Floods in the UK are often caused by heavy rainfall lasting from minutes to weeks. Houses in flat areas are particularly at risk: meeting the shortage of houses in the south-east requires building on these areas. To estimate the flood hazard risk in order to try to protect these buildings, accurate rainfall predictions are needed. However, the connection between rainfall and flooding is complicated, so that rainfall predictions must also say how likely rainfall is at any time - calculating the probability of rainfall. Extreme rainfalls caused devastating floods in Boscastle in 2004 and Lynmouth in 1952, but the causes and pattern of rainfall was different. Therefore, scientists also need to know what pattern of rainfall caused the flooding. This research aims to get good quality predictions of the probability of rainfall by combining advanced methods from statistics, the output from a new supercomputer model of the weather, and a new computer archive of exteme rainfalls going back to 1860, provided by a specialist company Hydro-GIS Ltd. It also aims to produce an automatic system for discovering the most likely pattern in the predicted rainfalls. The new prediction system and data will be freely available over the internet for use by the government and universities.

Biomass burning (BB) is a major dynamic of the earth-atmosphere system, emitting smoke pollutants in quantities that are highly variable in space and time. By transferring knowledge and results from NERC grant NER/Z/S/2001/01027 this NERC Knowledge Transfer (KT) Project will design, build and evaluate the first system for the global geostationary observation of BB emissions source strength by (i) adapting and implementing the fire detection and characterisation algorithms developed under the grant for use with the full suite of geostationary systems (currently 2 x Meteosat SEVIRI, 2 x GOES, and MTSAT; with the possibility of INSAT3D after launch), (ii) linking this to cloud-masking procedures developed in concert with the UK Meteorological Office, and (iii) running the resultant BB scene analysis on the Meteorological Office real-time feeds for these data. The proposal does not undertake new science, and so is ineligible for NERC Standard Grant funding, but will transfer knowledge to enable Meteorological Office to produce a unique and widely called for BB product and will transfer that product onward to project partners for use in forecasts of atmospheric state. Users of these forecast products will thus also receive the benefit of the KT. The output BB product will be synthesised to a uniform, consistent and validated datastream with quantified uncertainties that will be made available at the necessary timesteps for forecast purposes. Product specification and file characteristics will be informed by the requirements of the Meteorological Office and the European Centre for Medium Range Weather Forecasting (ECMWF). The BB product will be validated against simultaneous higher spatial resolution observations and will be incorporated into current procedures for forecasting of atmospheric state and the ascribing of causal mechanisms to noted changes in atmospheric trace gas, aerosol and CO2 concentrations.

There is currently a large effort in the development of general circulation model (GCM)-based seasonal to decadal prediction systems to provide climate forecasts. Such techniques are rather complex, technically challenging and still in their infancy. Any weather or climate forecast will be subject to three sources of uncertainty, namely observation uncertainty, the model-component of initial uncertainty, and model uncertainty over the forecast period. The aim of this proposal is to improve the reliability of extended range forecast of weather and climate, mainly focusing on the ocean component of the coupled system. We propose to develop and incorporate various tools based on stochastic physics to improve the reliability of forecasts focusing on a more accurate representation of ocean observations and model uncertainties. The individual impacts of the different developments on the reliability of the forecasts will be quantified to provide estimates of the different sources of uncertainties in the forecasts. The development of reliable extended range forecasts can be extremely beneficial with major economical and societal consequences.

