where the chemical contents are in milliequivalents per liter. The observed water temperature are in degrees Celcius.
The soil number, series name, textural classification, nearest town to site, and date of erosion are presented. The clock times for the start and end of periods one (Rain Only) and two (Rain and Flow Addition) are also given. The "sample" time for the Rain Only period was the clock time noted when there was sufficient runoff from the six rills to begin rill data collection.
The quality of the eroding water is presented. The chemical anlayses were carried out by the SCS National Soil Survey Laboratory in Lincoln, NE. The chemical contents are listed in milliequivalents per liter. SAR is the sodium absorption ratio (USDA SCS, 1984), and is found by:
where the chemical contents are in milliequivalents per liter. The observed water temperature are in degrees Celcius.
The results of the bulk density and soil moisture observations are presented. The soil series and date of experiment are recorded in the heading. The "Before rain" samples were collected before period one. The "After rain" samples were collected after period two for the soil moisture analysis, and after period three for the bulk density analysis. The plot numbers refer to the rill and infiltration plots
as numbered in
Figure 3. The bulk density "Depth" is the approximate depth of excavation for each of the observations. The bulk density is in grams of soil per cubic centimeter excavated. The moisture contents for the bulk density samples are recorded in both grams of water per gram of soil, and cubic centimeters of water per cubic centimeter of soil. The relationship between these two moisture contents is
The "Depth" for the moisture content analysis is an approximate depth of sample obtained by the soil probe. The moisture content (MC) is in grams of water per gram of soil.
On the Sharpsburg site, the soil surface was covered with large, dry, hard clods 50 to 100 mm in diameter. The presence of the clods made it impossible to measure the bulk densities before the start of the experiment of the Sharpsburg site.
Because of a break in the field-data collection, the wet weights for the Heiden bulk density samples were not obtained, and so no estimate of the moisture content is recorded for the soil from the bulk density analysis of the Heiden site.
On the Caribou site, the surfaces were covered with large stones at the completion of erosion. The stone cover was too great to allow the bulk density to be measured of the Caribou site. Other samples that were lost have been noted in the results.
The soil strength observations from the rill plots are recorded. Strength readings were made after period one. In 1987, the fall cone was used after period two, and the other strength readings taken after period three. In 1988, the fall cone was read after period three. In 1988, the fall cone was read after period three along with the other strength readings. Observations were made at distances of 2, 5, and 7 meters from the tops of rills 2, 4, and 6 (see Figure 3). The units for the torvane, internal vane, and pocket penetrometers are in kPa. For the fall cone penetrometer, a strength index (Towner, 1973) is recorded from:
The torvane and pocket penetrometer (Penetrometer measurements were carried out on both the side and the bottom the rill at each observaton location. The fall cone penetrometer (Fall Cone) and internal vane (Inter. Vane) observations were made on the rill bottom.
The fall cone results presented are means of three observations at each location. Because of the form of Equation {13}, when the fall cone registered a very low fall displacement due to a rock or root located immediately beneath the tip, a very large index was calculated.
Where no sample was taken because a restrictive layer had been reached, "NS" was entered in the table. In 1987, no internal vane readings were taken on some soils, and these sections of the results have been left blank. (Soil Strength Analysis result).
The soil strength observations from the external plot are presented. Units are the same as for the rill plots. The times recorded are clock times, but the time of wetting the plot has not been recorded. When more than a single observation was made with a torvane or penetrometer, all readings have been recorded. The fall cone readings are all based on the average of there readings.
A summary of the sediment size analysis is presented. The results combine the aggregates and particles that were retained on the seives, for all sizes over 63 microns. For under 63 microns, a particle bulk density of 2.63 was assumed in the calculation of the respective pipetting times for each analysis.
Plots 6 and 8 were interrill plots and plots 3 and 5 were rill plots (see Figure 3). Under "Type", I refers to interrill, and R to rill. In the time column, A was sampled near the end of period one, B near the end of period two, and C near the end of period three.
The percentage of soil by weight is given for each size category. The 32 to 63 micron size contains the "remainder" that was not accounted for in all of the smaller size classes. Those values in the 32-63 micron range, formatted to four decimal places, are the sum of sediment retained on a 38 micron sieve and the greater than 32 micron size category in pipetting.
Any negative percentages are a result of small discrepancies in sampling and weighing during pipetting. The large negative values noted for Williams are suspected to be due to failure to calibrate the laboratory balance prior to carrying out the final mass observations for this soil. For a discussion of the possible sources of error in the analysis method, refer to Meyer and Scott (1983).
