An Introduction to the PRECIS system
PRECIS was developed in order to help generate high-resolution climate change information for as many regions of the world as possible. The intention is to make PRECIS freely available to groups of developing countries in order that they may develop climate change scenarios at national centres of excellence, simultaneously building capacity and drawing on local climatological expertise. These scenarios can be used in impact, vulnerability and adaptation studies, and to aid in the preparation of National Communications, as required under Articles 4.1, 4.8 and 12.1 of the United Nations Framework Convention on Climate Change (UNFCCC).
- The PRECIS Handbook (PDF, 1 MB) - Generating High Resolution Climate Change Scenarios using PRECIS - April 2004. This handbook was published jointly by the UNDP and the Met Office Hadley Centre.
- PRECIS Regional Climate Modelling System (PDF, 104 kB) - Climate Impacts Model Capability
Why use a regional climate model?
A regional climate model (RCM) is a downscaling tool that adds fine scale (high resolution) information to the large-scale projections of a global general circulation model (GCM). GCMs are typically run with horizontal scales of 300km. RCMs can resolve features down to 50km or less. This makes for a more accurate representation of many surface features, such as complex mountain topographies and coastlines. It also allows small islands and peninsulas to be represented realistically, whereas in a global model their size (relative to the model grid box) would mean their climate would be that of the surrounding ocean.
RCMs are full climate models, and as such are physically based. They represent most if not all of the processes, interactions and feedbacks between climate system components represented in GCMs. They produce a comprehensive set of output data over the model domain. There are three types of technique for obtaining regional climate change projections: statistical, dynamical and hybrid (statistical-dynamical) techniques. RCMs fall into the dynamical category.
Formulation of the PRECIS model
Annexes I & II of the PRECIS handbook describes the (very similar) scientific formulations of the PRECIS regional model and HadAM3P, the model that provides the default lateral boundary conditions (LBCs). Both are based on the atmospheric component of the Met Office Hadley Centre's coupled climate model, HadCM3.
HadAM3P is a global atmosphere-only model with a resolution of order 150km, forced by surface boundary conditions (sea-surface temperature and sea-ice fraction) from HadCM3 and observations. It has been run for two "time slices": 1960-1990 and 2070-2100. The Met Office Hadley Centre is running a suite of climate change experiments, sampling a range of scenarios from the IPCC Special Report on Emissions Scenarios (SRES), using HadAM3P. See Obtaining PRECIS for more information about which experiments are available to drive PRECIS. Other modelling centre data is also available as input to PRECIS, including the Max Planck institute (Germany)'s ECHAM4 and ECHAM5, the ECMWF ERA-40 reanalysis, and the NCEP (USA) R2 reanalysis data.
What hardware and OS is required to run PRECIS?
PRECIS has been developed to run on PCs under 32-bit Intel (x86) compatible Linux-based systems. No particular Linux operating system is recommended; several distributions of SuSE, RedHat and Mandrake/Mandriva have been tested successfully. Dual-boot PCs can be used, but bear in mind that the Windows partition will be unavailable unless the experiment is interrupted and the machine re-booted. PRECIS's control scripts require the public domain Korn shell (pdksh) to be installed, which is not always done as part of a default Linux installation. However, it isn't necessary to use the Korn shell as the interactive shell.
At least 512MB physical memory is required, and 1GB or more is recommended. The difference in performance is small, but the smaller memory imposes a (possibly undesirable) upper limit on the size of the domain that can be modelled. Processor speeds of 2.0GHz or greater are recommended - in general, simulation speed is proportional to chip speed.
Depending on the output options chosen, a typical 30-year PRECIS experiment generates between 130 and 540GB of output data, whilst the boundary conditions for a thirty-year experiment occupy 40GB. A hard disk size of at least 100GB is therefore recommended, although with careful data management by the user, smaller disks can be used. It is recommended that DLT or DAT tapes are used for back-up and permanent storage of the output data. The preferred medium for supplying boundary data is via an IDE hard drive, typically 500GB.
The standard distribution of PRECIS, which includes sample boundary driving data, is on a dual layer DVD, so a DVD-ROM is required.
If power supplies are unreliable, PRECIS can be shut down cleanly (ready for a re-start) given approximately a 1 hour notice. If shut down without warning, the model can be re-started from the beginning of the model month in progress, which might involve repeating up to typically twelve (real) hours of simulation. A UPS (uninterruptible power supply) may therefore be appropriate if power outages are frequent. For more information, please see Chapter 2 of the PRECIS technical manual (PDF, 1 MB) .
How fast is PRECIS?
A typical experiment, covering a 100-by-100 grid box domain and including a representation of the atmospheric sulphur-cycle, run on a 1 core, takes 2.5 months to complete a 30-year simulation. With 4 cores it will take about 3 weeks and 8 cores about 2 weeks.
What output does PRECIS produce?
A comprehensive set of variables has been chosen for PRECIS's default output - for a listing, please see Appendix C of the PRECIS technical manual (PDF, 1 MB) . The user can specify the levels on which upper-air variables are output. Variables are automatically output as "climate" means (monthly, seasonal, annual, decadal etc.). Daily and hourly means can be selected if required.
If a particular variable is not included in the standard list or an existing variable on a different timescale is require, it is easy (with help from the Met Office Hadley Centre) to reconfigure PRECIS so as to include it. Similarly, custom meaning periods or time series of instantaneous values may be available.
The output data is in the Met Office's own binary data format (PP format). Various tools are supplied with PRECIS to process it. It can easily be converted into both GRIB and NetCDF formats, but not vice versa. PP format is the recommended output data format.
How are boundary conditions supplied?
Boundary condition data are supplied on an IDE hard drive of about 500GB. See Obtaining PRECIS for more information.
Is a Windows version of PRECIS available?
There are no plans to produce a Windows version of PRECIS.
Is a Macintosh version of PRECIS available?
Though Macintosh OS X is Unix based, PRECIS has not been compiled to support anything other than Intel or AMD processors. Thus, PRECIS is not supported on older Macintosh computers . This includes any Linux distribution for PowerPC. Now that Apple is using Intel processors, there's no reason why Precis wouldn't run on Linux running on an Intel Macintosh, although we haven't verified that it works.
Can we have the PRECIS source code?
The PRECIS source code is included with the standard distribution. Compilers will be required in order to produce new executables.
Can PRECIS be parallelized or run on a cluster?
The current version of PRECIS runs on either a single processor or on shared memory multi-core cluster systems in parallel MPP mode. We recommend that spare processing capacity be used to run separate experiments, such as baseline and climate change, or different ensemble members for a given scenario, simultaneously.
PRECIS will not run as installed on cluster systems using scheduling, queuing or job control software unless an Expert Scientific Software Engineer is sent to your institute at a cost. Please email precis for more details.
It could be much easier and cheaper running PRECIS on a multi-core desktop rather than a dedicated cluster.