Extending existing models to capture vegetation response to extreme weather events: the MODEXTREME project
Extreme weather events are combinations of meteorological drivers that exceed certain thresholds and occur with low frequency, negatively impacting human living conditions and economic systems including agriculture. The three-year European project MODEXTREME (started on November 1st, 2013) aims at improving the predictive capability of biophysical crop and grassland simulation models under extreme weather conditions (mainly high/low temperatures and water deficit/excess). Existing modelling solutions can be improved with re-usable software libraries to capture extreme weather impacts. Estimates from existing and new modelling solutions will be compared on a variety of datasets and evaluated with respect to medium-term trajectories of future climate (mid-21st century). This will be achieved via the multi-model platform for plant growth and development simulation BioMA (Biophysical Model Application) and will support short- and medium-term forecasts in Europe via the Monitoring Agricultural ResourceS (MARS) workflow of European Commission Joint Research Centre. Project results will also extend the toolbox for food security monitoring and early warning systems outside Europe (Argentina, Brazil, China, South Africa, and United States). This paper explores both conceptual and software challenges of the project.
Download full text in pdf format
G. Bellocchi, F. J. Villalobos, M. Donatelli, O. B. Christensen, O. Rojas, R. Confalonieri, I. N. Athanasiadis, I. Carpusca, C. O. Stöckle, Extending existing models to capture vegetation response to extreme weather events: the MODEXTREME project, 7th Intl. Congress on Env. Modelling and Software, pg. 2196-2202, 2014, International Environmental Modelling and Software Society (iEMSs).
You might also enjoy (View all publications)
- Learning long-term crop management strategies with CyclesGym
- Rapid turnover of sensor data to genetic evaluation for dairy cows in the cloud
- Mixing process-based and data-driven approaches in yield prediction