Saline soils in Pakistan are traditionally reclaimed by methods of continuous ponding of the soil. This method consumes large quantities of good quality water that otherwise may be used for irrigation of crops. Due to water scarcity there is an urgent need for efficient leaching methods. In this thesis laboratory experiments and theoretical modelling are reported that investigate water use efficient methods of leaching saline clay soils that would be applicable to the soils in Pakistan.

When the soil is aggregated, as is usually the case because of cultivation, micropores within aggregates act as sources or sinks for salts while the flow and mixing of salts takes place in the macropores around aggregates. Salts in the micropores within soil aggregates are transported to the macropores by diffusion, while they are transported by convection with the moving water in the soil macropores. The leaching efficiency of different desalinisation methods has been investigated by considering the diffusion of salts out of the aggregates that are subject to different boundary conditions in the surrounding macropores. The different methods considered were: (a) continuous leaching, (b) intermittent leaching with rest periods that provide time for salt diffusion and (c) a new method called the start-stop method in which the macropores are intermittently filled and drained without continuous flow through the macropores.

Because the leaching of salts from aggregated clay profiles is diffusion limited, experiments were first conducted to obtain the effective diffusion coefficient of solutes in baked clay, formed into spherical aggregates of various sizes. Analytical and numerical solutions of the diffusion equation were developed to determine the amount of solute leaving a single aggregate under conditions envisaged for the various leaching methods. These results indicated that the start-stop method used less water to reduce the solute content of the aggregate than either the continuous leaching or the intermittent leaching methods. This was confirmed in leaching experiments on columns of aggregates. These results showed that there was no advantage of using intermittent leaching over continuous leaching for removing 60% of the salts from columns if the macropores were initially saturated. However, water savings up to 12% were possible by using intermittent leaching when 80% of the salts were leached. Under initially unsaturated macropore conditions water savings up to 65% were possible with intermittent leaching for leaching 60% of the salts. Under our experimental conditions with baked clay aggregates, 12% more solute was leached with the same amount of water under saturated intermittent leaching than with drained intermittent leaching. The Burn’s leaching model was modified to take account of the bimodal pore distribution and diffusion. The model was able to simulate the leaching behaviour in the experimental macroporous soil columns.

Since solute transfer was more rapid when aggregates were initially unsaturated, methods are needed to de-saturate the soil before the desalinisation process. In the field, unsaturated soil conditions are obtained not only by surface evaporation but also by growing some plants. A pot experiment showed that plants can uniformly desaturate the soil in the root zone. Leaching then required only 37% of the water needed when there were no plants, suggesting that this would be a possible technique to apply to the soils in Pakistan and other parts of the world, depending on climate, available water, drainage and topographic conditions.

Of the methods investigated in this work the start-stop method was found to be the most efficient method with water savings up to 90%.