History of ecosystem model development at Colorado State University and current efforts to address contemporary ecological issues

Authors: DEL GROSSO, STEPHEN - U.S. DEPARTMENT OF AGRICULTURE (USDA); PARTON, WILLIAM - COLORADO STATE UNIVERSITY

Submitted to: American Society of Agronomy
Publication Type: Book / Chapter
Publication Acceptance Date: 2/16/2019
Publication Date: 2/15/2019
Citation: Del Grosso, S.J., Parton, W.J. 2019. History of ecosystem model development at Colorado State University and current efforts to address contemporary ecological issues. American Society of Agronomy. 8:53-70. https://doi.org/10.2134/advagricsystmodel8.2017.0012.
DOI: https://doi.org/10.2134/advagricsystmodel8.2017.0012

Interpretive Summary: Ecosystem models were initially developed in the 1960’s to explore connections between nutrient cycling and biology and how human activity alters these processes. One of the first ecosystem models developed in the 1970’s was the ELM grassland model. Extensive field observations were used to represent almost all processes including species level plant growth, soil biology, insect and mammal predation. Unfortunately, the ELM model was too complex to use for practical management questions. However, some of the sub models were used in later, simpler, ecosystem models including CENTURY. During the 1980’s CENTURY was used for site level simulations. Increased funding during the 1990’s led to model improvement and allowed for coarse resolution regional simulations and the first large scale ecosystem model comparisons using standard weather, soils and land use datasets. As interest in climate change and greenhouse gas emissions grew in the late1990’s and 2000’s, the DayCent model was developed to simulate fluxes of the full suite of biogenic GHGs. DayCent is used to estimate soil GHG fluxes for the US National GHG Inventory, and for decision support tools that compare management practices.. Current research needs include software to automatically generate model input files and mechanisms to conduct comprehensive model comparisons.

Technical Abstract: Ecosystem level models were initially developed in the 1960’s to explore connections between nutrient cycling and biology and how human activity alters biogeochemical processes. One of the first terrestrial ecosystem models developed in the 1970’s was the ELM grassland model. Extensive field observations were used to represent almost all processes including species level plant growth, soil C and N cycling, insect and mammal predation, etc.. Unfortunately, the ELM model was too complex to use for practical management questions. However, the soil temperature, water and nutrient cycling sub models were used in later, simpler, ecosystem models including CENTURY. During the 1980’s CENTURY was used for site level simulations while regional application was limited due to lack of needed datasets. Increased funding during the 1990’s led to model improvement and allowed for coarse resolution regional simulations and the first large scale ecosystem model comparisons using standard weather, soils and land use datasets. As interest in climate change and greenhouse gas emissions grew in the late1990’s and 2000’s, the DayCent model was developed to simulate fluxes of the full suite of biogenic GHGs. DayCent is used to estimate soil GHG fluxes for the US National GHG Inventory, to compare the impacts of conventional vs. improved land management strategies for decision support tools and perform life cycle assessments. Availability of multi-site, standardized data sets with comprehensive model driver and testing data has facilitated model evaluation and improvement. Current research needs include software to generate model input files and mechanisms to conduct comprehensive model comparisons.