An Adaptable DTS-based Parametric Method to Probe Near-surface Vertical Temperature Profiles at Millimeter Resolution

Abstract. High-resolution spatial temperature data is crucial to understand and describe near-surface atmospheric processes. Resolving temperature profiles at a sufficient resolution to capture near-surface temperature gradients is a challenge, especially in layers of short vegetation, such as grass. Grass has strong insulating properties, further promoting steep temperature gradients. To accurately quantify these gradients, sub-cm resolution temperature data is required. Distributed Temperature Sensing (DTS) allows for temperature measurement at a resolution of approximately 25 cm along a fiber optic cable. Such a cable can span several kilometers, resulting in many data points. Compacting this cable in a small space results in high spatial resolutions. This paper proposes a novel DTS-based temperature profiling method, which can probe temperature profiles at resolutions and accuracies up to the mm scale. This method revolves around a parametric script to generate files, which specify a cutout path for a laser cutter. The resulting parts are assembled into a coil-like structure, which holds a helically wound DTS fiber. Script input parameters can be changed to yield different, identically reproducible setup geometries. The performance of the method was tested and evaluated by fabricating a coil prototype and testing it in the lab and in the field at the CESAR atmospheric observatory in Cabauw, the Netherlands. This yielded high-quality vertical temperature data within the lowest meter above the soil, including a section of  10 cm tall grass. A vertical resolution and accuracy of 1.3 mm was attained and verified. The surveying period covers a broad range of weather conditions. The biases and artifacts that result from environmental parameters, such as solar radiation and precipitation were mapped and quantified. Despite the presence of minor biases and artifacts, the development of the method was concluded to be successful. As the method is modular and parametric, it can easily be applied in other research, potentially extending its application to other fields.