Grantee Research Project Results
Final Report: Close-coupling of Ecosystem and Economic Models: Adaptation of Central U.S. Agriculture to Climate Change
EPA Grant Number: R828745Title: Close-coupling of Ecosystem and Economic Models: Adaptation of Central U.S. Agriculture to Climate Change
Investigators: Antle, John M. , Capalbo, Susan M. , Hunt, William , Paustian, Keith , Mooney, Sian , Hoagland, Kyle D.
Institution: Montana State University - Bozeman
EPA Project Officer: Packard, Benjamin H
Project Period: October 1, 2000 through September 1, 2003
Project Amount: $1,420,860
RFA: Assessing the Consequences of Interactions between Human Activities and a Changing Climate (2000) RFA Text | Recipients Lists
Research Category: Climate Change , Air
Objective:
- Develop methods to more closely couple existing ecological and economic models that can be used to assess the impacts of climate change in agricultural ecosystems. This involves linking processes in ecological models with land use and input use decisions in economic models, so that the type and strength of feedback between ecological and economic processes is suitably represented.
- Simulate the ecological and economic impacts of climate change on agriculture in the 21-state central U.S. (CUS) region and in study areas in Montana and Nebraska, using data at various scales (field/farm, county and Major Land Resource Area (MLRA)), and using a range of climate change scenarios and sensitivity analyses (figure 1).
- Investigate the dynamic and spatial properties of agricultural ecosystems to assess how estimates of the impacts of climate change are affected by the choice of spatial scale, temporal scale, and degree of model coupling. These properties will be compared at the farm/field, county and MLRA scales in the central United States using primary data collected by the PIs, and secondary data collected by various state and federal agencies.
Summary/Accomplishments (Outputs/Outcomes):
Using the Century ecosystem model and outputs of the Canadian Climate Model, three climate scenarios were simulated: a climate change scenario with perturbed temperature and precipitation, referred to henceforth as the TP scenario; a CO2 fertilization scenario in which CO2 concentration is increased without changes in temperature or precipitation; and a combined temperature-precipitation-CO2 scenario referred to here as TPC. To simplify the presentation, we focus on results for the two scenarios of primary interest, the TP and TPC scenarios. These two scenarios provide lower and upper bounds on the like effects of climate change and enhanced atmospheric concentrations of CO2. The simulations represent climate change over the period 2000 to 2050.
Analysis of Climate Change Vulnerability
We developed economic measures of vulnerability to climate change with and without adaptation in agricultural production systems. With a model of the dryland grain production systems of Montana, we found support for the hypothesis that the most adverse impacts on net returns distributions tend to occur in the areas with the poorest resource endowments and when mitigating effects of CO2 fertilization and adaptation are absent. The results also show that relative and absolute measures of vulnerability depend on complex interactions between climate change, CO2 level, adaptation, and economic conditions such as relative output prices.
Ecological Impacts of Climate Change
Our analysis using the Montana case study also shows that the spatial distributions of net returns, and thus the choice of production system, vary systematically with assumptions about climate impacts, CO2 fertilization, and adaptation. Accordingly, the ecological impacts of climate change also vary according to these factors. We find that adaptation has substantial effects on both ecological outcomes (changes in soil carbon) and economic vulnerability. Without adaptation, each system is characterized by lower soil carbon stocks, as well as lower mean economic returns and higher spatial variability in returns, relative to the same system simulated with endogenous adaptation.
Chaos and sensitivity to initial conditions
Our work on the dynamic properties of agroecosystems includes studies of chaos and of multiple stable equilibria. We found that an empirical method widely applied for detecting chaos in ecological time series is unreliable when applied to time series data typically available for economic data. In addition, we have tested the model for the Central US region and found that while these perturbations may impart model divergence for long time periods (e.g., hundreds of years) the magnitude of the divergence is not large enough to substantially change the policy implications of the analysis.
