Work package 2: Climate model development
The objective of this work package is to undertake collaborative work to share expertise in coupled climate model development
The objective will be to improve capability and enhance the quality of data underpinning climate services (see WP3) with specific interest in simulation of precipitation over Brazil / South America at seasonal timescales.
Work can cut across a range of modelling approaches from high-resolution relatively short-term (seasonal-decadal) predictions to increased-complexity (including biogeochemistry) decadal-centennial projections. Both Met Office and CPTEC/INPE have forecasting capabilities on shorter timescales (NWP and up to 1 month) but they are not in scope for this CSSP project. The unifying theme of WP2 is, therefore, on rainfall at seasonal scales either within a seasonal forecast context or in the context of how long-term climate change may affect seasonal scale precipitation.
Similarly code optimisation and scalability for improved HPC exploitation are important activities but out of scope in this project.
Since 1995 CPTEC/INPE began operational forecasting in the form of ensemble seasonal prediction for the next season. North-east Brazil is one of the regions of the globe with a high level of predictability in seasonal forecasts - the Met Office publishes seasonal forecasts for this region. There is INPE expertise in tropical atmospheric dynamics and Earth System Modelling (especially land surface, vegetation, fire and biomass burning aerosols). Biogeochemical forcings and feedbacks play a leading order role in tropical climate change, but important processes such as vegetation dynamics, tropical wetland processes, wildfires and aerosols are not well represented in models. As model complexity increases it is vital that reliable and process-based evaluation is developed to ensure models and their projections are as robust as possible
The strategic objective of the CPTEC / INPE for 2016-2019 is to be a world reference, at least in South America, in research and modeling of an integrated system. This would allow better planning of activities related to potable water availability, electric power, the rational use of water resources and food safety in Brazil with direct benefits to CSSP Brazil Work package 3.
In this work package we propose sharing experience and expertise in model development activities, including development of processes relevant to South America and process-based evaluation metrics which are vital for ensuring realistic behaviour of the coupled Earth System model. Understanding of, simulation of, and evaluation of precipitation over South America at seasonal timescales is a particular focus.
- Seasonal forecasting. Understanding the predictability of South America on seasonal-decadal timescales including links to large scale dynamics and modes (ENSO, Atlantic SSTs) and drivers of change such as GHGs, aerosols and land-use. Met Office can make existing seasonal forecast data available for analysis over this region. INPE may perform new seasonal forecasts at a resolution TBD. This will allow a joint analysis of initialised predictability of seasonal rainfall over Brazil.
- Investigation of the role of forcings on seasonal-scale precipitation, and changes in seasonal precipitation. Development of event-attribution techniques to look at the role of forcings (such as aerosol or land-use) and feedbacks (such as fire emissions, ET or precip-recycling) on seasonal scale precipitation.
- Model evaluation including coupled processes. Global model evaluation, but with focus on South American performance, particularly on precipitation. Improved metrics of performance will be developed beyond simple time-mean measures. Development of routine process-based evaluation metrics that affect precipitation, including: microphysics, clouds, aerosols (from emissions to cloud interactions), large-scale circulation, hydrology and evaporation (tie in to observation campaigns such as SAMBBA and inversion modelling)
These separate strands of predictability, drivers and evaluation research could be brought together to explore and analyse a specific event such as the 2014 Sao Paulo drought.
This research would contribute to capacity building in climate modelling especially simulation of precipitation which is seen as a crucial output. It would also build confidence in the model projections for this region to inform both mitigation and adaptation policies across timescales.