Procedure to generate precipitation data for simulating long term soil erosion from a hillslope area with event oriented models: The design storm methodology
Abstract
Process oriented hydrologic/erosion models which predict long term soil loss from agricultural land with continuous weather data often use simplified relationships to reduce computational time. There is a great need to use the detailed, fundamental computations available with event oriented models to determine long term impact without having to simulate every precipitation event of a continuous weather sequence. This study developed a procedure to generate a few precipitation events which are suitable for evaluating average annual soil erosion from hillslope areas with event oriented models. This procedure, called the design storm methodology (DSM), preserved the seasonality of soil loss and allowed flexibility for determining the exact number of design storms to be simulated. The DSM is composed of three components: biweekly erosion probability distribution (EPD); a relationship between the fraction of average annual erosion and precipitation volume (FP); and monthly average total storm energy to precipitation volume ratios (EPR). The EPD and EPR can be developed from a single long term simulation of a landscape which represents the midpoint of landscape features; such as slope steepness and profile length. The FP curve varies with the runoff potential of the soil. The parameters of the FP relationship were related to the curve number, which was utilized to represent soil runoff potential. The CREAMS model was used in this study. CREAMS was modified to stochastically account for variable storm erosivity and peak runoff rates. Individual events were simulated with the erosion subprogram of CREAMS utilizing the DSM. The DSM was evaluated by comparing simulations made with the generated design storms with simulated long term results. The design storms simulation results adequately reproduced the long term average soil loss, the seasonal distribution of soil loss, and the primary particle sizes of the eroded sediment for soils with moderate runoff potential. The design storm simulations tended to underpredict average annual soil loss for soils with high runoff. The design storm simulations tended not to reproduce the eroded particle sizes for soils with low runoff. Overall, the DSM generated storms reproduced long term average results well.
Degree
Ph.D.
Advisors
Beasley, Purdue University.
Subject Area
Agricultural engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.