SURFACE RUNOFF MODELLING OF AN ALKALI WATERSHED UNDER RECLAMATION

Abstract

In India and also in many parts of the world, vast tracts of waste-lands are due to alkalinity of the soils. An alkali soil contains sufficient exchange able sodium, which causes the soil to disperse and to increase in the pH. The excess sodium causes the for mation of hard crusts at its surface-layer, and also adversely affects its physical and nutritional proper ties. In order to produce more food to meet the needs of increasing population, it becomes obligatory to reclaim these barren alkali lands. Reclamation of alkali soil requires the removal of excessive exchangeable sodium and its replacement by calcium, followed by the leaching of the soil. This is popularly accomplished by the application of chemical compounds such as gypsum, pyrites, etc., followed by leaching. . The hard crust formed at the top surface of the alkali soils reduces the infiltration considerably, and thus most of the incipient rainfall is transformed into runoff. As the alkali soil gets reclaimed, its hydrologic characteristics change considerably. The present v^rk is an extension of the research conducted in the Central Soil Salinity Research Institute (CSSRI), Karnal (India). The data used pertain to the iv Kachwa watershed, which underwent reclamation from 1976 to 1979. Detailed surface hydrologic investigations have been carried out for the runoff computations of this watershed. To ascertain the effect of reclamation on the surface runoff, the Instantaneous Unit Hydrographs and subsequently short duration unit hydrographs have been computed for different degrees of reclamation. The twoparameter conceptual model consisting of a cascade of linear reservoirs (Nash, 1957) is used for the purpose. The effect of reclamation on the model parameters, nK and K, is established. Also, the influence of the reclamation on the unit hydrograph parameters namely, the peak ordinate, the time lag, and time base, is estab lished. A coaxial graph illustrating these influences is prepared for the Kachwa watershed. The effect of reclamation on the specific peak discharge has also been worked out. The unit hydrograph studies thus indicate that the surface storages do change as reclamation proceeds. Attempts have been made to carry out a detailed study of the changing surface storage characteristics of the watershed sys+sm through conceptual configurations of linear reservoirs. The Kachwa watershed (though small in area) is treated as the pilot test watershed for the surface hydrologic studies that may be applied on medium to large alkali watersheds undergoing reclamation. For a small watershed, the entire system is considered to be lumped and time-invariant. To start with, the Single Linear Reservoir (SLR) Model has been tried for simul ating the hydrologic responses. For Kachwa watershed, the SLR Model did not give satisfactory results. Such a model did not take into account the surface storage function adequately. However, for this purpose an additional routing element of a linear reservoir, attached in series to the SLR, has been found to be of great use. Therefore, a simple model of two linear reservoirs in series has adequately represented the alkali watershed under recla mation. The additional linear reservoir has a damping effect on the response of the SLR Model. The effects of reclamation on the storage coefficients of the two linear reservoirs have been studied in detail. The proposed configuration of the cascade of two linear reservoirs in series is essentially a lumped system. A careful examination has revealed that the sur face storages behaved differently in unreclaimed and reclaimed parts of t\e watershed. As a first attempt for a medium watershed, a distributed system has been consi dered in which two linear reservoirs (one representing vi the unreclaimed part of the watershed, whereas the other, the reclaimed part) are connected to acommon routing element of another linear reservoir. The exact differ ential equations for the system have been worked out and the results obtained have been found to be encouraging. Further, for large alkali watershed undergoing reclamation, whose surface storages may not be taken care of adequately with the proposed distributed system of three linear reservoirs, amodified version, comprising of two strings, each having two linear reservoirs in series, has been conceived of. The reclaimed areas of the watershed are conceptually represented by one string, whereas the unreclaimed alkali lands are taken care of by the other string. Both of these strings are connected to arouting element, which is also a linear reservoir. The proposed model has been tested with the available storm events for different degrees of reclamation. A thorough scrutiny of the reclamation process reveals that the reclamation proceeds from plot to plot. Plots so identified are bunded, treated,and leached. Subsequently, paddy crops are sown to promote the process of leaching. Water stood in the paddy fields and the transfer of water (a.footing the surface storages) very much depended on the agronomic practices followed. These VI1 phenomena can be studied through distributed parameter approaches. For this purpose, the time-area-concentrat ion based approach has been tried. The genetic principle of runoff has been correlated with the convolution process to distribute the storage effects over the entire watershed. An isochronal pattern suited for reclamation process is determined. The runoff factors for different parts of the watershed and for various storm events are computed. The proposed model can give idea about the involvement of different parts of the watershed in the runoff process. The proposed lumped and distributed parameter models developed to account for the surface hydrologic responses have been tested for the available storm data of the Kachwa watershed. The results so obtained have been critically examined, and logical conclusions have been drawn.