Effects of Spatial Scale
The analysis of the effects of spatial scale involves the comparison of results from the Montana and Nebraska models to the results from the CUS model, where the results are overlaid by Montana sub-MLRAs and Nebraska MLRAs. Comparison across models constructed at different spatial scales is complicated by the fact that the farm-level models provide more detail on input use and cost of production, but aggregate models are able to represent all potential land uses. Our results indicate that the farm-level and county-level models gave substantially different predictions of climate change, with the farm-level models generally predicting smaller impacts than the county-level models. This finding is derived from the fact that most of the effects of climate change were associated with between-system adaptation (changes from pasture to crops), and that this form of adaptation is better represented with the county-level model than the farm-level models.
Impact of Climate Change in the CUS Region under Alternative Climate Change Scenarios with Static and Dynamic Economic Models
Comparing the alternative climate change scenarios under static and dynamic versions of the county-level model for the CUS region, we found (figure 2):
- Climate change impacts vary substantially within and across ecoregions.
- Both the static and dynamic models show that the effects of the TP scenario are mixed, with many regions experiencing negative impacts, whereas under the TPC scenario impacts are generally positive except for the Southern Plains region.
- The static model tends to show more extreme negative impacts in the Southern Plains region under the TP scenario.
- Spatial variability in value of production is generally lower and more symmetrically distributed under the TPC scenario, with the exception of regions 9 and 12 which experience much higher spatial variability and more positively skewed distributions under the TPC scenario.
Model Coupling and System Dynamics
Implementing feedbacks from the economic model to the ecosystem model is conceptually straightforward but operationally very difficult due to operating system incompatibilities and other programming issues. An additional challenge is that simulations have to be executed for a large number of sites (in our case, about 1720 counties for the central U.S., about 450 farms in Montana and 200 in Nebraska). Moreover, the simulations have to be executed for a large number of possible combinations of land use and management to represent the management alternatives that exist. Therefore, an efficient computational approach is needed.
We investigated two strategies:
- Developing an interface between the Century model, as it is being run on a Unix-based parallel processing system at Colorado State University, and the economic models that are being run in the Windows environment using the Statistical Analysis System (SAS).
- Utilize the Tradeoff Analysis software developed by the project PI (Antle) in collaboration with Wageningen University in the Netherlands.
The second strategy was successfully implemented, and future work will utilize it to investigate closer coupling between ecosystem and economic models. Analysis investigating the effects of coupling led to the following results:
- Dynamics and heterogeneity are important to understanding the behavior of the system, and the insights afforded by models that capture system dynamics and heterogeneity may have important implications for areas that are most vulnerable to impacts of climate change and other environmental disruptions.
- In the CUS data, there are strong and statistically significant effects of lagged land use on present land use, and this type of a dynamic model does result in substantially different predictions of land use and production under climate change than with a static model.
- There are statistically significant but quantitatively small feedbacks from land use to production. This finding suggests that, if the ecosystem model’s representation of feedbacks is similar in magnitude to these empirical results, that model feedbacks may not be as important as the dynamics within the economic model itself.
- Incorporation of a feedback mechanism in the CUS model showed that feedbacks could substantially change the model predictions. We can conclude that feedbacks have potentially important impacts on simulated impacts of climate change. Further research will be required to quantify the magnitude of land use and management feedbacks and implement them with the coupled ecosystem and economic models.
Effect of the Rate of Climate Change on Estimates of Climate Change Impact
Using the dynamic version of the economic model for the central US, we investigated the effects of changing the temporal rate of climate change on estimated impacts of climate change. Some economists have argued that one of the most important issues is the rate of climate change, due to the impacts the rate of change would have on the ability to adapt to climate change. We found that there is a positive effect on climate change impacts from reducing the rate of climate change, and an adverse effect of an accelerated rate of change (figure 3). Thus, our findings to confirm the hypothesis that a higher rate of change would have adverse effects on production, due to the reduced ability for the agricultural sector to adapt.
Changes in Location and Quantity of Production
Our simulations for the central United States provide changes in the location of production and the quantities produced, by county. In the model baseline, we found that that the historical pattern of westward and northward movement of the Midwestern corn-soybean system would continue, and that there would be a reduction in pasture and in crop-fallow rotations in dryland grain crops. The simulations showed that climate change will augment this trend (figures 4 and 5 present results for corn and pasture).
