Stream depletion due to groundwater pumping has been recognised as a significant environmental concern by water resource managers for some time. Methods currently used in New Zealand to estimate stream depletion overestimate the actual stream depletion, as no allowance is made for streambed clogging or partial stream bed penetration of the aquifer.

A solution has been obtained by Hunt (1999) for streamflow depletion created by pumping from a well beside a stream. This solution differs from earlier stream depletion solutions by the inclusion of a semipervious clogging layer along the streambed and partial streambed penetration of the aquifer.

The aim of this research was to investigate the use of the solution by Hunt (1999) for analysing stream depletion field data. Field work was undertaken adjacent to the Doyleston Drain during August - September 1998. Tests undertaken consisted of aquifer step and constant discharge tests, streamflow measurements, aquifer and stream water level measurements, and adaptations from data obtained from previous soakage testing in the dry streambed.

Two methods have been presented to estimate aquifer and streambed parameters from the constant discharge test. Method I used drawdown data only. Method 2 further developed the results obtained from Method I and incorporated the measured stream depletion data in an iterative technique. It was found that Method 2 almost simulates the actual measured stream depletion (△Q ). Two additional independent methods to estimate the streambed leakage coefficient (λ) were investigated.

Theoretical stream depletion estimated by Theis (1941) and Hunt (1999) were compared with the actual stream depletion measured during the constant discharge test. Hunt (1999) Method 2 was found to more accurately estimate stream depletion than both Method 1 and Theis (1941). Variations in the estimation of △Q between Method 1 and Method 2 were a function of the value calculated for λ .

It was found that significant stream depletion can occur from streams that partially intercept confined or semi-confined aquifers. Hunt (1999) is applicable to any aquifer with a partially penetrating stream ( either naturally or artificially formed) that is not perched above the aquifer piezometric surface.

It is recommended that Hunt (1999) be introduced as a standard water resource management tool to better assess stream depletion effects from groundwater pumping.