Climate Change and Tourism
David Viner and Bas Amelung

Climate change will have a range of direct impacts on the tourism industry by changing the environment of resorts (e.g., sea-level rise, temperature, etc.) and it will increase the vulnerability of the tourism industry to other environmental changes. There will also be a range of indirect impacts, for example: raising conflicts in water resources; health effects; impacts on the built environment; and detrimental impacts on the local environment. Tourism through increasing emissions (aviation emissions are estimated to be the fastest growing source) of greenhouse gases (GHGs) is in turn having an impact upon the climate system. Therefore, the introduction of GHG mitigation policies will impact on tourism. These interactions between climate change and tourism have to date not been examined on a large scale.

HadAT1: a new gridded radiosonde temperature product
Peter W. Thorne, David E. Parker, Simon F. B. Tett, Phil D. Jones, and Mark McCarthy

Recent studies highlight the degree of uncertainty in upper-air temperature records and the effect that this can have upon e.g. detection studies. It is imperative that we create multiple independent long-term homogeneous timeseries to fully investigate and understand this uncertainty. The Hadley Centre's current radiosonde-based global gridded upper-air product, HadRT2.1s, has major limitations regarding the selection of potential break-points and their treatment with MSU satellite based co-located series, limiting our corrections solely to post-1979. These corrections reduce the dataset's spatio-temporal consistency and independence from MSU. We aim to address these deficiencies in creating our new dataset: HadAT1.

New gridded daily rainfall data for the Sahel
Andrew Matthews, Tom Holt and Jean Palutikof

Rainfall in the Sahel, West Africa, is characterised by high interseasonal variability and has suffered frequent severe droughts since the late 1960s, suggesting a change in the rainfall regime (Hulme, 2001). In order to create a spatiially homogenous daily rainfall dataset suitable for unbiased statistical analysis, a gridded dataset has been created using stations in Burkina Faso, Mali, Niger and Senegal. A smoothing thin plate spline has been used to create a daily gridded rainfall series at a 1° resolution.

Assessment of HadRM3 Extreme Precipitation in the UK
Marie Ekström and Phil Jones

Climate change impacts on extreme rainfall in the United Kingdom (UK) present a major concern regarding their effects on water resources and hydrology. Climate model output gives some guide to the magnitude of future changes. However, the modelling involves many parameterisations of sub-grid processes. Validation of model estimates with observed data is therefore essential to assess the reliability of modelled rainfall extremes. This study compares return period magnitudes in rainfall based on the Hadley Centre regional model (HadRM3) data and on observed data for the period 1961-1990.

Statistically downscaling from an Earth System Model of Intermediate Complexity to reconstruct past climate gradients across the UK
Mark Hoar and Jean Palutikof

Earth System Models of Intermediate Complexity (EMICs) allow the simulation of climate over Glacial-Interglacial Cycles. To add finer detail to the coarse spatial resolution of these models, a statistical downscaling approach can be used. Based around the Louvain-la-Neuve EMIC, MoBidiC, we are developing a methodology which uses full three-dimensional General Circulation Models (GCMs) as an intermediate step in the downscaling. The first requirement is to select the GCM. Correlation matrices and hierarchical cluster analysis have been used to compare 21 PMIP GCMs and 6 sensitivity studies for a domain 38N13W to 65N30E for the present day, mid-Holocene and Last Glacial Maximum time slices. Downscaled information is required along four transects within this spatial domain. Along these, the GCMs have been compared with observed climate and proxy palaeoclimate data for the same time slices. This has been done using correlation matrices and root mean squared errors, with and without the linear trends of the transects removed. K-means cluster analysis is used to define the 'best' model or best group of models based on these statistics. The second step in the downscaling is to use an objective measure to fit a suitable polynomial to the 'best' curves obtained for each time slice. The polynomials can then be linked to the output of the Louvain EMIC to produce transects for any period during the last 123 kyr BP. Although this methodology has been developed for the Louvain models, it is hoped that it can be applied to the output from other EMICs and for long-term future climate change.

Regional Climate Model Simulations of Daily Maximum and Minimum Near-Surface Temperatures Across Europe 1961-1990: Comparisons with Observed Station Data
Anders Moberg and Phil D. Jones

High-resolution (~50 km) dynamic Regional Climate Models nested in Global Climate Models (GCM) are becoming an increasingly important tool in climate research. HadRM3H is the latest RCM from the Hadley Centre, with a resolution of 0.44° lat/lon. It is forced at its lateral boundaries by an atmosphere-only GCM (HadAM3H). Both models have 19 vertical layers. HadAM3H was run driven by observed 1961-90 sea surface temperatures and sea ice. Initial conditions were interpolated from a transient model run with a coupled ocean-atmosphere model, where greenhouse gases and aerosols were successively increased from 1860 to 1990. An ensemble of three members has been obtained for the 1961-90 period, by starting transient runs from different points in a long unperturbed control run.

Development of High Resolution Weather Scenarios for the EPSRC/UKCIP Initiative
Clare Goodess

In order to develop high-resolution climate change scenarios for key locations, this project will use computer-based weather generators and statistical models to produce common source datasets for the other projects in the programme. The basis of this additional data will be the UKCIP02 climate change scenarios, which will be developed for shorter time periods (daily and sub-daily) and higher spatial resolution (5 km grid boxes and point locations rather than a 50 x 50 km grid), focusing on climate variables which are of greatest concern to the built environment. The project will also address issues of scenario uncertainty and provide further information on potential changes in the "urban heat island" effect.

This project will act as a service to the other projects in the initiative and develop best practice in the application of climate change scenarios. At the end of the project, the new scenario data that is generated will be made more widely available.