Comparison to the U.S. Assessment
The results from the static and dynamic CUS models were compared to the results from the U.S. assessment as summarized in Reilly et al. (Climatic Change, 2003). The CUS models developed in this study generally predict the same pattern of changes in production, although it predicts smaller changes in the Lake States region and larger changes in the Corn Belt (figure 6). This difference is most likely explained by the fact that the model used in the U.S. assessment incorporates a down-sloping demand for U.S. grain production, and thus generates lower prices for corn and soybeans compared to the CUS model that fixes prices at 1997 levels.
Figure 1: Central 21 States of U.S. Ecoregion Grouping (Montana and Nebraska study areas highlighted).
Figure 2: Proportionate changes in value of production predicted by dynamic and static models, by CUS ecoregion, for TP and TPC scenarios.
Figure 3: Effect of changing the rate of climate change on predicted percentage change in value of production, TPC scenarios.
Figure 4: Acreage changes predicted by the dynamic CUS model, TPC scenario.
Figure 5: Pasture acreage changes predicted by the dynamic CUS model, TPC scenario.
Figure 6: Comparison of predicted percent changes in value of production by the US Assessment (Reilly et al, Climatic Change 2003) and the CUS dynamic and static models. (Data from Reilly et al for 2030 and 2090 averaged for 2050).
Journal Articles on this Report : 6 Displayed | Download in RIS Format
Other project views: | All 34 publications | 9 publications in selected types | All 8 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Antle JM, Capalbo SM, Elliott ET, Paustian KH. Adaptation, spatial heterogeneity, and the vulnerability of agricultural systems to climate change and CO2 fertilization: an integrated assessment approach. Climatic Change 2004;64(3):289-315. |
R828745 (2003) R828745 (Final) |
Exit |
|
Antle JM, Stoorvogel JJ. Incorporating systems dynamics and spatial heterogeneity in integrated assessment of agricultural production systems. Environment and Development Economics 2006;11(1):39-58. |
R828745 (2003) R828745 (Final) |
Exit Exit |
|
Antle J, Capalbo S, Mooney S, Elliott E, Paustian K. Spatial heterogeneity, contract design, and the efficiency of carbon sequestration policies for agriculture. Journal of Environmental Economics and Management 2003;46(2):231-250. |
R828745 (2002) R828745 (Final) |
not available |
|
Antle, J.M. and S.M. Capalbo (2002) “Agriculture as a Managed Ecosystem: Policy Implications.” Journal of Agricultural and Resource Economics 27(1):1-15. |
R828745 (Final) |
not available |
|
Hunt HW, Antle JM, Paustian K. False determinations of chaos in short noisy time series. Physica D: Nonlinear Phenomena 2003;180(1-2):115-127. |
R828745 (2002) R828745 (Final) |
not available |
|
Stoorvogel JJ, Antle JM, Crissman CC, Bowen W. The tradeoff analysis model: integrated bio-physical and economic modeling of agricultural production systems. Agricultural Systems 2004;80(1):43-66. |
R828745 (2003) R828745 (Final) |
Exit |
Supplemental Keywords:
Agriculture, adaptation, global climate, vulnerability, terrestrial ecosystem, central United States, economics, ecology, simulation, spatial scale, model coupling, dynamics,, RFA, Scientific Discipline, Air, Ecosystem Protection/Environmental Exposure & Risk, Ecology, Ecosystem/Assessment/Indicators, Ecosystem Protection, climate change, Ecological Effects - Environmental Exposure & Risk, Economics, Environmental Monitoring, Ecological Risk Assessment, Agronomy, Social Science, ecological exposure, anthropogenic stress, climate change impact, farming, human activities, meteorology, economic models, socioeconomic indicators, circulation model, agroeconomics, climate models, agriculture, environmental stressors, agro ecosystems, modeling ecological risk, ecological models, ecosystem sustainability, global warming, sensitivity, agriculture ecosystemsProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.