When soil is only partly wet, predicting how water will move through it is crucial—but measuring that movement directly is very difficult. This study finds a quicker route: let the soil’s water-retaining behavior reveal it. Partly wet soil matters whenever engineers need to understand water moving through the ground.

But the soil’s ability to let that water pass is very difficult to measure. The study uses a soil-water curve, describing how much water the soil holds as different pulling forces draw water through the soil. Using that relationship, the study predicts the unsaturated permeability coefficient, a measure of how easily water moves through partly wet soil.

The central idea is to avoid directly measuring the unsaturated permeability coefficient, using soil-water characteristic curves to predict that coefficient instead. The study gathers four values that describe how loess holds water, using measurements collected from the Loess Plateau in recent years.

It then maps how those values change from place to place and fits the link between the soil’s water-holding behavior and its ability to transmit water. That link becomes a practical calculation for water-related engineering problems in loess.

To turn scattered soil measurements into regional maps, the data are imported into mapping software and used to estimate values between measured locations. The prediction maps also allow for measurement errors, so the regional picture does not pretend that every observation is exact.

The method is checked against measured water-transmission results from other researchers and from different regions. That comparison asks a straightforward question: does the shortcut agree with direct measurements well enough to be useful? The predicted results are basically in line with the experimental results, with the prediction deviation controlled within a stated range.

When the soil sample has low saturation, the deviation is small. When saturation is high, the deviation becomes larger, although it remains within the reported range. The change from the predicted result to the fully wet soil value is relatively smooth and more in line with the actual situation.

This matters because water moves through soil very differently as the soil dries: the measured and predicted values track closely, with the largest mismatch staying within about five ten-millionths of a centimetre per second. Across the three regions, a place in the Loess Plateau can be given a quick and economical approximate estimate of its partly wet water-transmission behavior.

The estimate requires only the relatively simple experiment proposed in the study, rather than a full direct measurement at every location. The soil-water curve records the relationship between water content and the force holding water in the soil, and it is widely used to estimate loess properties.

Because loess water transmission is key to solving engineering problems, understanding and fitting this curve is important. The SWCC describes how soil water content relates to matrix suction, expressing the soil’s basic water-holding relationship in this study. Regional analysis then examines how the curve’s parameters vary and develops a method for calculating unsaturated permeability through linear regression.

Across the Loess Plateau, a simple water-and-soil test can provide a quick, economical estimate of how partly wet loess transmits water, with closer agreement at lower wetness and useful accuracy overall.