Transportation is one of the most challenging urban subsystems to transform in an environmentally friendly and futuristic way. Many city dwellers use a variety of modes of transportation. An efficient and sustainable urban transportation system must include many modes of transportation for a single trip. Intermodal combinations are essential for urban transportation efficiency. Public transportation and commuting are essential elements of multimodal travel. In urban areas, a mix of bicycles, vehicles, and public transportation is prevalent, while in rural areas, car, and public transportation are more prevalent. By examining the characteristics that lead customers to prefer water transportation over Metrobus and Marmaray, we hope to gain a better understanding of how the Asian and European sides of Istanbul are traversed. The number of participants in the "Maritime Transportation Satisfaction Survey" was 2,343. During this period, a model was built using the survey item "frequency of use" (dependent variable). Numerous survey examines and evaluation methodologies were utilized to determine the effectiveness of this strategy. The study examines the intermodal travel motivations and the evaluation of transportation options by multimodal users. For a successful urban transportation system, urban planning must take into account multimodal travel behavior and user expectations. There are initiatives to improve water transportation in Istanbul. Conventional maritime transportation is inadequate from start to finish. An integrated route optimization method is needed to increase the efficiency of maritime transportation. We believe that by strengthening maritime transportation links will increase water consumption. Before the coronavirus pandemic, 2,343 maritime carriers were evaluated on March 8, 2020. (Different surveys were conducted among the passengers of City Lines, Private Motors, Metrobus, and Marmaray to compare their choices and reasons.) SPSS will be used for data analysis. Multivariate Statistical Analysis relies on Multinomial Logistic Regression and Discriminant Analysis models, both of which use the K-fold and Leave-one-out criteria to decide which attributes are valid in the regression model and which are valid in the discriminant approach. The Hosmer-Lemeshow test criteria yielded a p-value greater than 0.05 for MLR characteristics.