Abstract: 【Objective】The ecological environment in Tarim basin of southern Xinjiang is extremely fragile, as it’s in the arid inland river basin. As the major high quality cotton production base, mulching, drip irrigation and other water-saving irrigation technologies have relieved the contradictions of the water shortages during cotton growth period, but the winter-spring irrigation quota is relatively high. The infiltration characteristics and its influencing factors of drip irrigation cotton fields during fallow period under different surface mulching and winter irrigation quotas in south Xinjiang were studied with the aiming atproviding a theoretical basis for the spring irrigation quota and irrigation efficiency.【Method】The double-ring infiltrometer was used to measure the infiltration in a constant-head experiment. Six treatments for winter irrigation with bare fields (WIB, and the winter irrigation quota was 200 mm, the same as below), winter irrigation with high stubble fields (WIS), winter irrigation with straw mulching fields (WIM), no-winter irrigation with bare fields (NWIB), no-winter irrigation with high stubble fields (NWIS) and no-winter irrigation with straw mulching fields (NWIM) were designed. The soil infiltration characteristics and simulation infiltration process were compared and the infiltration factors were analyzed.【Result】The results showed that winter irrigation reduced initial infiltration rate, stable infiltration rate and cumulative infiltration, while WIB, WIS and WIM decreased the initial infiltration rate, stable infiltration rate and cumulative infiltration compared with NWIB, NWIS and NWIM by 22.51%, 21.49% and 12.94%, 75.86%, 27.27% and 42.86%, 61.98%, 22.18% and 29.31%, respectively. NWIM and WIM reduced the initial infiltration rate compared with bare field and high stubble fields by 14.74%, 15.20% and 4.20%, 5.96%, respectively. NWIM and WIM stable infiltration rate and cumulative infiltration showed bare fields＞straw mulching fields＞high stubble fields, stable infiltration rate of NWIB and WIB were 2.96 times, 4.39 times and 1.25 times, 1.46 times larger than straw mulching fields and high stubble fields. Cumulative infiltration NWIB was 2.13 times, 2.39 times, 2.63 times, 3.01 times and 3.07 times larger than NWIM, NWIS, WIB, WIS and WIS. In consideration of determination coefficient (R²) and root mean square error (RMSE), regression analysis of model simulation indicated that general empirical model and Kositiakov model were optimal for describing the soil infiltration process while Horton model was the second, Philip infiltration model was the third. Correlation analysis showed that the soil initial water content and soil bulk density have a negative relationship, but soil temperature and soil porosity have a positive relationship with initial infiltration rate, stable infiltration rate and cumulative infiltration. Initial moisture content of the shallow affected soil infiltration parameters obviously compared with deep soil water content. Initial infiltration rate showed a significant correlation with initial soil water content, soil temperature, soil bulk density and soil porosity at the depth of 0-10 cm soil, stable infiltration rate and cumulative infiltration showed a significant correlation with initial soil water content, soil bulk density and soil porosity at the depth of 0-10 cm soil. 【Conclusion】Winter irrigation and mulching reduced soil infiltration capacity during spring fallow period in south Xinjiang, and this method is beneficial to reducing spring irrigation quotas and improving irrigation efficiency.

Key words: South Xinjiang, drip irrigation cotton fields, soil infiltration, soil infiltration model