Estimating Saturated Hydraulic Conductivity Along a 860 Km Transect in the Loess Plateau of China.

Saturated hydraulic conductivity (Ks) and its spatial variability are essential for the modelling of field water and solute transport. The measurement of Ks, however, is expensive and time-consuming. The objectives of the current study was to estimate Ks in China’s Loess Plateau based on routinely measured soil properties, e.g., soil texture, bulk density, soil organic matter, as well as climatic and terrain factors, e.g., precipitation, elevation, using multi-regression, hierarchical neural network and autoregressive state-space approach. A transect of 860 km (34°05’N-39°23’N, 108°37’E-110°19’E) was selected along the south-north plane of the Loess Plateau, on which 86 sites were investigated at an interval of 10 km from Apr. 15 to May 15, 2013. Using a 5-cm soil auger, soil cores of 0-10, 10-20 and 20-40 cm were collected and measured for Ks, bulk density, soil texture, soil organic matter and saturated soil water content. In the meanwhile, mean annual precipitation and temperature for each site were derived based on the meteorological data during 1953-2013 derived from China Meteorological Data Sharing Service Systems. Elevation and local slope were measured with a GPS receiver (5-m precision) and a geological compass, respectively. The Ks estimations obtained by the three methods were then compared and the primary factors affecting Ks were further explored. These results hold important implications for hydrological modelling and irrigation management.