The results from the interrill runoff samples are presented by site. The number and shape are given for each plot (Figure 3). The rainfall rate (INTENSITY) collected at the upslope side of each plot (Figure 4) is given in the first column. The clock time is the time from the start of the rainfall until the start of the respective sample collection. The runoff flow rate is in millimeters per hour. The erosion rate is in grams per minute per square meter. The concentration of sediment in the runoff (Conc.) is listed in the final column in grams per liter. The four detachment values that were used to calculate Ki have been underlined. In cases where one of the final four values was not considered as a true indication of the erosion rate, the reason for not selecting that value has been noted.
The most common sources of error during sampling these plots were from mud splashed into the bottle by the person collecting the sample, the bottle overflowing, the bottle tipping over, or runoff leaking around the sampling trough. Much of the erosion observed on the covered plots was probably from raindrop induced splashing into the collection trough from outside the plot. Another source of variation between plots was due to the angle at which rainfall contacted the plots because of the horizontal velocity imparted to the drops by the simulator. This effect, which lead to differences between the left side and right side plots was not found to be statistically significant (Liebenow et al, 1989).
On some sites, wind may have influenced the raifall distribution. The rainfall and erosion of those plots located near the outside of the wetted area on the side facing the wind were sometimes lower than the other plots. An example of this can be seen on the the Hersh reults where the three plots furthest downslope, plots 3, 7, and 10 had rainfall intensities about 20 percent lower than the other seven plots.
One complete set of plot data, plot 8 on the Heiden clay, was not used because the erosion rates were much lower than had been observed on the other Heiden plots. Field notes stated that the thickness of the topsoil on this plot was less than on the others, and this may account for the lower runoff and erosion rates observed.
During 1988, additional runoff samples were collected from plots 4 and 5 during period two. The intensity during this period as well as the runoff observations are recorded. The time recorded for these samples is from the beginning of the rainfall for period two (click here).
Several crates of bottles were dried before initial weighing on the Tifton site. For these plots, only sediment content was available to allow calculation of the erosion rate.
The calculation of the interrill erodibiltiy, Ki is presented. "Plot" refers to the plot number (Figure 3). "type" is entered as "R" for ridge, interrill plots, and "F" for flat uncovered, infiltration and erosion plots. The "Side" column indicates the side of the simulator on which the plots were located, when facing down-slope. The plot means of the four Erosion rates underlined on the INTERRILL DATA are entered as Di. "I" is the rainfall intensity for each respective plot in millimeters per hour.
"Slope" is expressed as a percentage. For the flat infiltration plots, the slope was that of the nearest rill plot. For the ridged, interrill plots, the slope was calculated from topographic data available from photogrammetric analysis when available, or was assumed to be 57 percent on clays and 60 percent on silts for 1987, and 51 percent on all soils for 1988. The slope factor (Equation {3}) is not sensitive to variations in slopes at these magnitudes. Ki was calculated by solving Equations {2} and {3} for Ki.
The units have been standardized to give Ki in kg s m -4. The means are presented for each soil. The standard deviations and coefficients of variation for each soil are included in the summary.
The runoff rates and sediment concentrations from the rill plots for period one are reported. The clock time recorded in the first column applies to all rills. The starting time for each site can be found click here. The rill flows are given in liters per minute,and the sediment concentration of the out flow in grams per liter. Because no velocity measurements or photographs were taken during this period, it was not possible to calculate a rill width, hydraulic radius, or erosion rate for a given area, or erodibility. The time required to reach flow equilibrium varied with soils, so there are different sizes of data sets.
The sediment contained in the runoff during this low-flow period was detached by interrill erosion, and depended on the rill flow to transport it to the outlet. On most soils, there was net deposition in the rill bottom during these periods.
No runoff occurred during period one for the Bonifay series due to the high infiltration rates. Other missing data resulted from spilled bottles or other problems encountered during sampling or laboratory analysis.
The data used to calculate the rill erodibility are presented. The heading for each soil summarizes the date, specific weight of water (gamma), the transport coefficient (B), the Kinematic viscosity (mu), the velocity factor, and the mean interrill detachment rate (Di). Some of the Di's used differed slightly from the final values found in the interrill analysis, but the small discrepanices would have a negligible effect of the rill erodibility, and so they were not altered. The specific weight and the viscosity were based on the observed water temperature. The slope for each rill is recorded in the first line of each individual set of rill data.
In the source column, the codes refer to the nominal flow addition rates. R+0 was rain plus zero flow addition, R+2 was rain plus 2 gallons per minute, F+2 was flow addition only (no rainfall) at 2 gallons per minute, etc. The outflow rate is in liters per minute.
The maximum velocity (M Vel) observed during the dye addition is in meters per second. The average velocity (A Vel) in meters per second, is the maximum velocity multiplied by the Velocity Factor given in the heading. The Area in square centimeters is the observed Flow divided by the average velocity.
