Adaptive modelling of water-logged groundwater regime

Author : Patmajum

Solution to the complexity of conjunctive use planning using mathematical models does play around the source/sink term. This temperament perpetuates the concept of water logging in the command areas resulting from excessive supply of irrigation water. This thinking has provoked the model users to incline towards approaching the water balance input in the model and suggesting consumptive use at many times as remedial measures out of immaturity. Elsewhere even with sprinkler techniques, water logging and salinity problems have been noticed.

Let us analyze the causes for water logging and salinity first and then apply it to a conceptual model to avoid notices of uncertainties. This manuscript deals with Dharoi Command Area in Sabarmati River Basin using MODFLOW.

Introduction

In many of the irrigation commands, sustainability of agriculture production goes down all of a sudden due to degradation of land because of salinization, alkanization, water-logging and soil acidification. In the long-term, one of the most frequent problems of irrigation schemes is the rise in the local water-table termed as water-logging. A high water table also makes the soil difficult to work. Good irrigation management, closely matching irrigation demands and supply are sighted as control measures to reduce seepage and increase irrigation efficiency, thereby reducing the groundwater recharge. Apart from measures to improve water management, two options to reduce seepage are to line canals in highly permeable areas and to design the irrigation infrastructure to reduce wastage. Low irrigation efficiencies, as low as 20 to 30% in some areas, are cited as one of the main causes of rise of water table. Shirahatti (2005) found that increased area under irrigation and cropping intensity especially with paddy cultivation, the rise in water table was significant after 1996 in the Stage-I, Phase-1 area of the UKP command. Poor water distribution systems, poor main system management and archaic in-field irrigation practices are described as the main reason.

The International Commission of Irrigation and Drainage has recommended increasing field application efficiency to even 50%, to significantly reduce the rise in the groundwater. Dhindwal and Kumar (2005) tested the efficient water management practices on the selected distributaries of the JLN command in southern Haryana and assessed their impact on water productivity. The groundwater level rise can be spectacularly fast in flat areas where the water table has a low hydraulic gradient. The critical water table depth is between 1.5 and 2 m depending on soil characteristics, the potential evapo-transpiration rate and the root depth of the vegetation/crops. Modi et al. (2007) observed that the area falling in command of Kosamba and Hansot Brach canal is nearly level to very gently sloping and some places have concave topography. Thus natural drainage cannot be considered adequate to take care of excess water of irrigation. This drainage condition is responsible for water logging and salinity in the command.

Theoretical aspects of groundwater modelling are included in Bear and Verrujit (1987), Franke et al. (1991) and Anderson and Woessner (1992). Numerical model application in command areas has been practiced since last four decades, initially dealing with quantity aspects and later on extended to quality. Extensive literature is available dealing with numerical modelling of groundwater flow in command areas (Kashyap, 1989). Majumdar (2001) reported the conceptual development of groundwater recharge and draft in Upper Palar zone in Palar River basin in Tamilnadu and found that recharge has relation with groundwater contamination. Majumdar et al. (2002) conceptualised the flow domain of Nadia district of West Bengal to disagree with the existing concept of water balance estimates and thereby role of recharge in arsenic pollution. In fact they found dynamics of draft in majority causing the problem. Payne et al. (2003) reported many other MODFLOW applications. Satluj-Beas command in central Punjab has been modelled by Majumdar et al. (2005) to conclude that irrigation return flow does influence the river water levels more than the groundwater table. Such unusual observations were also reported in Kenedy and Brodhoeft (2006) in Galletin valley in Montana. Ahmed and Umar (2009) carried out groundwater flow modelling in a part of central Ganga Plain in Uttar Pradesh. Majumdar (2010) applied MODFLOW to simulate irrigation return flow in a coastal command of Andhra Pradesh.

