Meteorological and soil factors affecting evaporation from fallow soil

Using the conception of natural periods, for which the difference between rainfall and drainage can be equated to the evaporation, the mean daily rates of evaporation from a block of fallow soil at Rothamsted are examined for 94 periods varying in length from 13 to 45 days. In seeking correlations with single daily meteorological observations two types of treatment are employed. (1) The year is considered in three seasons of four months each—summer, winter and two equinoctial pairs of months—and it is shown that an almost complete description of evaporation can be obtained in terms of rainfall only, the nature of the correlation varying from season to season. (2) A general treatment is attempted in physical terms, considering evaporation as due to diffusion across a non‐turbulent boundary layer whose thickness is determined by wind velocity, the soil surface being assumed to be continuously at 100 per cent R.H. The general agreement between observed and predicted values is very good in winter. The summer data are shown to lie between the theoretical limits imposed by the assumptions of (i) continuous 100 per cent R.H. at the surface, and (ii) a steady retreat of the 100 per cent R.H. layer into the soil, i.e., no upward movement of liquid during the evaporation process. The considerable scatter in the data is attributed partly to the inadequacy of single daily meteorological observations but chiefly to the lack of knowledge of the conditions existing at the soil surface.