Towards uniformity in the near-surface temperature profile

Accurate estimations of the surface temperature are crucial to determine the exchange of energy and moisture between the surface and atmosphere. Temperature profiles in the atmospheric surface layer (ASL) are generally represented by the Monin-Obukov Similarity Theory (MOST). The logarithmic behavior in the lowest part of the ASL requires the introduction of the thermal roughness length. However, existing models for the thermal roughness length, such as the ratio to the roughness length for momentum, are not well-defined and may be based on incorrect physical assumptions. Therefore, it remains a challenge to predict the surface temperature from standardized temperature measurements at higher levels.

A reason for this uncertainty is the high vertical measurement resolution that would be required to validate the models near the surface. In this study, we designed a Distributed Temperature Sensing (DTS) structure that measures the air temperature vertically up to 1 m above the surface with a resolution of 2 cm and from 1-8 m with a 30 cm resolution. We placed this structure at a grassland site in the Netherlands in May 2022. We saw how the conventional MOST model accurately describes the logarithmic layer. However, we noticed a strong misalignment with the observed temperatures in the lowest 1 to 2 m.

We have introduced an alternative length scale into the existing model concept, omitting the need for the thermal roughness length. We show how one month of temperature profile observations under stable conditions takes one universal form by normalizing the temperature observations with θ∗ and normalizing height by the new length scale. We can describe this form, from directly above the vegetation up to and including the logarithmic layer, using a fully physically based model. Comparisons are also made against alternative models, including a tall canopy model and an adapted MOST model.