Hi,

I think this depends to a major extent on the amount and quality of your data in to the model. Sparse and low-quality input data will give similar output. Dense and high-quality input data will give high-quality output. In general, I think Modflow is one of the more complete models in its category.

best

rob

I suggest you as a first attempt to use a crop model at the scale of the uppermost few meter to compute the return of irrigation water to the aquifer. Hydrus1D is able to do that. This kind of model will require sub-metric resolution in the upper soil layers and a time step less than 1 day. 

This will allow you to force a Modflow model for groundwater with the results of the crop model.

Best Regards, 

Pierre

Because of monthly temporal constraints, MODFLOW is unable to capture daily changes in well yields and borehole-scale water table drawdown. I read about Simulating Pumping Boreholes with a Darcy-Forchheimer Regional-Radial Flow Model - SPIDERR) and it seems more viable than MODFLOW

Looks like you need to interface with a surface model to accurately model the return flow nd drawdowns from withdrawal. Have you tried couple models like https://water.usgs.gov/ogw/gsflow/ or SWAT-MODFLOW at http://swat.tamu.edu/software/swat-modflow/

I agree with Dr Randhir regarding SWAT-MODFLOW. What the situation at hand requires is integration of subsurface hydrology with what happens at the surface - to water and land use.

However, you might also want to take a look at SHETRAN. It's a surface-subsurface integrated model. I had worked in a paper that involves surface water cum ground water modeling in a rapidly urbanizing basin with high groundwater abstraction issue. However you will find that apart from other applications, it's been widely used as a robust contaminant transport model as well.

SHETRAN webpage (software and related publications ): 

http://research.ncl.ac.uk/shetran/

Above mentioned paper:

http://www.tandfonline.com/doi/full/10.1080/19475705.2017.1289985

You may try to combine the modflow and swat model. the latter is strong in handling crops and crop water use. somebody coupled modflow and swat. however, it may not easy to set up and run the coupled model. you may run them offline with the output of swat as input of the modflow. this is much easier to do. good luck. 

Dear Ammara Talib, you have a good question for which perhaps you already have the correct groundwater flow functioning of your study area. If that is the case, possibly you already have enough elements to decide that there is no vertical flow from beneath your study land. Maybe, a little bit of chemical questions to the groundwater flow involved as were it comes from, might be of some use to your investigation. I am taking the liberty of including a publication on how we found the importance of the different components of groundwater flow. This importance is to be manifested in the groundwater flow model. Please do have from Mexico saludos,jjoel carrillor  

Thank you all for valuable information. SWAT-MODFLOW seems like a good option. the only limitation I can see for now is that it is coupled one way or not fully integrated. Since in my study area, fluctuation of surface water is effected by groundwater, model with  two way coupling will be more suitable. 

Attached article is modification of SWAT to SWAT-LU with two way coupling.

I believe modflow would be your best bet, and probably focusing on the appropriate parameters to simulate.