Reply to comment by Zhi-Hua Wang and Elie Bou-Zeid on “Impact of wave phase difference between soil surface heat flux and soil surface temperature on soil surface energy balance closure”

Abstract

[1] We thank Wang and Bou-Zeid [2011] (hereinafter referred to as WB) for their comments about the phase difference between soil temperature and soil heat flux in our paper [Gao et al., 2010] (hereinafter referred to as GHL). In our previous work [Gao et al., 2008], we gave lots of derivations about the phase difference mentioned above. The main difference between WB's comments and our work (GHL) is the third term of their equation (11), which denotes the transient disturbance caused by starting the oscillation of surface temperature from the initial condition. In micrometeorological experimental research, this term is usually neglected because it is small. For soil surface heat flux, the transient term vanishes. [2] Our assurance (see GHL) that all of Rn, H and LE have phases identical to that of soil surface temperature may be more complicated for vegetation (such as dense grassland) surfaces because there are complicated physiological and physical processes [e.g., Sellers et al., 1996a, 1996b]. In the database used by WB, the sensible heat flux and latent heat fluxes partly come from grass canopy and partly come from ground soil surface, and the proportion depends on leaf area index, vegetation type and canopy height, etc. From the Web site WB provided, we failed to find how to decide the soil heat flux G0 which WB used in their comments. For grassland surface energy analysis, we should account for the heat storages in both grass and soil into soil surface heat flux. The same principle applies to soil surface temperature, and it should represent the canopy temperature and ground soil surface temperature [e.g., Sellers et al., 1996a, 1996b]. We suggest WB provide more information around soil surface heat flux and soil surface temperature determinations in their comments for further discussion. We found that WB's data of LE and H varied quite a lot (big changes from time to time, particularly around 1000, 1100, and 1200 EST) in their Figure 1b. Since Rn is quite smooth, G0 should have varied from time to time based on G0 = Rn-H-LE. The net radiation (Rn) shown in WB's Figure 1b was derived from four components of radiation budget: incoming and outgoing components of short- and long-wave radiation. I suggest WB insert a figure to show the diurnal variation courses of these four components to see whether the surface temperature measured by a thermal infrared radiometer has a phase identical to the outgoing long-wave radiation. Furthermore, there are three essential issues WB should address for their Figure 1b. (1) Rn suddenly decreased for a while at 1315 and 1415 (EST), and it was most probably caused by the intermittent cloudiness. However, both H and LE courses have a nick around 1200 (EST). As is well known, both H and LE should respond to change in Rn accordingly. (2) Both H and LE reached their peaks of 120 W m−2 and 160 W m−2, respectively, around 1115 (EST) when Rn and G0 were 260 W m−2 and 20 W m−2, respectively. It is obvious that Rn < H+LE under the condition that G0 > 0. This result may be caused by the different footprints of Rn, H and LE. (3) WB computed the eddy covariance fluxes over 5 min periods. Why select 5 min rather than 20 min or 30 min for average as is usual? [3] This study was supported by the National Program on Key Basic Research Project of China (973) under grant 2010CB428502, by National Natural Science Foundation of China under grants 40975009 and 40906023, by China Meteorological Administration under grant GYHY201006024, and by the Centurial Program sponsored by the Chinese Academy of Sciences. I am grateful to Heping Liu (from Department of Civil and Environmental Engineering, Washington State University) for his valuable comments.