Agriculture model development to improve performance of the Community Land Model
The important relationships between climate change and agriculture are uncertain, particularly the feedbacks related to the carbon cycle. Nevertheless, vegetation models have not yet considered the full impacts of management practices and nitrogen feedbacks on the carbon cycle. We are working to meet this need.
We have integrated three crop types (corn, soybean, and spring wheat) into the Community Land Model (CLM). In developing the agriculture version of CLM, we added plant processes related to management practices and nitrogen cycling. A manuscript documenting our changes to CLM has been accepted for publication in Geoscientific Model Development Discussions (“Modeling Agriculture in the Community Land Model,” by B. Drewniak, J. Song, J. Prell, V. R. Kotamarthi, and R. Jacob).
Our agriculture model, CLM-Crop, uses the Agro-IBIS growth scheme based on four stages of crop development: planting, emergence, grain fill, and harvest. Planting date and growth period are based on the Crop Calendar Dataset published in 2010 by Sacks and colleagues. Fertilizer application during emergence is based on typical U.S. rates. Soybean has additional nitrogen fixation, as in the SWAT (Soil and Water Assessment Tool) model. A dynamic root scheme allows roots to grow linearly during the growth period and expand in soil layers in coincidence with water and nutrient availability. During grain fill, nitrogen is mobilized from leaves and stems to meet organ demands. At maturity, all grain and portions of leaves and stems are harvested. Several of these components of CLM-Crop will be included in the CLM4.5 release, scheduled for May 2013.
Results of our case study with CLM-Crop showed the following:
- Productivity and yield declined when residue was harvested for biofuel use and increased when nearly all the residue was returned to the soil. Therefore, using residue to meet biofuel demands might not be a sustainable farming practice.
- When planting date responded to climate, the temperature at planting was critical.
- When planting date was earlier than average, soybean yields increased, but corn yields decreased because of nitrogen limitation.
- Gross primary productivity decreased in regions where crops were grown intensively.
- Yields of corn and wheat simulated by CLM-Crop compared well with published U.S. observations for 2008 (Monfreda and coworkers), but soybean yield was overestimated.
Our analysis of CLM showed strong sensitivity of crop productivity to soil nitrogen. CLM-Crop indicated a negative correlation between yield and temperature and a positive correlation between yield and precipitation, demonstrating that as temperature increases, yields could decline.
We are working to improve the parameter calibration of CLM-Crop in the next release of CLM, in hopes that the new below-ground biogeochemistry n-layer model will improve nitrogen cycling. We will then consider the impacts of agriculture on soil carbon storage.