Some further considerations of the measurement and indirect evaluation of natural evaporation

A previous observational study (Pasquill 1949 a) has been supplemented by further measurements of evaporation and the associated vertical profiles of wind velocity and vapour pressure in the first two metres above a clayland pasture with grass of moderate length, mainly in effectively adiabatic conditions of flow. The evaporation measurements, using a simple form of soil evaporimeter, included two series of statistically designed measurements, in order further to examine the “reliability” of the technique. The data so obtained, though not comprehensive, indicate no substantial error due to the necessary isolation of test soil from underlying layers, but there is a suggestion that the rates of evaporation ultimately adopted might be underestimates, owing to unavoidable disturbance of root and soil continuity in the vertical at the perimeters of the soil cores contained in the evaporimeters. The observed rates of evaporation (E o ) are compared with those computed (E c ) from the following formula, which is based on the Prandtl form of the law relating drag and wind profile over a rough surface in adiabatic conditions, E c = ρ k 2 ( u 2 – u 1 ) ( q 1 – q 2 )/[log ( z 2 – d/z 1 – d )] 2 where u 1 , u 2 , q 1 , q 2 are the wind velocities and specific humidities at heights z 1 , z 1 above the soil surface, d is a‘zero point displacement’ associated with the vegetation cover, ρ is air density and k is von Kàrmàn's constant, of magnitude 0.4. For the present measurements the usual “roughness parameter,” deduced from the wind profiles, was 0.51 cm as compared with 0.25 cm for the previous series. The two series include 19 separate determinations of rate of evaporation over periods of about one hour, and in terms of different sections of the observed profiles, provide 41 values of E c /E o in effectively adiabatic conditions, with wind velocities at a height of 2 m ranging from 3 to 9 m/sec. The mean values of E c /E o is 0.99 and the individual deviations from 10 are within 10, 20 and 30 per cent in 24, 66 and 90 per cent of cases. It is demonstrated that deviations of this order are to be expected from local variations in the various factors involved. Although some a priori uncertainty remains concerning the precise validity of the evaporation measurements, the consistency so demonstrated in terms of a semi‐theoretical treatment which has been established in other connections (Calder 1949) suggests that the ultimate result may be accepted with confidence. Brief consideration is then given to the validity and practicability of applying the above formula to evaporation from areas of growing crops.