Modeling Introduction

A computer-based model is a computer program that simulates what did or what might happen in a situation in order to make more informed decisions. Computer models are common tools used in research and in many industries for decision making, including agriculture. Examples include:

  • Crop models – used to make crop and fertilizer management decisions;
  • Economic models – used to minimize input cost and maximize returns;
  • Hydrologic models - used to estimate water movement across the landscape; and
  • Weather models - used to predict present and future growing season conditions

The MDA is using computer modeling to estimate nitrate leaching losses in areas with vulnerable groundwater. These models incorporate the physical, chemical and biological processes related to nitrogen conversion, uptake, release, turnover and transport within soils and crops.

The Value of Computer Models to Estimate Nitrate Leaching Loads

Field research evaluating nitrate leaching under different management practices are used to develop and calibrate these models. However, field research has practical limitations in the length of study, what field conditions are tested, and the number of management practices that are to be evaluated. Computer models can extrapolate out beyond the initial research fields to a larger region with similar management, soil, and weather conditions. They also allow for evaluation over a longer time period than most research projects. This is useful to provide a greater understanding of the range of nitrate leaching loss which could be expected under different conditions such as wet vs. dry years.

How the MDA will Use Computer Models

The MDA will model nitrate leaching losses before and after implementation of Best Management Practices (BMPs) and Alternative Management Tools (AMTs) for common cropping systems, soil types, and climate conditions within Drinking Water Supply Management Areas (DWSMAs). Monitoring data from research and demonstration projects conducted by the University of Minnesota is used to calibrate the models so they are representative of local conditions. Model results provide an estimate of the expected nitrate leaching reductions when certain practices are implemented and help support discussions about the effectiveness of different management practices on reducing nitrate leaching losses. Also, model results can provide input for local teams about what management changes may be needed to achieve nitrate leaching reduction goals including:

  • The MDA will provide the Local Advisory Team (LAT) estimates of BMP and AMT impacts on nitrate losses based on computer modeling.
  • The LAT may use this information to determine BMPs and AMTs to implement.
  • The computer models can be used to estimate nitrate leaching
    • From past cropping and weather conditions;
    • From current crop and weather conditions; and
    • After implementation of BMPs and AMTs.

Computer simulation models are one of many tools the MDA uses to understand nitrate leaching in agriculture systems. In addition to modeling, the MDA supports in-field research and on-farm demonstrations. These are complementary and related efforts. In-field research and demonstration data can help calibrate computer modeling tools for different crops in different areas of Minnesota.

Generalized Modeling Process

Graphic illustrating the modeling process. Steps include: Initial model run by MDA; Review and Comment by LAT; Next Level Modeling; Final Estimates. The model can be revised and cycle through the Review and Comment by LAT and Next Level Modeling stages before a Final Estimate is determined.

Modeling Output Example

Scenario Nitrogen Applied (lb/ac) Split N Application Reduced N Rate Cover Crop Nitrate load reduction example
Baseline* 30% Fall; 70% Spring No No No -
Scenario 1 April 15: 110; June 1: 60 Yes No No 16%
Scenario 2 April 15: 80; June 1: 40 Yes Yes No 24%
Scenario 3 April 15: 80; June 1: 40 Yes Yes Yes 34%

*Baseline conditions reflect the following fertilizer practices: 70% of the watershed was simulated with spring fertilizer application rates of 170 pounds of nitrogen per acre and 30% was fall applied at a rate of 180 pounds of nitrogen per acre. The rotation is 50% corn and 50% soybeans.

Where We Get the Data for the Model