Ronald L. Bingner

EVALUATION OF ANNAGNPS ON MISSISSIPPI DELTA MSEA WATERSHEDS

Sediment and its associated pollutants entering a water body can be very destructive to the health of that system. Best Management Practices (BMPs) can be used to reduce these pollutants, but understanding the most effective practices is very difficult. Watershed models are the most cost–effective tools to aid in the decision–making process of selecting the BMP that is most effective in reducing the pollutant loadings. The Annualized Agricultural Non–Point Source Pollutant Loading model (AnnAGNPS) is one such tool. The objectives of this study were to assemble all necessary data from the Mississippi Delta Management System Evaluation Area (MDMSEA) Deep Hollow watershed to validate AnnAGNPS, and to use the validated AnnAGNPS to evaluate the effectiveness of BMPs for sediment reduction. In this study, AnnAGNPS predictions were compared with three years of field observations from the MDMSEA Deep Hollow watershed. Using no calibrated parameters, AnnAGNPS underestimated observed runoff for extreme events, but the relationship between simulated and observed runoff on an event basis was significant (R2 = 0.9). In contrast, the lower R2 of 0.5 for event comparison of predicted and observed sediment yields demonstrated that the model was not best suited for short–term individual event sediment prediction. This may be due to the use of Revised Universal Soil Loss Equation (RUSLE) within AnnAGNPS, and parameters associated with determining soil loss were derived from long–term average annual soil loss estimates. The agreement between monthly average predicted sediment yield and monthly average observed sediment yield had an R2 of 0.7. Three–year predicted total runoff was 89% of observed total runoff, and three–year predicted total sediment yield was 104% of observed total sediment yield. Alternative scenario simulations showed that winter cover crops and impoundments are promising BMPs for sediment reduction.

Runoff and Soil Erosion Evaluation by the AnnAGNPS Model in a Small Mediterranean Watershed

In order to evaluate prediction models of runoff and sediment yield in a Mediterranean environment, the distributed parameter, physically based, continuous simulation, daily time step AnnAGNPS model was applied to an experimental watershed of mainly pasture in Sicily. Results from AnnAGNPS simulations were evaluated using 7-year data monitored at this watershed. The model showed satisfactory capability in simulating surface runoff at event, monthly, and annual scales after calibration. Peak flow predictions were generally good for low flow events and poorer for higher flow rates. A high model efficiency was achieved for the 24 suspended sediment yield events recorded during the entire period of observation after reducing the roughness coefficients for both rangeland and cropland areas. The overall results confirmed the applicability of the AnnAGNPS model to the experimental conditions.

Simulating ephemeral gully erosion in AnnAGNPS

Ephemeral gully erosion can cause severe soil degradation and contribute significantly to total soil losses in agricultural areas. Physically based prediction technology is necessary to assess the magnitude of these phenomena so that appropriate conservation measures can be implemented, but such technology currently does not exist. To address this issue, a conceptual and numerical framework is presented in which ephemeral gully development, growth, and associated soil losses are simulated within the Annualized Agricultural Non-Point Source (AnnAGNPS) model. This approach incorporates analytic formulations for plunge pool erosion and headcut retreat within single or multiple storm events in unsteady, spatially varied flow at the sub-cell scale, and addresses five soil particle-size classes to predict gully evolution, transport-capacity and transport-limited flows, gully widening, and gully reactivation. Single-event and continuous simulations demonstrate the models utility for predicting both the initial development of an ephemeral gully and its evolution over multiple runoff events. The model is shown to recreate reasonably well the dimensions of observed ephemeral gullies in Mississippi. The inclusion of ephemeral gully erosion within AnnAGNPS will greatly enhance the models predictive capabilities and further assist practitioners in the management of agricultural watersheds.

Annualized Agricultural Non-Point Source model application for Mississippi Delta Beasley Lake watershed conservation practices assessment

The Annualized Agricultural Non-Point Source (AnnAGNPS) model has been developed to quantify watershed response to agricultural management practices. The objective of this study was to identify critical areas where conservation practices could be implemented and to predict their impact on Beasley Lake water quality in the Mississippi Delta using AnnAGNPS. Model evaluation was first performed by comparing the observed runoff and sediment from a US Geological Survey gauging station draining 7 ha (17 ac) of Beasley Lake watershed with the AnnAGNPS simulated runoff and sediment. The model demonstrated satisfactory capability in simulating runoff and sediment at an event scale. Without calibration, the Nash-Sutcliffe coefficient of efficiency was 0.81 for runoff and 0.54 for sediment; the relative error was 0.1 for runoff and 0.18 for sediment, and the Willmott index of agreement was 0.94 for runoff and 0.80 for sediment. The quantity of water and sediment produced from each field within the Beasley Lake watershed, quantity of water and sediment reaching Beasley Lake, and the potential impact of various USDA Natural Resources Conservation Service conservation programs on water quality were then simulated. High sediment-producing areas for nonpoint source pollution control were identified where sediment loads could be reduced by 15% to 77% using conservation practices. Simulations predicted that converting all cropland to no-till soybeans (Glycine max [L.] Merr.) would reduce sediment load by 77% whereas no-till cotton (Gossypium hirsutum L.) would reduce it 64%. The approach taken in this study could be used elsewhere in applying AnnAGNPS to ungauged watersheds or watersheds with limited field observations for conservation program planning or evaluation.

