SUMMARY Urea application to soil raises the pH and ammonium concentration, thus providing ideal conditions for ammonia volatilization. A mechanistic model is presented, which combines the process of ammonia volatilization with the simultaneous transformation and movement of urea and its products in soil, for predicting the concentration profiles of urea, ammoniacal‐nitrogen and soil pH, and ammonia losses, following application of urea. The model consists of continuity equations describing the diffusion and reaction of urea, ammoniacal‐nitrogen and soil base; it takes into account the volatilization of ammonia and the concurrent acidification of the soil surface; and considers a variable PCo2 profile due to soil respiration and urea hydrolysis. The derivation of the continuity equations and their boundary conditions, calculations of ammonia volatilization, and appropriate methods for numerical solutions are described.

SUMMARY A sensitivity analysis of the model is given for conditions in which there is no plant uptake of N and over the range of conditions likely in tropical lowland ricefields, and the mechanisms governing inorganic carbon, nitrogen and acidity dynamics in the floodwater and soil are discussed. Changes in the rate of CO2 production in the soil, floodwater algal activity and floodwater carbonic anhydrase activity altered diurnal changes in floodwater pH but little influenced the daily mean pH, and consequently little influenced cumulative N loss by NH, volatilization. Soil pH buffer power and rate of urea application also little influenced cumulative percentage N loss. Increases in soil ammonium buffer power, floodwater depth and downward water flux, and decreases in floodwater‐ and air‐turbulence, moderately decreased N loss. N loss greatly decreased with decrease in urea hydrolysis rate and with urea incorporation in the soil. Incorporation decreased N loss more in the absence of standing floodwater. With low urease activity, deep floodwater gave consistently lower N loss; with high urease activity, shallow floodwater gave lower N loss when urea was incorporated.

A sensitivity analysis of the model described in Part I showed that the proportion of N lost as ammonia from surface applied urea is very sensitive to the initial pH of the soil, its pH buffer capacity, the rate of urea application, and the soil urease activity. Under the conditions tested, the diffusion of bicarbonate ion to the soil surface, to neutralize the acid generated when NH4+ is volatilized as NH,, appeared to be the main process controlling the rate of ammonia volatilization.The amount of ammonia volatilized was not very sensitive to the value of the transfer coefficient between the soil surface and the atmosphere, nor to the soil moisture status if this was around field capacity. Adsorption of ammoniacal-nitrogen was less important than the soil pH buffer capacity in influencing the ammonia volatilization.Further applications and extensions of the model are discussed.