lakes. However, due to the special shape and inflow conditions of the Dongting lake, the lake water level has a strongly spatial and temporal variability during the dry season, so the response mechanism of the water level in different lake areas to the release from the the TGR is still worthy of further study. In this project, taking the Dongting lake, a typical river-connected lake in the middle reach of the Yangtze River as an example，first we will expound the spatio-temporal evolution of water level in different lake areas in Dongting lake based on long time series water level data from different lake areas, and explore the driving factors that cause the water level variation according to the relationship between the water levels of Dongting Lake and discharge from the Yangtze River and its tributaries; then we will reveal the mechanism of the Zhicheng discharge on water level of different lake areas using the data-driven model based on support vector machine optimized by particle swarm optimization algorithm(PSO-SVM); at last we will establish the dynamic and quantitative responses of the discharge at Zhicheng station to the release from the TGR based on the random forest algorithm, and ultimately establish the dynamic and quantitative responses of the lake water level to the release from the TGR. The research results will certainly provide an important theoretical and scientific basis for formulating the water supplement programs to improve the Dongting lake ecological environment.

长江中游通江湖泊连续多年出现特枯水位引起的生态环境问题备受关注，通过三峡水库补水调度最有可能成为近期改善湖泊生态环境问题的有力措施，然而由于洞庭湖特殊的湖盆形态入湖水流条件导致枯水期水位具有明显的空间异质性，不同湖区水位对三峡水库下泄流量的响应机制仍需进一步深入研究。本项目以长江中游典型通江湖泊洞庭湖为研究对象，基于不同湖区长时间序列水位资料阐明枯水期洞庭湖水位分异特征，依据湖区水位与干支流流量的关联性识别不同湖区水位变化的驱动因素；构建基于粒子群优化支持向量机的洞庭湖水位预测模型，揭示枝城站流量对不同湖区水位的作用机制；采用随机森林回归模型建立枝城站流量对三峡水库下泄流量的动态定量响应关系，最终确定洞庭湖水位时空分异特征对三峡水库补水调度的响应机制。研究成果可为改善洞庭湖生态环境的三峡水库补水调度方案的制定提供理论基础和科学依据。

本项目以长江中游典型通江湖泊洞庭湖为研究对象，采用不同方法揭示了洞庭湖水位时空分异特征对三峡水库补水调度的响应机制，在此基础上为定量描述三峡水库枯水期不同运行方式对洞庭湖生态补水效果的空间差异及候鸟栖息地生境的影响，以洞庭湖典型的珍稀越冬水鸟—白鹤为指示性候鸟，以白鹤摄食对栖息地水深需求作为关键生态因子，建立白鹤摄食对水深需求的栖息地适宜度模型。构建涵盖长江干流、三口河系、洞庭湖及其四水尾闾河段的江湖一体化耦合水动力模型，实现了洞庭湖水动力分布特征的精确模拟，建立了面向白鹤摄食对三峡水库出库流量需求的物理栖息地模型，量化不同补水方式对白鹤摄食栖息地面积的影响。结果表明：（1）枯水期洞庭湖不同湖区水位受三峡出库流量和四水入湖流量共同影响，但各湖区的主要影响因子存在差异，其中城陵矶和东洞庭湖鹿角站水位受三峡出库流量影响最大，关联度分别为0.9837和0.9829；南洞庭湖受资水影响最大，关联度为0.9817；西洞庭湖受沅江影响最大，关联度为0.9819；（2）三峡水库提前消落和高水位消落方式对洞庭湖水位的影响具有明显的空间异质性，对东洞庭湖北部影响较为明显，而对于东洞庭湖南部、南洞庭湖和西洞庭湖影响较小。提前消落较均匀消落方式城陵矶水位平均提高0.120m，对应的三峡电站发电量减小0.30%，而高水位消落方式城陵矶水位平均降低0.09m，发电量增加0.28%；（3）枯水期三峡水库不同出库流量下洞庭湖白鹤潜在摄食栖息地面积保持稳定并随出库流量的增加呈增大趋势，维持在101.40-121.84km2之间，其中东洞庭湖摄食栖息地面积在7.49-9.86km2之间，南洞庭湖（含横岭湖）摄食栖息地面积在47.37-60.34km2之间，西洞庭湖摄食栖息地面积在46.54-51.64km2之间。不同湖区摄食栖息地面积随着三峡水库出库流量的增加均呈增大的趋势，说前三峡水库枯水期补水调度对于维持栖息地面积具有重要作用。较三峡水库运行前相比，白鹤摄食栖息地面积最大增加20.44km2，对应的增幅为20.16%。成果明晰了三峡水库运行对洞庭湖白鹤摄食栖息地面积的影响规律，可为通过三峡水库补水调度改善洞庭湖越冬水鸟摄食栖息地生境提供理论基础。