The biggest difference between many of the approaches is flexibility and empowerment. With SimCLIM and ArcGIS overlays, one can easily: 1] create customized scenarios for different years, emission scenarios and/or climate sensitivities? 2] examine the range of uncertainties in different combinations of GCM patterns, emission scenarios and sensitivities? 3] combine months into seasons particularly relevant for, say, agriculture or water supply? 4] analyze observed time-series station data? 5] generate scenarios of future sea-level rise which take account of global, regional and local factors and which are consistent with the scenarios of climate changes? 6] systematically perturb time-series data (monthly, daily or hourly) using those climate change patterns? estimate return periods for extreme climate events (e.g. heavy rainfall, hot days)? 7] estimate changes in return periods for climate extremes, with and without climate changes? 8] seamlessly run sectoral impact models (e.g. DHI water, DSSAT crop)? 9] have the option of attaching additional impact models which are driven by observed data and scenarios of change? update the system as new climate change patterns become available without retaining consultants?

What format are the outputs of PRECIS delivered in? With SimCLIM you can import from a number of formats (I know this question is raised elsewhere). For example, ArcGrid, Grass Ascii and Idrisi files. Shape files for things like roads and landcover etc can also be added to a SimCLIM system. More to this . . . SimCLIM uses a pattern scaling method similar to that used by MAGICC/SCENGEN. Thus prior to being imported into SimCLIM Precis outputs must be processed off-line in order to get change pattern (the difference between the future climate and control run climate)which then needs to be normalised to get the changes per degree of global warming. This is not a difficult thing to do and we have done it many times. We can assist with this.

With reference to DSSAT we have worked with Gerrit Hoogenboom to build a seamless interface with the DSSAT model system. Therefore a DSSAT user can import weather files from DSSAT into the SimCLIM system and perturb them using the various GCM and SRES drivers and then bringing the pertubed files back into DSSAT to run the model with perturbed weather files. In terms of MAGICC/SCENGEN we work closely with its developers and the GCM patterns we use are derived from it. The biggest difference between MAGICC/SCENGEN and SimCLIM is that SCENGEN is only a scenario generator which applies form global to regional scale (at a rather coarse resolution). In comparison SimCLIM is an integrated model that focuses from regional down to local scales and is essentially a \"tool kit\" that contains methods and models for analyzing observed climate, future climates and with links to impact models. Further to this SimCLIM, to a large extent, contains the results from MAGICC/SCENGEN, for generating regional climate scenarios. The SimCLIM team could further specify the data formats that SimCLIM works with.

As regards to the \"capacity constrained\" issue: The particular advantage of SimCLIM is that it is very \"user-friendly\". It was designed for end-users. So we have conducted many training sessions and are able build the capacity of people in small islands to use system within a few days. As regards the \"data-poor\" issue: Data is always a problem -- whether for SimCLIM or for any other model or system. SimCLIM can handle limited data situations -- one just has to be aware of the limitations and prepared to work within them. However, in our experience it is rare to find a place that does not have any data. Sometimes quite a bit of data can be derived from global datasets. One of the advantages SimCLIM has is its flexibility - it can facilitate analyses according to the availability of data (you do more detailed analyses if you have a lot of data, otherwise you can still perform crude, first-order assessments even if you only have limited datasets. Perhaps the SimCLIM team can further elaborate on this and on the capacity requirement aspect.

Yes, you are correct, the number of sectoral impact models that come with SimCLIM is limited. However, there are several ways in which this limitation is overcome: (1) we have special versions that link to well-known sectoral models -- e.g. SimCLIM-DHI links to the suite of Danish Hydraulic Institute hydrological models, and SimCLIM-DSSAT links to the DSSAT crop modelling system that is used world-wide. (2)SimCLIM has in-built spatial modelling tools (the \"map calculator\"). With this, one can manipulate spatial data with user-created instructions or equations. So, if one knew some basic relationships between climate variables and disease incidence, this could be modelled within SimCLIM directly and quickly; (3) at the request of countries, we can add to the impact model library by identify and linking \"off the shelf\" models; (4) country has impact models, we can build a custom \"plug-in\" to incorporate the model into the systemj; (5)if need be, we can work with the country to develop new impact models

Re extent of technical support: We have been providing on-going technical support for countries using SimCLIM and are committed to extending that support to new users. The extent of support varies from provision of software to full training and technical assistance. Re cost: depends on the extent of technical assistance and the circumstances of the end-users.

Nothing special as regards to CPU and/or hard-disk space. We recommend Pentium4-plus, especially if you have created a fine spatial resolution. It runs on Microsoft Windows systems.

