Met Office climate models have made a significant input to the international CMIP5 modelling project for IPCC
Every 5-6 years the Intergovernmental Panel for Climate Change (IPCC) publishes a report into the most up to date understanding of climate science, including an assessment of the physical science, vulnerability of socio-economic and natural systems and mitigation options. Met Office climate research has always played a key role in informing this process and the next report, the Fifth Assessment Report (AR5), due out in late 2013 will be no different with several Met Office scientists playing major roles as lead authors and coordinating lead authors for the report. One crucial aspect of the assessment report is in making projections of future change which draw on state-of-the-art climate models.
The 5th Coupled Model Intercomparison Project (CMIP5) is an internationally coordinated activity to perform climate model simulations for a common set of experiments across all the world's major climate modelling centres. Over the last two years Met Office climate models have been used to perform all of the CMIP5 experiments and deliver the results to a publicly available database. The CMIP5 modelling exercise involved many more experiments and many more model-years of simulation than previous CMIP projects, and has been referred to as "the moon-shot of climate modelling" by Gerry Meehl, a senior member of the international steering committee, WGCM. Throughout the entire CMIP5 process, Met Office scientists have been instrumental in informing the experimental design, and performing the required experiments. The Met Office has been the first centre to start producing data and to have finished all the experiments using our state-of-the-art "Earth System" model, HadGEM2-ES.
HadGEM2-ES is a configuration of the Unified Model (MetUM), and is the Earth System version of our family of climate models HadGEM2. The tag "Earth System" refers to the fact that it represents many more processes than just the physical atmospheric and oceanic processes commonly included in General Circulation Models (GCMs). HadGEM2-ES has, at its core, the atmospheric and ocean components of the MetUM but also includes representations of the global carbon cycle, dynamic vegetation, atmospheric chemistry and ocean biology. This means instead of using pre-determined inputs of atmospheric composition such as aerosols and greenhouse gases, the model can simulate how these components change over time in response to anthropogenic activity and changing climate conditions. More than a decade ago Met Office research was the first to show that coupling the global carbon cycle to a climate model could significantly accelerate climate change due to the "climate-carbon cycle feedback". This research, published in Nature in 2000 by Peter Cox and colleagues has now been cited more than 1,000 times and established the Met Office as world leaders in Earth System science.
The Met Office Hadley Centre remains at the forefront of Earth System research.
HadGEM2-ES was the focus of an editorial in the journal Nature in 2010 and was described as "...possibly the world's most sophisticated climate model". The diagram shows the complex web of interactions represented in HadGEM2-ES - making it more complex than any other model that has been used for all of the CMIP5 experiments. Not only does the atmospheric composition affect the weather and climate, but interactions between the components are also important, such as dust emissions from bare soil which act as a source of iron, an important micro-nutrient, to oceanic plankton which in turn can emit DMS - an aerosol pre-cursor which acts to form cloud condensation nuclei and thus affect the Earth's radiation balance and vegetation growth. The model is also able to represent direct human influence over the land surface, through deforestation, as well as through emissions of CO2 and other greenhouse gases.
The climate simulations with HadGEM2-ES will help to understand key climate and biogeochemical processes and also to answer questions about the future such as what climate change might look like under un-checked emissions of CO2, or what emissions reductions would be required to achieve a given climate "target" to avoid dangerous climate change. Relative size of data provided to IPCC 4th and 5th assessment reports Enlarge
In addition to performing climate projections for the 21st century with HadGEM2-ES, CMIP5 for the first time included experiments to make near-term climate predictions. In this context, "predictions" differ from "projections" in that they attempt to capture natural climate variations as well as the long-term average response to human influence. The Met Office is also a world-leading centre for decadal predictions, and for these near-term simulations we used the HadCM3 climate model initialised with observations of the ocean and atmosphere state. This has already been shown to give improved forecasts of global temperature and Atlantic hurricane frequency, and there are signs that skilful predictions of rainfall and drought might be possible in some regions including parts of Africa, America, India and Europe.
The sheer number and diversity of experiments has meant that this activity has required huge effort from across all our science and IT areas. About 10,000 model years have been simulated and more than 165,000 data files have been processed and sent to the "Earth System Grid" - the distributed database that will make the CMIP5 results from all modelling centres available for researchers around the world to analyse.
All of the hard work developing and running our models would count for nothing if we were unable to deliver our data in a timely manner in the required format to the end users. This in itself has been a huge undertaking, and has required world-leading IT support and infrastructure every bit as much as the modelling has required world-leading science expertise. To date we have delivered 45 TB of data - or around 90 times the amount of data we sent for the equivalent exercise for the last IPCC report (AR4), with up to 5 TB still to come over the next few months. We are very grateful to colleagues at the British Atmospheric Data Centre (BADC) for their help and support in hosting this data.
The HadGEM2 model is described in these publications:
The CMIP5 simulations with it are documented here:
The decadal prediction system is described here:
The Nature News Feature on HadGEM2 can be seen here: