Abstract. In recent years, wireless sensor network (WSN) has emerged as a new technique to collect Earth observation data at a relatively low cost and minimal labor over large areas. However, WSN observations are still point data. To determine the spatial distribution of a land surface parameter, interpolation of these point data is necessary. Some geostatistical interpolation methods, such as the Ordinary Kriging method, Co-Kriging method, and Regression Kriging method, have been used in various fields. However, capturing the spatial distribution pattern of heterogeneous land surface parameters is still difficult. For example, near-surface soil moisture is a critical parameter for agriculture management, and hydrological and ecological research. However, as soil moisture is related to many factors such as topography, soil type, and vegetation, even a WSN observation grid is not sufficiently dense to reflect its spatial distribution pattern. This study developed a method to interpolate WSN-measured soil moisture with the aid of remote sensing images. The underlying idea is extension of the traditional Kriging algorithm by introducing spectral variables, specifically, vegetation index (VI) and albedo, from satellite imagery as supplementary information to aid interpolation. Thus, the new Extended Kriging algorithm operates on spatial and spectral combined space. The algorithm has been applied to WSN-measured data in the HiWATER campaign to generate daily soil moisture maps in the 5 km ×: 5 km oasis area in the middle reaches of the Heihe River, western China, from June 10 to July 15, 2012. Visual inspections indicate that the result from the Extended Kriging algorithm shows more spatial details than that of the traditional Kriging algorithm, and the temporal variation of patch-average soil moisture is, in general, consistent with precipitation/irrigation data. Leave-one-out cross-validation was also adopted to estimate the interpolation accuracy. The Root Mean Square Error (RMSE) of the Extended Kriging method was also found to be smaller than that of the original Ordinary Kriging method. Analysis with minimum variance of error (σ_k), a self-uncertainty indicator given by the Kriging algorithm, also gave the same conclusion. Further testing also indicated that if high-resolution land surface temperature maps are available, they can be added to the spectral variables and further improve the interpolation accuracy.