CHARACTERIZATION OF A PERCHED GROUND WATER SYSTEM IN IRRIGATED SUGAR CANE USING ELECTRICAL RESISTIVITY PROFILING, PUUNENE, HAWAII

Between 2010 and 2012 we took part in a study to assess the practicality of converting the last large-scale sugar cane plantation on Maui Island, Hawaii to biofuels production. The plantation currently has 14,380 hectares of sugar cane under drip irrigation. The irrigation system supplies up to 900,000 m of water per day through a system of 177 km of concrete-line canals and 43, mostly unlined reservoirs, used for temporary storage. Our part of the study was to assess the water loss rate from the irrigation system. Because the unlined reservoirs represent the largest potential loss of water, we started by deploying seepage meters in six representative reservoirs. We found net gain, rather than loss in all reservoirs tested. To explain this surprising result, we collected a toposequence of 2D resistivity profiles from above irrigated cane in the upper elevations of the plantation, down slope across a canal, and then across a reservoir. These profiles show high resistivity (5000 Ohm-m) at the ground surface, up-slope of irrigated cane. Starting at the up-slope-most irrigated cane, there is a down-slope-thickening, lowresistivity (10-200 Ohm-m) layer extending from the surface to depths of 0 to 10 m. At the canal crossing, this layer thickens to 15 m and is less resistive upslope of the canal, but thins and becomes more resistivity down-slope of the canal. Beneath a reservoir, the layer thickens to 20 m, arches beneath the reservoir dam and up to the base of the field beyond. We interpret this low-resistivity layer as a wet zone, fed by irrigation water that has bypassed the cane root zone to form a perchedgroundwater system. Because the canal is concrete lined and extends 3 to 5 m beneath the surface grade, it acts as a barrier to perched water flowing down-slope from above. The reservoir bottom was also excavated below grade. Hence, perched water flowing down slope enters the reservoir along its upslope side. It then flows down and underneath the dam through a narrow belt at the deepest part of the reservoir, where the sealing material forming the base of the perched system was either absent or removed during construction of the reservoir. We conclude that this previously unrecognized perched water system explains the observed widespread lack of leakage from the most of reservoirs and suggests that the bulk of the leakage is occurring in the down-slope parts of the reservoirs and the dam faces. Significant leakage from the canals may be limited to areas of high crack density in the concrete linings. The existence of this perched water system has important implications for the operation of the plantation in an era of increased demands on the fresh water resources of Maui.