Development of an air-entrained rainfall simulator for field plots

Rainfall simulators have been widely used for many years to accelerate and extend studies of soil erosion, infiltration and run off. Simulators make possible the production of rainfall of selected intensities and durations at any desired time and place. To obtain desirable drop characteristics, most existing simulators produce high application rates. These rates are reduced to desirable intensities by means of intermittent movement of the drop formers. The purpose of this study was to develop a rainfall simulator which would produce non-intermittent rainfall of desirable intensities and durations. The objectives of this research were (1) to design and construct a non-intermittent rainfall simulator, (2) to evaluate the uniformity of distribution obtained with the rainfall simulator and (3) to determine the fall velocity and size distribution of raindrops at selected intensities of simulated rainfall.

An air-entrained rainfall simulator capable of producing nonintermittent rainfall of selected intensities was developed with locally available and easily fabricated materials. It is easy to assemble, disassemble and transport, covers an effective area of 6.1 m X 6.1 m (20 ft x 20 ft) and has a provision for connecting additional units to cover large field plots. To vary the rainfall intensity, compressed air was injected into the water conduit through slits located 180 degrees apart and transverse to the direction of water flow. Photographs of the raindrops were taken using a high speed camera in order to determine fall velocity and size distribution of the raindrops. Depending upon combination of water-air pressure, one type of nozzle produced intensities ranging from 65 to 107 mm/h (2.6 to 4.2 in/hr). Distribution uniformity was satisfactory and median drop diameter (mass basis) was in the 2.0 to 2.5 mm range. Using a larger nozzle and various combinations of water-air pressure, intensities ranging from 85 to 154 mm/h (3.3 to 6.1 in/hr) were obtained. In one test, however, excessive air pressure resulted in a median drop size too small for this magnitude of intensity. Fall velocities of the simulated raindrops produced by one nozzle were approximately 15 percent more than the terminal velocities of natural raindrops of equal size. The high velocities were attributed to insufficient fall distance due to laboratory constraints and to inability to precisely determine drop size. While results were generally satisfactory, several minor modifications are suggested to improve performance and testing.