The widths (wr), in centimeters, were measured on site photographs for each flow rate for period two. On Sharpsburg, and on all soils for period three, the iteration to ensure that the calculated widths matched the observed widths was not carried out. No photographs to measure widths were available for Sharpsburg, and time has not been available to carry out this analysis for period three. For these observations, the cross-sectional from the rill meter before a given set of flows were used for R+0, R+2, R+4, R+6, F+2, F+4, and F+6. The cross-sections from the rill meter after the event were used for R+6, R+8, R+10, F+6, F+8, and F+10. For the R+6 and F+6 conditions, both the before and after results were used in the regression analysis to find Kr and Tauc.
The hydraulic radius (rh) in centimeters was found by methods discussed on previously, from the width, the cross-section area, and the rill meter data. The sediment transport rate, Qs, is the product of the flow and the concentration, in grams per second. The hydraulic shear (tau) in Newtons per square meter, was calculated by equation {5}:
The interrill contribution to detached sediment, E, in grams per second per square meter of rill, was found by the following formula:
The dimensionless Darcy-Weisback friction factor, F was found by the following equation:
On the Woodward soil, a nonerodible layer was reached at the end of the second period. This may have occurred near the end of period three on other soils, but not to the extent that was observed at Woodward.
The summary of Kr, Tauc, and Ki values are shown in Figures 5, 6, and 7, and presented in tables. The mean, standard deviation, and coefficient of variation are given for each soil. Regressions were carried out for both detachment rate, Dr, and detachment capacity, Dc, as a linear function of shear, tau. The results of both regressions are presented.
Those sets of Kr and Tauc in brackets were considered to be outliers, and were not used in calculating the mean and standard deviation. Where "unstable" has been entered for a given rill, the iteration method to find Kr and Tauc did not converge.
The critical shear for Keith on Figure 7 has been shown as zero, and not the calculated value of -2.8.
A summary of the soil properties measured by the SCS National Soil Survey Laboratory is presented. These properties are from the A horizon of the central pit. Other horizons, and other pits were also sampled, and other properties were analyzed, but the persentation of the 2500 pages containing these results was beyond the scope of this publication. The methods for finding these properties were described by USDA SCS (1984 and 1989).
The clay, silt, very fine sand (VF Sand) and water dispersible clay are given as percentage of the sample that passed through a 2 mm sieve. The very fine sand is included in the total sand content. The rock is the percentage of the total soil sample that was greater than 2 mm.
The coefficient of linear extensibility (COLE) denotes the fractional change in clod dimension from a dry to a moist state. When coarse fragment are absent:
The soil moisture retained by a soil sample at 1/3 bar moisture tension and at 15 bar moisture tension, in percent by weight, are given in the 1/3 B and 15 B columns. The percentage of aggregates in the 2-1 mm range that were soaked overnight and then retained on a .5 mm sieve after 20 oscillations are recorded as the aggregate stability (Ag Stab).
The surface area of a soil sample can be found using ethylene glycol monethyl ether (EGME) retention. The amount of EGME retained is presented. This value can be converted to surface area by the relationship:
The contents of Calcium (Ca), Magnesium (Mg), Sodium (Na), and Potassium (K) are given in milliequivalents per 100 grams, as found by atomic absorption. The cation exchange capacity (CEC) in milliequivalents per 100 grams as found by Ammonium Acetate titration is presented. Iron and aluminum contents given in percentages were found using dithionite-citrate extraction methods.
The sodium absorption ratio (SAR) was found by:
where Na, Ca, and Mg are milliequivalents per liter.
The conductivity in micromhos was measured in a paste of the soil. The percentage of organic carbon (Org C) was found by acid-dichromate digestion. The percentage of calcium carbonate was found using HCl treatement.
Descriptions of the locations of each erosion site, as provided by the SCS soil survey report, are summarized.
Sketch maps showing the approximate location of each site are presented. Local assistance should be obtained to pinpoint exact plot locations.
The authors wish to thank the numerous individuals who have been of assistance in planning, collecting, and analyzing this data. For the full list click here. The list includes:
ARS scientists, technicians, administrators, and secretaries who helped plan, coordinate, and carry out the project, and who shared research facilities and equipment
University faculty and staff who helped coordinate the experiment, shared research facilities, and assisted with data collection and analysis
Undergraduate and graduate students from several universities who assisted in data collection and analysis.
SCS field and laboratory staff who assisted in identifying sites, carrying out site soil analyses, collecting data, and identifying and analyzing soil properties
Farmer cooperators and land owners who allowed access to their fields, prepared experimental sites, and assisted with data collection
The first name listed under each soil was the contact person for that site. Without the cooperation and assistance of these individuals the completion of this major research project would not have been possible. The authors wish to apologize if any individuals have been unintentionally overlooked.
