利用长江防洪实体模型进行了2002年10月地形条件下长江荆江河段遭遇“54年型”和“98年型”特大型洪水时的洪水演进特性实体模型试验研究。研究表明：试验条件下洪峰从枝城传播至监利需14～21 h，平均传播速度3.2～4.8 m/s。当遭遇“54年型”洪水且不考虑三峡水库调蓄时，各站试验洪峰水位与1954年实际洪水相比普遍升高，荆江面临极其严峻的防洪形势，沙市站洪峰水位超过堤防设计水位约 1.60 m，石首、监利和城陵矶等站超过设计堤顶高程达 0.4～1.23 m；考虑三峡水库调蓄后，各站试验洪峰水位较不考虑调蓄情况明显降低，防洪形势有明显缓解，但依然不容乐观，其中沙市站洪峰水位超过堤防设计水位0.30 m，莲花塘站超过设计堤顶高程0.28m。当遭遇“98年型”洪水且不考虑三峡水库调蓄时，各站试验洪峰水位与1998年实际洪水相比普遍升高，荆江防洪形势严峻，其中沙市、石首、监利和莲花塘站分别超出堤防设计水位达1.20，1.25，1.73，1.68 m。

Adopting the Yangtze River Flood Protection Physical Model, experimental study on the propagation features of the 1954 and 1998 floods in the Jingjiang Reach of the Yangtze River have been conducted under the topographical condition of Oct. 2002. The experimental results reveal that it takes 14～21 hours for the flood peak to transport from Zhicheng to Jianli with an averaged speed of 3.2～4.8 m/s. When experiencing the 1954 floods, the experimental flood peak stages (FPS) are generally higher comparing with the prototype floods in 1954 when operation of the Three Gorges Project (TGP) is not considered. The flood situation of the reach is extremely severe: the FPS at Shashi exceeds the design stage of the dike by 1.60 m , and the FPS at Shishou, Jianli and Lianhuatang exceed the design crest level of the dike by 0.40～1.23 m. After consideration of the TGP the experimental FPS drop considerably. The flood protection situation of the reach is relaxed substantially, however, not enough to be optimistic. The FPS at Shashi exceeds the design stage of the dike by 0.3 m, and that at Lianhuatang exceeds the design crest level of the dike by 0.28 m. When experiencing the 1998 floods, the experimental FPS are generally higher comparing with the prototype floods in 1998 when operation of the TGP is not considered. The flood protection situation of the reach is still severe: the FPS at Shashi, Shishou, Jianli and Lianhuatang exceed the design stage of the dike by 1.20 m, 1.25 m, 1.73 m and 1.68 m respectively.