Spatial pattern and interrelation of soil water content and thermal conductivity in greenhouse

Abstract: Soil hydrothermal coupling process widely exists in engineering and agriculture field. As 2 important physical features of soil, soil water content and thermal conductivity are meaningful for soil organic matter accumulation, chemical reaction rate and coupled transfer of water and heat. Many relevant tests were completed in laboratories, and test conditions were mostly artificial conditions instead of open air conditions, under which research area was different from one another, which could make the stability influenced easily by other factors. This experiment was conducted under a stable greenhouse condition and research area was in small scale, which helped to make sure the similarity of different sampling areas. To explore the distribution characteristics and relevance between soil water content and thermal conductivity, an 8 m × 8 m tomato field in the experimental greenhouse was selected. The division was conducted with 1 m × 1 m grid and 64 sampling points were set in total. The sampling method was to use soil drill to obtain samples from 0-40 cm depth soil layer. Drying method was used for measuring soil water content. By means of the portable soil thermal conductivity instrument (MTN01), which was based on the non-steady-state probe (NSSP) technology, thermal conductivity of the 0-20 cm depth soil layer was measured. On the basis of classical statistics, geostatistics, regression analysis and spectrum analysis theories, the distribution characteristics and relevance between soil water content and thermal conductivity were studied. Contour map were used to analyze the spatial distribution characteristics of soil water content and thermal conductivity. Also spectrum analysis of soil thermal conductivity on the sequence of soil water content was carried out. According to the principle of spatial spectrum analysis, soil water content was used as distance sequence with thermal conductivity as variable sequence, 32 intervals were set up with step length of 0.35%, and then the Fourier transform was conducted. The result showed that soil water content at the 0-40 cm depth had a trend of increasing firstly and then decreasing, and the peak of the average value was observed in depth of 20-30 cm. For the thermal conductivity, the 10-20 cm depth layer was 15.60% higher than the 0-10 cm layer. There was a strong spatial relationship between soil water content and thermal conductivity in all depths of soil layer. In addition, in the experiment, the spatial correlation range caused by random factors was relatively low. The smallest range (1.31 m) was larger than the sampling distance (1 m), which meant the grid arranged can meet the requirement of spatial analysis. On the condition of uniform water supply, the evaporation intensity of soil layers of different depths in the experimental field and the soil water content of adjacent field would influence the spatial distribution of soil water content. When the soil water content varied between 17% and 28%, for 0-20 cm soil layer, the spatial distribution of soil water content and thermal conductivity were highly correlated, and the result of regression and fit showed that it was a linear positive correlation (R2=0.837). The result of spectrum analysis showed thermal conductivity had a long-term negative correlation in the sequence of soil water content.