Measurement of soil‐surface heat flux with a multi‐needle heat‐pulse probe

Summary Soil‐surface heat flux ( G 0 ), an important component of the surface energy balance, is often determined by summing soil heat flux ( G z ) at a depth ( z ) below the surface and the rate of change in soil heat storage (Δ S ) in the layer above z . The soil heat flux G z is commonly measured with passive heat flux plates, but self‐calibrating plates or additional corrections are required to obtain accurate data. In some cases, Δ S is neglected because of the difficulty of monitoring the dynamics of volumetric heat capacity ( C ), which might lead to erroneous estimates of G 0 . To overcome these limitations, we introduce the heat‐pulse method for measuring G 0 with a multi‐needle heat‐pulse probe (HPP). Soil temperature ( T ) distribution, thermal conductivity (λ) and C of the 0–52‐mm layer were measured hourly on five consecutive days with an 11‐needle HPP, and G z at 50‐mm depth ( G 50 ) and Δ S of the 0–50‐mm layer (Δ S 0–50 ) were determined by the gradient and calorimetric methods, respectively. Independent measurements of G 50 with a self‐calibrating heat flux plate and Δ S 0–50 calculated with the de Vries model C were used to evaluate the HPP data. With reliable G 50 and Δ S 0–50 measurements, the HPP‐based G 0 data agreed well with those estimated from the independent method (with a mean absolute difference of 4.5 W m −2 ). Supporting measurements showed that determining G z at the 50‐mm depth minimized the likelihood of errors from evaporation below the measurement depth. The multi‐needle HPP provides a reliable way to determine G 0 in situ . Additional analysis demonstrated that by reducing the number of needles from 11 to 5, the datalogging requirement was reduced by half and G 0 was still determined with acceptable accuracy. Highlights Heat flux at the soil surface ( G 0 ) was monitored by the heat‐pulse technique. A multi‐needle heat‐pulse probe (HPP) was used to measure subsurface soil heat flux ( G z ) and heat storage concurrently. Appropriate measurement depths of G z were determined to minimize the effects of subsurface latent heat sink on G 0 . A simplified calculation reduced the datalogging requirement of the multi‐needle HPP.