Modeling Soil Organic Carbon in the Soil and Water Assessment Tool (SWAT)
Methodology
The Soil and Water Assessment Tool (SWAT) is a modeling tool that is a continuation of the USDA Agricultural Research Service (ARS) modeling that has been around for over thirty years.
It is designed to predict the impact of management on water, sediment, and agricultural chemical yields in ungauged watersheds
Major model inputs include weather, land slope, soil temperature and properties, plant growth, nutrients, pesticides, bacteria and pathogens, and land management (Gassman et al. 2007).
These components help to build models that simulate:
- runoff
- evapotranspiration
- crop yield and growth
- erosion
After the area was selected data for the weather and land were needed to proceed with the SWAT modeling.
The other data such as wind speed, humidity, and solar radiation were not influential enough to affect the final outcome of the result so the SWAT weather generator was used for simplicity.
Before calibrating parmeters the evaporation method was changed from Penman – Monteith to the Hargreaves calculation and the daily curve number (ICN) calculation was also changed from soil moisture to the plant evaporation method.
Then the default simulation was run in order to get a basic knowledge of how well the simulated results matched up with the actual results.
Each run must start two years prior to the desirable first year of results to allow SWAT to warm up and allow streamflow and groundwater to reach a steady state.
Changing parameter at times can seem to be a tedious and ineffective. In order to counteract that finding the more influential parameters is the best solution.
This can be done by allowing SWAT to run a Sensitivity Analysis on the model and after it has run (caution: it may take a few days for test to complete) SWAT will rank the top influential parameters (Figure 2).
A sensitive analysis estimates an outputs rate of change directly from the change of the input.
This test was run to predetermine the parameters that had the most affect on the runoff. By running this test it gave a direction for calibrating and reduced the amount of time spent on calibrating.
Figure 2. displays part of the SA for the Runoff flow used in this research.
Once a Sensitive Analysis had run the more influential parameters were identified and the calibration began Figure 3.
Optimally, when changing parameters for a model that will be reused the smallest amount of parameters should be changed to adequately match the actual data. This simplifies the usability of the model and how well it can be rebuilt by another trying to match results.
Three to four main parameters were found for each calibration.
Runoff Flow
- CN2 - The curve number is an empirical parameter that predicts a soil runoff
- CNCOEF - Plant ET curve number coefficient
- Plant ET Method
Crop/Biomass Yield
- BIO_E - Biomass-energy ratio
- HVSTI - Harvesting Index
- WSYF - Harvesting Index under highly stressed growing conditions.
- BLAI – Maximum potential leaf area index
Figure 3. Displays all of the parameter changes made during calibration in order to fit the model to the actual data collected.
Calibration