Measurement of soil water content using distributed temperature sensor with heated fiber optics

Abstract: Soil water content has great spatial-temporal variability, so accurate field-scale soil water content acquisition with high spatial-temporal resolution is of grave significance for precision agriculture. The distributed temperature sensing (DTS) technology combined with the principle of heat pulse probe is expected to achieve this goal; however, there isn't research to compare and evaluate the advantages and disadvantages of different methods to estimate soil water content using heated DTS technology. In this study, the indoor soil tank tests were carried on to heat the carbon-fiber optical cable embedded in sand, and the temperature variations along the fiber optics at different water content were measured by the DTS. The relationship between soil water content and maximum temperature rise, cumulative temperature rise and thermal conductivity were established, and the measurement accuracies of soil water content derived from the above mentioned 3 estimated methods were compared. The results showed that the temperature fluctuation of the fiber optics decreased with the increase of the sampling spacing or the time interval, and the reasonable sampling spacing and time interval could control the temperature fluctuation within a range between -0.1 and 0.1 ℃. The temperature rise value of fiber optics decreased with the increase of soil water content. The temperature rise was the highest for the dry sand and was the smallest when the soil was saturated. The maximum temperature rise and cumulative temperature rise had a similar trends with the change of water content, and decreased exponentially with the increase of water content. The slope of curve decreased gradually with the increase of water content, and the sensitivity to water content gradually reduced. However, there was an increased exponential relationship between thermal conductivity and water content. With the increase of water content, the slope of curve did not decrease obviously. In the whole range of water content of sand, thermal conductivity had a good sensitivity to water content. For thermal conductivity method, at the all range of water content, the scatter points between measured and predicted values were on or near the 1:1 line, showing good predictions. For the maximum temperature rise and cumulative temperature rise methods, the scatter points were all around the 1:1 line when the water content range was in the 0-0.1 m3/m3, which had better prediction results, while the water content range was in the 0.1-0.25 m3/m3, where the scatter points were below the 1:1 line, which wound underestimate the moisture content. When the water content range was greater than 0.25 m3/m3, the scatter points were mostly above the 1:1 line, which overestimated the moisture content. The measurement accuracy of the thermal conductivity method was higher than that of the maximum temperature rise method and of the cumulative temperature rise method regardless of the low (0-0.1 m3/m3), medium (>0.1-0.2 m3/m3) and high (>0.2-0.35 m3/m3) water content ranges. The measurement accuracies of the 3 methods decreased with the increase of water content. The root mean square error of the thermal conductivity method was 0.015 m3/m3, which was lower than that of the maximum temperature rise method (0.038 m3/m3) and the cumulative temperature rise method (0.050 m3/m3). All of the 3 methods could measure soil water content accurately, but the accuracy of thermal conductivity method was the highest. However, although the maximum temperature rise and cumulative temperature rise methods could achieve certain accuracy, they had no physical meanings, the relationships between those 2 and water content were influenced by many factors, such as fiber optics characteristics and physical properties of soil. The thermal conductivity method had physical significance and was only related to the physical properties of soil. Moreover, the relationship between thermal conductivity and water content had been studied, and a lot of thermal conductivity models have been developed, which provided a simple and feasible method for estimating water content through soil thermal conductivity. Therefore, it was very attractive to measure water content by the active heating fiber optics-DTS using thermal conductivity method. This study provides guidance for water content measurement methods using DTS. It is of great significance to develop high-time-resolution in-situ monitoring techniques for soil water content at different spatial scales, and the ultimate goal is to accurately understand the water content dynamics in the field to guide the precision irrigation.