Sensitivity of the surface layer nocturnal energy balance closure to soil heat flux estimation

The surface energy budget shows how the available net radiation (Q*) is split among sensible heat (H), latent heat (LE) and soil heat flux (G). It is well known that field observations of the surface energy balance often are not able to close the budget, as usually (LE + H + G < Q*). This problem tends to be enhanced at nighttime, and this is often explained by the fact that in stable conditions a proper determination of the heat fluxes by the eddy covariance technique may not be feasible.

In observational studies, the different components of the soil energy budget are not measured at the same level, the heat storage within the soil layer between the observations needs to be added to the balance. As the soil heat flux sensor needs to be installed a few centimeters below the air-soil interface surface, one needs to have a good knowledge of the temporal evolution of soil temperature in the layer that goes from the surface to the heat soil sensor. However, it is also the layer that experiences the largest temperature variations, and vertical soil temperature gradients can be appreciably large near just below the surface.

In the present study, we look at soil heat storage, and investigate how much of the missed energy budget at nighttime may be attributed to the lack of exact knowledge of this component. An estimate of this term's relative importance and its variability range is given. The modified surface energy budget is assumed to be (H+LE+G+S=Q*), where the storage term is estimated through a simple relation C(DT/Dt)DZ, C is the Soil Heat Capacity, (DT/Dt) is time variation of temperature and DZ is the depth of the soil layer.

The data used in this study was obtained from four Soil Temperature Sensors, two Heating Flux Sensors, one Eddy Covariance System and one Net Radiation Sensor. The experimental site is one of the LBA's sites that is located in eastern Amazon region(3.0120S, 54.5370W), Santarém, Pará, Brazil over a deforested land (Pasture).