The latter

TrainCLIM is different from SimCLIM, in the sense that it is \"hard-wired\" and you can\'t add data. SimCLIM is an \"open-framework\" system which accepts new data and which the user can build up as required. Usually this means that we \"pre-load\" SimCLIM with regional and local data pertaining to the user\'s geographical focus. Re the issue of global data (GCM) for smaller states: This is important point. From a \"top-down\" scientific point of view, there is reluctance to interpolate these data down to fine resolution without proper downscaling (either dynamic modelling or statistical). However, from \"bottom-up\" point of view, those concerned with impact and adaptation need something to work with -- the \"best-guess\". This is our primary approach. If proper downscaled data are not available, we use simple interpolations (or \"re-gridding\") to provide this \"best-guess\".

SimCLIM can be run either in single site or spatial mode. And the resolution of spatial analysis is variable: it depends on the baseline datasets (including climate and other sectoral variables) you can supply the model with. In other words, it can be as high as you wish providing you have sufficient baseline data. In addition, some downscaling/interpolation software packages can be applied to increase the resolution if needed. This is user defined. It depends on data availability, size of geographical area, and the purposes of the assessment. The important point is, using SimCLIM the spatial resolution can be defined or adjusted. \"Areas within areas\" can also be created with finer resolutions and interpolated data.

SimCLIM is an evolved version of VandaCLIM with a large number of new features and new components (e.g., impact modules). Re impact models: yes sectoral impact models (e..g, human health impact model) can be linked to SimCLIM. The software has an interface to allow users to link impact models through a registration process. No SimCLIM is an \"open-framework\" system which means you can customize it for your particular needs. VandaCLIM is like TrainCLIM in that it was hard-wired. The methods are similar between all the models. We have answered the second part of your questions elsewhere.

re hydrological models: The DHI models are not built into the system and have to be obtained separately under license from DHI. Once they are up and running on your computer, the SimCLM-DHI version will interface with them. Alternatively, there is a range of simpler hydrological models available. These can be run in conjunction with SimCLIM, using the time-series data and the SimCLIM scenario generator to perturb the time-series data for analyses under future change. Re weather generator: SimCLIM does not come with a weather generator. However, we do have code for several weather generators and can incorporate these into the system. Re cost: This is difficult to say. The cost depends on the types of models and analyses that are required.

Yes SimCLIM can be used to do integrated watershed analysis, through hydrological modules which can be linked into the system. Re local data requirements for sectoral impact analysis: ideally time series data can extend to a length of 50 years or longer to derive robust trend. But if you do not have such long time series, there are different techniques (e.g., using weather generator for climate data) that can be applied to overcome the limitation. Re integrated watershed analysis: The off-the-shelf version of SimCLIM cannot, itself, provide integrated watershed analysis. However, it can provide the platform on which to add models and data for this purpose. For example, we customized a SimCLIM for a watershed analysis in the Cook Islands, using a simple hydrological model, a function to simulate flooding, functions to simulate flood damages, and functions to assess the benefits and costs of adaptation options. Re data: depends on which sector and which models. The SimCLIM model accepts monthly, daily and hourly time-series data. At least 30 years is preferable, but the longer the better.

Yes the types of data required to build a customized SimClim system include: (i) digital elevation model (whole or part of the island(s), time series data - rainfall, temperatures, rel. humidity, wind, sunshine and sea-level data. You will also need data on agriculture (soils, agriculture census data, water resources - e.g. groundwater data, catchment data, catchment roof areas, water demand statistics, health statistics, diseases (VECTOR- AND WATER-BORNE), fisheries data, etc. or any other biophysical system that is of interest in terms v&a assessment. Re degree of training: you could get proficient after a couple of days training, and, with a week, get a full training with some applications under the belt. Re costs: depends on the number of participants, where, length of time, etc. But not costly compared to other professional training courses available.

This depends on the extent of engagement. If, for example, you just wish the software, the process is: you license it, we pre-load existing data and data provided by you, and we send the package to you. If, however, you wish associated training and perhaps technical assistance, then a ToR needs to be prepared and a contract signed. How long will be the startup process take? It largely depends on data availability. If there are no publicly available data for your country and it has to be obtained from in-country sources, there could be delays depending on security and confidentiality issues. For example, with one small island country in the Pacific with whom we are working, it is taking a least a month to sort out data acquisition through government. When the customised SimCLIM is built usually about 2-3 weeks depending very much on supply and availability of data)it will be used in an in-country training of national experts. A two-week (10-days) intensive training will not only introduce to the experts the SimCLIM system but also facilitate V&A assessment and technical back-up support. The timing of training will depend on the circumstances within the country and how fast data can be made available.