Modelling of waterlogged command areas has been carried out in past to project the impact of conjunctive use planning of surface and subsurface water. Srivastava (1999) in a modelling study of Sharda Sahyak Canal command area has concluded that regulating groundwater draft and roistering of canals in different crop seasons can provide a better solution to arrest water-logging. Improvements in cropping and water use pattern can provide better results in arresting of water-logging than mere lining of canals. Goel et al. (2005) adopted a GIS based model for integrated water resources Management in an irrigation system. Based on the irrigation demands, canal water availability and groundwater conditions, the weekly model optimally allocate the surface water and groundwater. Rajput and Patel (2005) concluded that CROPWAT is a powerful tool to simulate different crop water need scenarios under different planting dates and thus enables the user to select most optimal sowing date to realize higher yields and water use efficiencies by matching the probable canal water supplies with crop needs. A calibrated and validated three layered MODFLOW based groundwater model has been developed by Goyal and Arora (2007) for the area under Indira Gandhi Nahar Project (IGNP), which has subsequently been used for prediction of future water budget and groundwater levels showing waterlogged area in different categories for different groundwater management strategies. Jain and Sondhi (1998) studied Water balance of Bist-Doab tract of Punjab state including waterlogged area of Mukhtasar district in Punjab and suggested for conjunctive use of surface and groundwater. All these studies and many other not cited here, have one concept in common: excessive irrigation causing the problem, which is very often not true. Therefore classical advice based upon modelling studies rarely worked on the field, resulting in to blame about ineffective capacity of numerical modelling techniques.

Consumptive use of water in the farm lands can not be denied, however sole reason for water-logging may not be excessive use of water and canal seepage. Groundwater rising under capillary action generally evaporates, leaving salts in the soil. The problem is of particular concern in arid and semi-arid areas with major salinity problems. Hydrus-2D was calibrated to simulate the water distribution pattern for different type of soil under varying emitter discharge, by Ajdary et al. (2005). They indicated that irrigation interval of 48 hr is appropriate for sandy clay loam soil. Singh (2005) presented study, aimed at diagnosing the impact of different irrigation management scenarios on crop performance in the absence of sub-surface drainage system through validated simulation model SWAP, ultimately suggesting the irrigation schedule and necessity for installation/non-installation of subsurface drainage system for sustainable crop production. These are altogether different concept, where soil properties are in the core of thinking. In the background of such convincing findings it is imperative that MODFLOW application in waterlogged areas be adaptive. Present paper explains it using the base data of Dharoi Command area in Mahsana district of Gujarat, India.

Concept Development for Water-logging

In respect of groundwater flow, water-logging may be seen as a condition of degraded horizontal flows with time. The soil approaching fully saturated condition from an initial unsaturated stage. Once voids in the soil are completely filled with water, it is similar to the concept of clogging around a recharge well conceptualised using cake thickness in filtration theory (Majumdar et al., 2007). The parameters which could influence the groundwater flow across the cake developed in MODFLOW model are: hydraulic conductivity (k) and width of the cake (l), initial water table condition and location of observation. Figure 1 to 4 shows effect of these parameters on head build up in the location of water application surrounded by the cake and also at a point on the other side. Figure 1 indicates increase in head buildup with cake formation, with width of the cake as major influencing factor. This is synonymous to water logging condition in a discrete point. Head buildups in the location of water application for the horizontal and inclined initial water table condition are identical. However it is different in any other location in the outer side of the circular cake. Figure 2 depicts such head build-ups at a radial distance all around in case of horizontal initial water table, where as the same for inclined initial condition are shown in figure 3 and 4 on the downstream and upstream of the point of application respectively. Overall impact of cake width is most significant in association with its reducing hydraulic conductivity with time. These discrete locations, when widely spread in a model area, give continuous rise to the existing water table condition irrespective of quantity of water application. However, rise in water table can certainly be controlled by limiting the water application in proportion to the current hydraulic conductivity. Hence objective of the present modelling study is to develop a relationship between applied water and hydraulic conductivity. Input of non-consumptive nutrients though remain attached with the reason behind gradual reduction of hydraulic conductivity and the relationship between the two could be useful in establishing water logging and remediation, same is not included in the present paper considering that could be readily available with soil investigation agencies. Elsewhere water-logging conditions are also seen, solely dominated by the local topography and anisotropy. Conjunctive use planning has very limited scope in such situations.