Dear Dr. vittala,

You know very well, that aquifers are being over exploited and you can't stop that that. Hence what ever you do, it is not possible to manage. however, if you able to make water budgeting, and you find that some surface water is going waste during rainy season, you try to capture that. Apart from that, you have given only average rainfall. Certainly, there may be high rainfall years after every drought years (extreme events). If we able to capture that water and make arrangements for artificial recharge, it may help to a certain extent. 

Hi Vittala, Pl have a look at the attached paper to see approaches adopted in few other countries..

Hi Vittala. Knowledge of connectivity is essential. I encourage you to have a look at the attached publications, especially N. Guihéneuf et al.

Best

Laurent

Article Groundwater flows in weathered crystalline rocks: Impact of ...

Article Determining the vertical evolution of hydrodynamic parameter...

the basic tasks need as follows:

(1) apply the surface  or airbone geophysical exploration to know the fracture media aquifer profile.

(2) according to the step (1) results, set up the  GW level monitoring network and long term monitoring the level variation.

(3) according the hydrology data, such as precipitation, evaporation rate, estimate the region infiltration rate.

(4) integrate the monitoring data and predict by model the future variation included the man made and natural phenomena.

One key point I would suggest is change the irrigation pattern, if irrigation is done extensively, which ensures sustainability,

and methodology is described well above

The successful water resources management anywhere is crucially dependent on the knowledge of the available water resources and the ability to predict the future impact of  the management alternatives. Prediction of the the future impact is usually done by mathematical modeling. Without the full understanding of the water resources system, one can not develop, calibrate, and validate GWF model. Before going to management scenarios, you should have good monitoring network, good knowledge of hydrogeological system, fracture system, analysis of climatological system. I just notice that the amount of annual precipitation is relatively high and thus, rainfall harvesting techniques, suitable ASR projects are needed for increasing recharges. Surface geophysics (2D and 3D imaging earth resistivity tomography) are good techniques for 

delineating sturated facture.

MONITORING...   Monitor groundwater levels, monitor rainfall (per event), monitor pumping discharge. Once you have been monitoring for 3 or 4 years, the groundwater balance will begin to become clear. After this, your groundwater management decisions will be based on some hard facts. Keep monitoring thereafter and don't stop. 

In the interim, you can take some temporary management decisions based on such things as groundwater recharge estimates, anecdotal information about the movement of the water table etc.  

Crop yield sustainability is a different question entirely. There are water efficient methods such as drip irrigation, but these MAY have adverse effects such as soil salinisation especially in dry climates. 

Several good comments are already provided, Richard Owen in particular. What is the thickness and grain size distribution of the "sandy soils"? Is it a deposit or a regolith due tong term weathering? Is the 50 m depth to the water level you mention implying unsaturated conditions in the sandy soil or a perched aquiifer with another aquifer in the bedrock beneath? Iff you happen to have a flow logging device handy I would propose measurements during pumping in the differet wells to find out the location of the flowing fractures. Btw, does the bedrock outcrop in the area of interest or how far away is the recharge area?

While you keep on the crusial monitoring as Richard says you could set up a web page and share details. There are several of us willing to discuss the conceptual models that you need to interpret the monitoring data - one model for the regolith and one for the crystalline bedrock.

I suggest you to read this book: "Water Circulation in Roks" (L. Scesi and P. Gattinoni).

In this book you can find the problem of connectivity and the main flow direction.

It is extremely difficult to decipher the aquifer domain in hard rock  terrain. It is so complex that accuracy cannot be predicted by any means of technicality. Due to heterogeneity, aquifer parameter is highly variable in hard rock aquifer. The flow is draft dependent and there is possibility of reversal of gradient conditions against natural slopes and flow direction. Hence, management is terrain specific and site specific perhaps may differ from prediction.