The construction of reservoirs by "gully land consolidation" leads to soil salinization

and the permeability of saline soil is an important factor affecting engineering design and construction safety. Therefore

it is important to clarify the remolded loess permeability under different salt content and type conditions. In this study

a series of indoor variable head permeability tests was carried out to investigate the variation of permeability coefficient of remolded loess with salt content by preparing sodium and calcium salt remolded loess specimens with different salt contents. The results indicated that with the increase of salt content

the saturation permeability coefficient of remolded loess gradually decreases

and the two are negatively correlated. Moreover

the permeability coefficient of salt-containing remolded loess decreased to varying degrees compared with that of remolded loess without salt. Under the same salt content

the permeability coefficient of calcium-salt loess was smaller than that of sodium-salt loess. The variation of pore water fugacity and pore structure with salt content and type within the specimen was tested using nuclear magnetic resonance (NMR) techniques. When the salt content was below 2%

the permeability of remolded loess decreased rapidly with the increase of salt content

due to the changes in pore water occurrence state and pore structure. However

when the salt content exceeded 2%

the permeability of the remolded loess increased with the change of pore water occurrence state

while the permeability of the remolded loess decreased with the change of pore structure

and the change of pore structure had a dominant influence on the permeability

thus the permeability coefficient of the remolded loess decreased slowly. Furthermore

under identical salt content

the changes in the pore water occurrence status and pore structure of the calcium salt remolded loess were more significant than those in the sodium salt remolded loess.