|Institution:||Western Washington University|
|Keywords:||Information storage and retrieval systems – Soil science – Evaluation; Soil surveys – Washington (State) – Whatcom County – Mathematical models – Evaluation – Databases; Land capability for agriculture – Washington (State) – Whatcom County – Mathematical models – Evaluation – Databases; Soils – Washington (State) – Whatcom County – Mathematical models – Evaluation – Databases; Watersheds – Research – Washington (State) – Whatcom County; Water quality – Washington (State) – Whatcom County – Data processing; Biology; Whatcom County (Wash.); Academic theses|
|Full text PDF:||http://cedar.wwu.edu/wwuet/514|
Controlling pollution from agricultural lands is a priority for improving watershed health. Best management practices (BMPs) recommend strategies such as riparian buffers and altered fertilizer application timing and rates for reduction of nutrient and sediment export from agricultural watersheds, but BMP effectiveness in nutrient retention can vary greatly depending on differences in crops, soils, and topography. Conducting nitrogen (N) and phosphorus (P) measurements in all BMP projects is generally not feasible, so well-validated models can help estimate benefits on the watershed scale. This project uses the Agricultural Policy/ Environmental Extender (APEX) model to simulate crop yield, streamflow, and surface runoff in a small watershed in Whatcom County, Washington, to prepare the model for future use in estimating nutrient and sediment retention benefits by BMPs. The APEX model requires detailed inputs for soils, climate, cropping system, and agricultural management; outputs must be calibrated and validated against existing environmental data. No current consensus exists as to the ideal set of soil data for the APEX model. I tested the APEX model for three different soils datasets: the Soil Survey Geographic Database (SSURGO), the National Cooperative Soil Survey (NCSS), and the Nutrient Tracking Tool (NTT), to determine the best dataset to use in terms of ease of use and model fit. I modeled the northern Kamm Creek watershed, a 227 hectare watershed that contains a diverse representation of Whatcom County cropping systems. As the first APEX modelling effort in western Washington, this study investigated parameters for blueberry and raspberry, two crops new to the APEX model, while testing model performance with three different sets of soils data. I manipulated key parameters in two of the datasets to evaluate their effects on hydrology and yield. The model performed well for streamflow and surface runoff across all soils during calibration, with satisfactory validation for surface runoff, but not streamflow. Performance for crop yields, however, varied across both crop type and soil data sets. Simulated crop yields fell within 10% of county-reported average yields for four of the five soils for blueberry, raspberry, and corn silage crops, whereas NTT soils drastically underestimated yields of both berry crops. I recommend applying the SSURGO soils dataset to future APEX modelling in Whatcom County, as it had the best model fit for hydrology and crop yields. Further recommendations are made for obtaining data to parameterize, calibrate, and validate the model to assure accuracy for future APEX modelling efforts. Advisors/Committee Members: Hooper, David U., Helfield, James M., Miner, Benjamin G., 1972-.