Direct Measurement of Flood Water Percolation in Arid Lands

Project Number:

03R-21

Project Duration:

29 months

Dates:

November 18, 2003 to April 20, 2006

Institution of Principle Investigator while on this project:

Ben-Gurion University of the Negev

Investigators (most current known information)

Dahan, Ofer (PI)

Lecturer, Department of Environmental Hydrology & Microbiology, Ben-Gurion University of the Negev, The Jacob Blaustein Institute for Desert Research, Midreshet Ben Gurion 84990, ISRAEL

EL: +972-8-659-6917, FAX: +972-8-659-6909, Email: odahan@bgumail.bgu.ac.il

McDonald, Eric

Earth & Ecosystem Sciences, Desert Research Institute, 2215 Raggio Parkway, Reno NV 89512

TEL: +1-775-673-7302, FAX: +1-775-673-7485, Email: emcdonal@dri.edu

Yechieli, Yosef

Israel Geological Survey, Malchey Israel 30, Jerusalem 95501, ISRAEL

TEL: +972-2-531-4326, Email: yechieli@mail.gsi.gov.il

Proposal Abstract

Summary and Conclusions

The research studies the recharge processes of shallow alluvial aquifer as a result of flood events in the Arava Valley, Israel. In this extreme arid region two different conditions for infiltration were examined. One was infiltration under natural ephemeral stream and second was under percolation reservoir. These kinds of reservoirs were built in the Arava Valley in the last thirty years to prevent water flow to the Dead Sea and to increase the potential of water capacity in the alluvial aquifer by enhanced percolating through the reservoirs.

In order to estimate recharge rates as a result of flood events under natural ephemeral stream and percolation reservoir four monitoring stations were built. Two stations (NAM1 and TZ3) were built in wadi Arava as a natural ephemeral stream and two others (TZ1 and TZ4) were built in Tzukim reservoir to represent a percolation reservoir. However, only two of them, NAM1 and TZ1 stations successfully monitored flood events during this study period. Each of the monitoring stations was equipped with real-time monitoring devices to measure parameters at surface, vadose zone and groundwater. The measured parameters were water level, electrical conductivity, water temperature and water content. Recharge rates were calculated using various methods to allow a broad accurate assessment of the recharge process. The methods which were used to estimate recharge rates were based on surface water loss, ring infiltrometers and water front propagation through the vadose zone and groundwater response to flood event. At NAM1 station two events were monitored, a controlled percolation experiment and a natural flood event. Recharge rate estimations under controlled conditions were calculated as 10 cm hr-1 in average and under natural small flood event 3.2 cm hr^-1. Although measurements were taken by the same system at the same monitoring station the differences in flux rates between the controlled percolation experiment and the natural flood event are mainly attributed to the flooding conditions. The water level in the controlled percolation experiment was ~2.5 times higher than that of the natural flood event and so was the measured flux. At TZ1 two flood events were monitored, one during January 2004 and second during October 2004. The reservoir was chosen to host the station mainly because its bottom was cleaned from deposits of fine material which accumulate to 4 meters along 10 years. This fact allowed the evaluation of the development of infiltration capacity from a refreshed reservoir. First 4 hours from flooding water percolate in rate of about 10 cmhr-1. Then in a two days period the flux rates were decreased to ~1.5 cm hr^-1. After the first two days no infiltration was observed and the rest of the reservoir water evaporated completely in the next six months. As a side effect of the water loss to the atmosphere surface water gets more concentrate. During summer the new clay deposits at the reservoir bottom dried revealing 40 cm silt and clay columns. Before the next monitored event on October 2004 took place the reservoir extension was completed. This new section allowed comparison of the infiltration from old and new reservoirs. First hours estimation of recharge rate in the old part of the reservoir was 1-2 cm hr^-1 which mostly happened through the big cracks of the shrinking clay deposit. After two days infiltration was no longer observed. On the other hand in the new reservoir rates were 1.5 cm hr^-1 at the first stage and reduced to 0.2 cm hr^-1 after two days (much more than zero in the old part), and after two weeks no water was left on top of the bottom of the new section of the reservoir.

The percolation process under natural conditions seems to be controlled by the flood water head. Percolation under reservoir conditions seem to be limited by silt and clay deposition. Through cleaned reservoir water infiltrates rapidly, similar to natural stream conditions. However in a two-day period, the silt and clay deposit which carry along with the flood water and sink on the bottom, starts to seal off the reservoir bottom, prevent water infiltrate and reduce the aquifer recharge.

In general, percolation under natural ephemeral stream and new reservoir showed same magnitude of percolation rates, 5-15 cm hr^-1. However new reservoir is no longer new after two days from the first flood event entering the reservoir, due to silt and clay deposit.

Minimal estimation of water volume recharge to the shallow alluvial aquifer by flood water under natural ephemeral stream of average width of 100 m and duration of 8 hours is 24000m3 km-1 (~2x106 m3 for the central and north Arava).