Comparative assessment of groundwater resource and recharge potential in barind area using modflow and mike she models

Abstract

MIKE SHE and MODFLOW are used to simulate the regional groundwater flow in the High Barind of northwest Bangladesh. The study area is the driest part of Bangladesh and has an upper thick and compacted clay layer of Pleistocene deposits. The low permeability of this clay layer limits its recharge to the aquifers. The geology of the area encloses two aquifers which are interconnected in many places. The aquifers are very thin locally. Heavy abstraction of groundwater is taking place for agricultural, industrial and drinking purposes. The study area faces problems like: almost no rain from November to April, limited scope for surface water storage, inadequate knowledge related to existing hydrogeology, etc. The models have been applied for the simulation of the groundwater flow for a period 1997 2005 over an area of about 2236 km2 . The area has been divided into 186 rows and 85 columns in layers of five distinct vertical deposits. Considering lithological variations and groundwater flow capacity, total 5 hydro-stratigraphic units have been demarcated within the studied depth in the study area (Clay Top, Upper Aquifer, Clay Middle, Lower Aquifer and Clay Bottom). At the outset MIKE SHE has been calibrated for 1997 to 2003 and validated for January 2004 to December 2005 against the observed groundwater levels at different locations. There is reasonable match between the computed and observed levels. Overland leakage coefficient is one of the major calibration parameters as unsaturated zone of the study area is thick. Soil properties like soil moisture tension/content relationship, unsaturated hydraulic conductivity and maximum bypass ratio of net rainfall are also important parameters for calibration. From the calibrated output, recharge values are extracted and used as input to the MODFLOW. During simulation of the MODFLOW the other input data and parameters are kept same to examine the results both from MIKE SHE and MODFLOW. The main difference between the two models is that MIKE SHE includes unsaturated zone, so it calculates infiltration, actual evapotranspiration and recharge from their physical laws. On the other hand MODFLOW deals with saturated zone only. From model result it is apparent that if recharge is calculated properly from other source, then MODFLOW will give good result. So in the case of groundwater flow study where irrigation is not present, MODFLOW can be used. Otherwise MIKE SHE will be more appropriate. Due to the occurrence of thin aquifers the study area shows a lower transmissivity locally. The study area being a ridge-shaped terrain, however, shows an outward flow of its groundwater all through the year to its surrounding low-relief areas. The recharge from rain starts in May and continues up to the end of October-November. It is found that for 1997-2005 the average recharge within the study area is about 443 rom, varying in the range of 308 rom in Sapahar Upazila to 607 rom in Chapai- Nawabganj Upazila. Simulation shows that recharge occurs mainly due to rainfall, while the contribution of irrigation (in the winter) is very negligible. Upazila-wise groundwater resources have been assessed based on safe-yield criteria, where groundwater table would be replenished every year. MIKE SHE result shows that the available groundwater resource (before irrigation starts) varies in the range of 180 rom in Tanore Upazila to 913 rom in Nawabganj Sadar Upazila. Usable resource has been assumed as 90% of the available resource as there are some natural losses set-out the study area during irrigation season. For the study area, potential recharge has been estimated considering water balance of the period from May 2001 to October 2001 (as water table attains the lowest and highest elevation in May and October, respectively). The potential recharge of the present study varies from 395 rom in Porsha Upazila to 621 mm in Nawabganj Upazila For every year and every upazila, plots on cumulative recharge against cumulative rainfall have been drawn and found that the plots are following some exponential relations. With the increment of rainfall, recharge increases gradually in the initial stage and comparatively faster in the later stage. Some upazilas are withdrawing more water than the maximum capacity of recharge (i.e., potential recharge). In many cases water table goes beyond suction limit, which leads shallow tube wells to inoperable condition. The adverse effect of over exploitation of groundwater may be overcome by placing shallow tube wells with appropriate spacing. There is a possibility of fault location near Tanore Upazila, which should be investigated further.