The limitations of SimCLIM are not really documented. If you mean \"limitations\" as in model limitations, then it has to be kept in mind that SimCLIM is not so much a \"model\" but rather a \"modeling system\" that contains arrays of data, models and tools for performing a large number of spatial and temporal analyses using observed and future scenarios of climate change. Thus, the \"limitations\" pertain more to the quality of data and models that are incorporated into the system. Having said that, the core of SimCLIM -- the scenario generators -- use the pattern-scaling approach. The major limitation of this approach is the assumption that the patterns of climate change remain constant over different forcings and time periods. This assumption is valid for most GCM runs, but one can detect changes in the patterns for others. Also, with the current version of SimCLIM, you cannot do \"ensembles\" of GCMs (averages of multiple GCM runs),but rather the outputs have to be examined individually, GCM by GCM. The next version of SimCLIM will have this capacity.

Data Requirements for SimCLIM (1) spatially-interpolated monthly climatologies (we usually preload SiMCLIM with these data): SimCLIM needs spatially-interpolated, mean-monthly values, based on a suitably long period of record (for example, the 1961-90 30-year record), for the following variables: minimum temperature, maximum temperature, mean temperature and precipitation. Thus, you need to provide 48 image files (4 variables x 12 months). If you think that you may eventually need other climate variables, like radiation, they can also be included. By \"spatially interpolated\", I mean that the data should be like maps. That is, there should be a defined geographical area (all of your country? Selected regions of the country? It is your choice). The data should be interpolated to an appropriate spatial resolution (the spatial resolution refers the grid size to which the data have been interpolated). (2)time-series climate data for selected stations (monthly and/or daily data)(We can pre-load some data and assist with formatting of data that is provided): Historical time-series data, along with their coordinates and elevation. The length of record does not matter, but, generally speaking, the longer the better of course. For some stations you might only have limited variables (for example, only precipitation). This is OK; we can just put whatever you\'ve got into the database. (3)downscaled, \"standardized\" patterns of monthly temperature (minimum, maximum and mean) and precipitation changes from GCMs: The SimCLIM will be preloaded with an initial set of GCM patterns. (4) digital elevation model (optional): if you think that elevation data will eventually be needed, also send these data for the same area and at the same spatial resolution.

The SimCLIM Open Framework System is a software product within the CLIMsystems Ltd range of climate simulation and impact model tools. SimCLIM is designed to support decision making in a wide range of situations where climate and climate change pose risk and uncertainty. A user customised SimCLIM Open Framework System software package has the capacity to assess baseline climates and current variability and extremes. Risks can be assessed both currently and in the future. Adaptation measures can be tested for present day conditions and under future scenarios of climate change and variability. With the program, users can conduct sensitivity analysis and examine sectoral impacts of climate change. The SimCLIM Open Framework System supports integrated impact analysis at various scales. The SimCLIM Open Framework System software provides an easy-to-navigate windows platform for creating a wide range of climate change scenarios using the latest scientific knowledge. The software is unique as it is the first of its type to provide users with the capacity in one software program to examine a range of climate parameters. For example, past climate data can be queried through an extreme event analysis tool which can, among other things, determine the probability of a particular extreme event, such as heavy rainfall or extremely hot or cold temperatures. The probabilities and return periods for such extreme events can also be queried for the future using a wide array of future scenarios of climate change as released by the Intergovernmental Panel on Climate Change (IPCC). This type of information is vital to engineers designing infrastructure to withstand future climatic events and a whole host of other individuals and organisations, such as the insurance industry, that must factor in potential climatic risk when planning for developments that will persist into a climate changed future. The software includes options for importing your own local data through the use of a data importing wizard. The software also incorporates the latest available Intergovernmental Panel on Climate Change (IPCC) greenhouse-gas emission scenarios marked by fundamentally different assumptions about fossil fuel consumption, demographic change and economic structures. A total of six emission scenarios (SRES) are included in the package and they can be queried for their associated changes in temperature, sea-level rise (total), sea-level rise (thermal expansion only) and CO2 concentration as produced by the MAGICC/SCENGEN model for IPCC. For each SRES the program produces a graph of projections from 1990 to 2100 with low, medium and high estimates. These same scenarios can be used to explore the likelihood of extreme events and future changes in sea level and coastal erosion. A water balance impact model is also included in the software package. SimCLIM is referred to as an Open Framework System owing to its flexibility. Individual users can easily customise the package through the importation of local data and development of customised impact models that can be easily attached to SimCLIM. Here are a few weblinks to research and projects associated with SimCLIM: Climate Proofing: A Risk-based Approach to Adaptation http://www.adb.org/Documents/Reports/Climate-Proofing/default.asp New Developments of the SimCLIM Model for Simulating Adaptation to Risks Arising from Climate. Variability and Change. http://mssanz.org.au/modsim05/papers/warrick.pdf SimCLIM: Recent developments of an integrated model for multi-scale, risk-based assessments of climate change impacts and adaptation http://www.anzsee.org/anzsee2007papers/Abstracts/Warrick.Richard.pdf Climate Change Modelling: The South East Queensland (SEQ) Regional Climate Change Adaptation Demonstration Project http://www.seqcatchments.com.au/climate.htm