Some questions from a follower:

Thanks for agreeing to help. I developed a simple mode just to check if it works. But I get this error (attached). There are NO streams in my study area so do I still need to give my model a input for drainage package? I wanted to go around it but I could not proceed, therefore I created streams as drainage lines from DEM.

Answer: You dont need to apply drains if you dont have drains. But, where water is going out of your system.

Secondly in my study area top surface geology will be heterogeneous and I have one confined aquifer underneath (2 m below the surface covering just a small part of the area not total area); and the task is to see if the a nearby road will influence this confined aquifer. Do you think a simple groundwater flow model can answer the question?

Answer: The geological setup is completely feasible with a multilayered groundwater flow model

How do I assign different geology on top model (first layer)? can I do that by importing a shape file and assigning different K values?

Answer: Please see this video: http://hatariwater.tumblr.com/post/143269466034/tutorial-to-create-a-modflow-grid-by-shapefile-and

I do not know, can we asign a time series for recharge values (in your video you give model an average value).

Answer: You can apply time series boundary conditions as recharge

I will be using 100 years forecasted data (incorporating climate change), what is the best way to run the model, using average monthly values or creating a time series for 100 years and running the model for 100 years?

Answer: Actually there is no limits for the time steps on MODFLOW. However is not practical to manage 1200 time steps.