Conservation practice effects on sediment load in the Goodwin Creek Experimental Watershed

Water quality and aquatic habitat due to unstable stream channels and high sediment concentrations during storm runoff events are major environmental concerns on the 2,132 ha (5,266 ac) Goodwin Creek Experimental Watershed in north Mississippi. Effects of enrolling erodible lands in the Conservation Reserve Program (CRP) and instream grade stabilization structures were evaluated using measured rainfall, runoff, and sediment concentration data and model simulations. Signatures of naturally occurring radionuclides indicated that 78% of the total sediment load originated from channel sources. The change of land to a CRP-like state (reducing cultivated land from 26% to 8%) reduced erosion and runoff from fields and thus decreased total sediment concentration by 63% between 1982 to 1990. Simulations using the Fluvial Routing Analysis and Modeling Environment model indicated that mean sediment yields would increase from 15% to over 200%, depending upon location in the watershed, if in-channel structures were not present. The combined effect of the grade control structures and the change of lands to a CRP-state was to reduce sediment yields by 78% near the outlet of the watershed.

Modeling the Evolution of Incised Streams. III: Model Application

Incision and the ensuing widening of alluvial stream channels represent important forms of channel adjustment. Two accom- panying papers have presented a robust computational model for simulating the long-term evolution of incised and restored or rehabili- tated stream corridors. This work reports on applications of the model to two incised streams in northern Mississippi, James Creek, and the Yalobusha River, to assess: 1 its capability to simulate the temporal progression of incised streams through the different stages of channel evolution; and 2 model performance when available input data regarding channel geometry and physical properties of channel boundary materials are limited in the case of James Creek. Model results show that temporal changes in channel geometry are satisfactorily simulated. The mean absolute deviation MAD between observed and simulated changes in thalweg elevations is 0.16 m for the Yalobusha River and 0.57 m for James Creek, which is approximately 8.1 and 23% of the average degradation of the respective streams. The MAD between observed and simulated changes in channel top width is 5.7% of the channel top width along the Yalobusha River and 31% of the channel top width along James Creek. The larger discrepancies for James Creek are mainly due to unknown initial channel geometry along its upper part. The model applications also emphasize the importance of accurate characterization of channel boundary materials and geometry.

Modeling long-term soil losses on agricultural fields due to ephemeral gully erosion

It is now recognized worldwide that soil erosion on agricultural fields due to ephemeral gullies may be greater than those losses attributed to sheet and rill erosion processes. Yet it is not known whether the common practice of repairing or obliterating these gullies during annual tillage activities exacerbates or mitigates soil losses over long time periods. Here, a numerical model is used to demonstrate the potential effects of annual tillage on cumulative soil losses from four geographic regions plagued by ephemeral gullies as compared to no-till conditions where the gullies are free to grow and evolve over time and space. Historical precipitation data and field measurements were compiled for specific sites in Belgium, Mississippi, Iowa, and Georgia, and the model simulated ephemeral gully development and evolution during the growing seasons over a continuous 10-year period. When agricultural fields are not tilled annually, the simulations suggest that gullies attain their maximum dimensions during the first few years in response to several relatively large runoff events. During subsequent runoff events, the gullies no longer erode their channels significantly, and soil losses due to gully erosion decrease markedly. When agricultural fields are tilled annually, the ephemeral gully channels are reactivated, thus causing significant soil losses each year in response to runoff events. Over the 10-year simulation, the modeling results suggest that erosion rates in these four geographic regions can be 250% to 450% greater when gullies are tilled and reactivated annually as opposed to the no-till condition. These results reveal that routine filling of ephemeral gully channels during tillage practices may result in markedly higher rates of soil loss as compared to allowing these gullies to persist on the landscape, demonstrating a further advantage of adopting no-till management practices.

