Skip to main content
ARS Home » Northeast Area » University Park, Pennsylvania » Pasture Systems & Watershed Management Research » Research » Publications at this Location » Publication #257071

Title: Development of a web-based runoff forecasting tool to guide fertilizer and manure application in the Chesapeake Bay watershed

item Buda, Anthony
item Kleinman, Peter
item Feyereisen, Gary
item Bryant, Ray
item KNIGHT, PAUL - Pennsylvania State University
item MILLER, DOUGLAS - Pennsylvania State University

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 7/16/2010
Publication Date: N/A
Citation: N/A

Interpretive Summary: An interpretive summary is not required.

Technical Abstract: Managing the land application of fertilizers and manures is critical to protecting water quality in the Chesapeake Bay watershed. While modern nutrient management tools are designed to help farmers with their long-term field management planning, they do not support daily decisions such as when to apply fertilizers and manures to farm fields. Applying fertilizers and manures at the wrong time (e.g., immediately preceding a rainfall event that produces surface runoff) unnecessarily increases the risk of surface water contamination. This poster describes efforts to develop a web-based tool that will utilize daily weather forecast data from the Pennsylvania State Climatologist (rainfall amount, rainfall intensity, soil moisture) to predict the probability of surface runoff over 24, 48 and 120 hour periods and display the risk of nutrient loss in the form of a “stop light” map (red = high risk, yellow = moderate risk, green = low risk). In support of tool development, a proof-of-concept study was conducted using a three-year runoff dataset from the Mattern experimental watershed in central Pennsylvania. Preliminary results suggest that the probability of surface runoff at a given landscape position can be quantified using simple logistic regression models that relate daily runoff occurrence (1 = yes, 0 = no) to weather forecast variables. This effort will be expanded to other sites in Pennsylvania with long-term surface runoff data in order to evaluate the suitability of logistic regression models for predicting the probability of surface runoff at the field scale and determine whether the models can be applied to develop a statewide runoff forecasting tool.