Reducing the error in infiltration determinations by means of buffer areas

A method is described for determining the infiltration capacity of a soil by using large plots protected against loss of applied water through lateral seepage. Approximately square 0.016-acre plots were laid out on a 4.6% slope and surrounded by metal borders sunk into the ground to a depth of 6 inches. Surrounding the plot was a border space which had the same treatment as the plot itself. About 4.5 feet outside the plot lines a trench was dug and water run through this until a belt of soil 7 to 10 feet deep around the plot was saturated. This served as a buffer area through which water and air could not pass laterally from the plot area on which infiltration tests were to be made. The plot together with the border area was spaded and covered with straw and then sprinkled by means of a multiple set of stationary sprinklers. The runoff from the 0.016-acre plot was collected and measured and the total intake and infiltration rate determined. A small plot, 16 by 72 inches, was placed in the middle of the 0.016-acre plot area and given the same preparation and straw treatment. The application of water and runoff were determined by a special system entirely separate from that used on the large plot. The results show that the intake of water on the large plots and on the small plot within were similar. Other of these small plots were placed in the open field where they had no water applied to the surrounding soil and no wet buffer belt to prevent lateral movement of air or water. The intake on these isolated small plots having no prewetted border protection was 75% greater than for the large plots. These results indicated that lateral movement of water had allowed these small plots to take in more water than would be possible if it were raining over the entire surface. Since the large plots were protected against lateral movement of water beneath the surface, the results on these plots should represent about what would be the intake possibilities if rain were falling over an entire watershed having the same soil and surface conditions as did these plots. The method used on these large plots might provide a means for determining with a fair degree of accuracy the infiltration capacities of a wide range of soils and surface conditions under natural rainfall. The method might be made to supplement the use of small plots when total intake capacities rather than the effect of specific surface treatments on intake are desired.