PHOSPHORUS COMPONENT IN ANNAGNPS

The USDA Annualized Agricultural Non-Point Source Pollution model (AnnAGNPS) has been developed to aid in evaluation of watershed response to agricultural management practices. Previous studies have demonstrated the capability of the model to simulate runoff and sediment, but not phosphorus (P). The main purpose of this article is to evaluate the performance of AnnAGNPS on P simulation using comparisons with measurements from the Deep Hollow watershed of the Mississippi Delta Management Systems Evaluation Area (MDMSEA) project. A sensitivity analysis was performed to identify input parameters whose impact is the greatest on P yields. Sensitivity analysis results indicate that the most sensitive variables of those selected are initial soil P contents, P application rate, and plant P uptake. AnnAGNPS simulations of dissolved P yield do not agree well with observed dissolved P yield (Nash-Sutcliffe coefficient of efficiency of 0.34, R2 of 0.51, and slope of 0.24); however, AnnAGNPS simulations of total P yield agree well with observed total P yield (Nash-Sutcliffe coefficient of efficiency of 0.85, R2 of 0.88, and slope of 0.83). The difference in dissolved P yield may be attributed to limitations in model simulation of P processes. Uncertainties in input parameter selections also affect the model’s performance.

Conservation practices and gully erosion contributions in the Topashaw Canal watershed

Quantifying the effectiveness of conservation practices at the watershed scale throughout the nation has been identified as a critical need. Our objective was to determine the effectiveness of these conservation practices for reducing sediment yield. The Topashaw Canal watershed (TCW), an 11,000-ha (27,181-ac) area in northcentral Mississippi, exhibits flashy stream response to storms with mean sediment concentrations (117 mg L-1 [117 ppm]) almost double the median sediment concentration (60 mg L-1). The most prevalent conservation practice imposed by acreage, since 1985, is enrollment in the Conservation Reserve Program (e.g., planting of pine trees). Grade-stabilization structures (e.g., drop pipes) are the most common conservation practice used to control gully erosion within the TCW. These structures are estimated by the USDA Natural Resources Conservation Service to reduce annual sediment yield from 11.5 to 0.1 Mg ha-1 yr-1 (5.13 to 0.05 tn ac-1 yr-1), but measurements have not been made to determine the accuracy of these estimates. Nonetheless, an average of 58 drop pipes have been installed annually within the TCW using Environmental Quality Incentives Program funds, and an additional 5.4 large drop pipes have been installed each year using US Corps of Engineers funds. Annual gully erosion accounted for 54% of the total sediment yield of over 73,000 Mg (80,445 tn) from TCW. The shift in land use to Conservation Reserve Program, combined with channel incision, has resulted in streambank failure and gully erosion being the primary sources of sediment currently leaving the watershed.

Effect of Topographic Characteristics on Compound Topographic Index for Identification of Gully Channel Initiation Locations

Abstract. Sediment loads from gully erosion contribute to water quality problems, reduction in crop productivity by removal of nutrient-rich topsoil, and damage to downstream ecosystems. The identification of areas with high potential for gully channel development is often performed using spatially derived stream power estimates from second-order topographic indices, such as the compound topographic index (CTI). The utilization of CTI to identify where gullies develop is affected by field and local topographic characteristics and DEM resolution. In this study, the effect of overall terrain slope, local relief variance, and raster grid cell size on CTI cumulative distribution values was investigated using theoretical and observed catchment methodology. In the theoretical analysis, stochastic methods were used to generate simulated catchments to quantify the influence of overall terrain slope, local relief variance, and raster grid cell size (each considered individually). The observed methodology used three sites with distinct topographic characteristics, measured gully channels, and high-resolution topographic information. Raster grids for the three observed study sites were generated at varying raster grid cell sizes. Critical CTI values were determined through comparison of measured gully thalwegs with threshold CTI raster grids of the observed watersheds at different resolutions. Results from the theoretical investigation indicate that CTI values were linearly influenced by changes in relief variance and overall slope, while variations in raster grid cell size caused an inverse power variation in CTI values. In addition, variations in raster grid cell size, produced changes in cumulative distributions of the top 0.1% CTI values. The use of normalized CTI values (CTI n ) produced merged cumulative distribution curves when varying overall slope, terrain relief variance, and to a lesser degree DEM resolution. Similar findings were obtained from the analysis of observed catchments. When DEM resolution varied, the differences in critical CTI n values in the same field were significantly reduced when compared to original critical CTI values, although differences were not fully eliminated. Normalization of the CTI cumulative distributions improved comparisons between different sites with distinct drainage area sizes and topographic characteristics, providing a possible alternative for investigations of large watersheds with more than one topographic characteristic. Results suggest that a normalized critical CTI between 1 and 2 could be used for the identification of areas with high potential for gully development. Knowing where gullies develop is important in understanding the effect of conservation practices on soil erosion through the use of field-scale and watershed-scale simulation models. Effective watershed management plans depend on this information to target the placement of conservation practices for the efficient use of available resources.