Final Report Summary - MTVEGMOD (Advancing dynamic vegetation modeling for mountain systems vulnerable to climate and land-use change)
Project context and objectives
Assessing the impact of climate change on terrestrial ecosystems is confronted with many challenges, including those associated with the handling of scale and complexity. Ecosystem models used to evaluate vegetation response to climate change consist of either locally applicable forest gap models or globally generalised dynamic global vegetation models (DGVM). At regional scales, assumptions from local and global models are not easily scalable and few studies exist to evaluate the consequences of scaling assumptions on ecosystem vulnerability to climate. Climate change is also a complex issue that involves not only changes to temperature and precipitation, but also changes to land use, disturbance and atmospheric chemistry, in particular, ozone formation. Here, we designed an experiment to evaluate the challenges for modelling climate impacts in mountain systems using the dynamic global vegetation model approach.
Project methodology
Three tasks were defined for the project duration: Task 1 was concerned with the development of regional-scale plant functional types, Task 2 concentrated on developing a new module to represent a regional global change driver (in this case, ozone); and Task 3 implemented a series of dynamic numerical simulations of climate change and interactions with local drivers on ecosystem biogeography and biogeochemistry. A proposal was submitted to the TRY Global Plant Traits database requesting data on shrub and grass species relevant to the Yunnan region of China, where the model simulations took place (Task 1). An ozone module was implemented to the Lund-Potsdam-Jena (LPJ) DGVM based on the concept that uptake of ozone is a function of stomatal conductance, and that damage to photosystems and productivity is related to ozone uptake (Task 2). A downscaling technique was applied to Fourth Assessment Report (AR4) models of the Intergovernmental Panel on Climate Change (IPCC) to provide 1 km resolution temperature, precipitation, and cloud-cover forcing (Task 3).
Project results
Most of the AR4 IPCC climate projections agree that both temperature and precipitation will increase. This suggests that vegetation will move toward higher elevations as well as becoming more productive in the mid-elevational belt, increasing the uptake of ozone and countering the effects of elevated CO2. Task 1 enabled us to access data for regional-model species fitting, but missing data for the shrubs of regional interest prevented detailed species parameterisation and we opted for a sensitivity-analysis approach. The ozone module was successfully implemented and calibrated with observations of ozone damage and productivity. In comparison to previous work with different models, LPJ produced consistent ozone-damage responses. Regional simulations with climate and ozone forcing produced upslope vegetation shifts and increases in productivity, but the effect of ozone on plant production was greatest in the midrange of species distributions (coinciding with the optimal growth conditions and highest stomatal conductance).
Conclusion
Our work illustrates that regional climate impacts can be assessed with global vegetation models that are updated to include forcings of local significance, whereas the parameterisation of regional species traits remains challenging. The downscaling of coarse-scale climate data to local scales is relatively well founded, but adaptation of model drivers to handle large file sizes, and parallel computer processing, is critical. The research developed here also suggests that monitoring of contemporary climate effects should expand from treeline sites to mid-range sites because these will be most sensitive to indirect global-change forcing, such as ozone. Complementary research is recommended to support databases of plant structure and environmental response.
Assessing the impact of climate change on terrestrial ecosystems is confronted with many challenges, including those associated with the handling of scale and complexity. Ecosystem models used to evaluate vegetation response to climate change consist of either locally applicable forest gap models or globally generalised dynamic global vegetation models (DGVM). At regional scales, assumptions from local and global models are not easily scalable and few studies exist to evaluate the consequences of scaling assumptions on ecosystem vulnerability to climate. Climate change is also a complex issue that involves not only changes to temperature and precipitation, but also changes to land use, disturbance and atmospheric chemistry, in particular, ozone formation. Here, we designed an experiment to evaluate the challenges for modelling climate impacts in mountain systems using the dynamic global vegetation model approach.
Project methodology
Three tasks were defined for the project duration: Task 1 was concerned with the development of regional-scale plant functional types, Task 2 concentrated on developing a new module to represent a regional global change driver (in this case, ozone); and Task 3 implemented a series of dynamic numerical simulations of climate change and interactions with local drivers on ecosystem biogeography and biogeochemistry. A proposal was submitted to the TRY Global Plant Traits database requesting data on shrub and grass species relevant to the Yunnan region of China, where the model simulations took place (Task 1). An ozone module was implemented to the Lund-Potsdam-Jena (LPJ) DGVM based on the concept that uptake of ozone is a function of stomatal conductance, and that damage to photosystems and productivity is related to ozone uptake (Task 2). A downscaling technique was applied to Fourth Assessment Report (AR4) models of the Intergovernmental Panel on Climate Change (IPCC) to provide 1 km resolution temperature, precipitation, and cloud-cover forcing (Task 3).
Project results
Most of the AR4 IPCC climate projections agree that both temperature and precipitation will increase. This suggests that vegetation will move toward higher elevations as well as becoming more productive in the mid-elevational belt, increasing the uptake of ozone and countering the effects of elevated CO2. Task 1 enabled us to access data for regional-model species fitting, but missing data for the shrubs of regional interest prevented detailed species parameterisation and we opted for a sensitivity-analysis approach. The ozone module was successfully implemented and calibrated with observations of ozone damage and productivity. In comparison to previous work with different models, LPJ produced consistent ozone-damage responses. Regional simulations with climate and ozone forcing produced upslope vegetation shifts and increases in productivity, but the effect of ozone on plant production was greatest in the midrange of species distributions (coinciding with the optimal growth conditions and highest stomatal conductance).
Conclusion
Our work illustrates that regional climate impacts can be assessed with global vegetation models that are updated to include forcings of local significance, whereas the parameterisation of regional species traits remains challenging. The downscaling of coarse-scale climate data to local scales is relatively well founded, but adaptation of model drivers to handle large file sizes, and parallel computer processing, is critical. The research developed here also suggests that monitoring of contemporary climate effects should expand from treeline sites to mid-range sites because these will be most sensitive to indirect global-change forcing, such as ozone. Complementary research is recommended to support databases of plant structure and environmental response.