Biomass burning (BB) and wildfires release huge quantities of particulates and trace gases into the atmosphere in amounts highly variable in space and time. Plume rise means these that under certain conditions these emissions can be injected into the atmosphere at heights far above the Earth surface, enhancing their long-range transport and altering their atmospheric chemistry, radiative budget, and air quality effects. Results from past project show that UK air quality can be signficantly affected by long-range transport of smoke from European and Russian wildires, and smoke from fires in Canada can be detected in air samples at DEFRA monitoring stations in e.g. Mace Head. Near real-time (NRT) atmospheric modelling and forecasting schemes aiming to realistically represent these aspects of the Earth system must include a high temporal resolution, non-retrospective source of BB emissions information - which generally comes from satellite Earth Obervation data. However, as discussed above, a fires smoke plumes buoyancy characteristics can strongly influence its atmospheric impact, and this is increasingly realised to be an important term to represent when modelling the long-range effects of wildfire smoke emissions. However, a lack of a priori information and, until recently, a directly-related EO observable, has meant that parameterisation of smoke plume injection height has received far less attention than has estimating the magnitude and variability of the smoke emissions. This KE Project will exploit the findings from two successful NERC research projects to provide major improvements to the current (ad hoc) prescription of wildfire smoke plume injection height in the prototype GMES UK/European atmospheric monitoring and forecasting scheme (the 'GMES Atmospheric Core Service', which is based on the world-leading integrated forecast system (IFS) of ECMWF in the UK and which is being desiged to provide the public, policy makers and downstream organisations with access to state-of-the-art atmospheric chemistry monitoring and forecasting data. The GACS serves a broad community of users, for example those involved in environmental policy development and policing, those delivering downstream services related to the health community (warning of increased asthma incidence during air pollution episodes), and those aiming to reduce public exposure to air pollution. We will work with Project Partners developing the GACS to exploit the research on plume height rise developed in NE/E016863/1 and the EO data processing procedures developed in NE/H00419X/1 to provide a much more realistic representation of smoke injection height in the GACS system; one that takes account of both fire and atmospheric characteristics such that the atmospheric transport of these emissions, including to the UK, can be better represented. The Project Partners are ECMWF, who lead GACS development in the UK and who operate the global model within which the plume rise scheme will be embedded, and Jülich Research Centre who are experts in the chemistry and transport of smoke emissions and who are a main partner in the GACS development. The KCL Environmental Research Group (KCL-ERG) are a 'down-stream' user of global atmospheric model output, funded by UK Government to provide regional air quality (AQ) monitoring and modelling, and this KE project will support them in starting to use the enhanced GACS outputs in their UK regional and London-wide AQ modelling schemes, in particular to take better take account of smoke-polluted air that is known to move into the UK from e.g. eastern Europe or western Russia, and which at present causes enhanced discrepancies between the AQ models and measurements (see DEFRA letter of support). All model outputs incorporating the new scheme will be made available freely through the GMES GACS system interface http://www.gmes-atmosphere.eu/ and for the UK region throught the online public interface www.londonair.org.uk/

Data assimilation is a method to combine numerical models with observations. It is used in all environmental sciences and essential to be able to simulate the real world, instead of a pure model world which has little to do with reality. With the increasing resolution of geophysical models both the size and the nonlinearity of these models increase. Also the number of observations increases and the observation operators, which connect the model variables to the observations, become more and more complex and nonlinear, like new satellite observations and radar observations in weather forecasting. Obviously, the data-assimilation methods have to fully allow for these nonlinearities. Present-day implementations in numerical weather prediction are all based in linearisations. For example, the (Ensemble) Kalman Filter assumes linear updates, and variational methods like 4DVar solve a weakly nonlinear problem through linear iterations. A further problem with variational methods is that error estimates are hard to obtain, and for highly nonlinear problems inaccurate. A few operational weather prediction centres have started experimenting with ensembles of 4DVar's. This has the potential of solving the nonlinearity problem, and at the same time provides an error estimate. Recently, the European Centre for Medium Range Weather Forecasts (ECMWF) started experimenting with ensembles of 4DVar solutions, generated by perturbing the observations, with very promising results. It is known from Kalman Filter (or rather Smoother) theory that when this ensemble is cycled through several data-assimilation cycles its spread will approximate the error covariance of the system. In that case, the ensemble is a sample from the correct distribution. However, for a strongly nonlinear system the Kalman filter theory does not hold, and it is unclear what the ensemble means, and there is a strong scientific and operational need to understand what these ensembles mean, and how we can improve them. On the other hand, it is well-known that we can represent the underlying distributions by a set of particles, i.e. a set of model states, in a so-called particle filter. Particle filters are fully nonlinear both in model evolution and analysis step. A fundamental problem, the so-called 'curse of dimensionality' has hampered their use in geoscience applications. Very recently a solution has been found by the PI that has the potential to revolutionize data assimilation in highly nonlinear geophysical systems (Van Leeuwen, 2010a; Van Leeuwen, 2010b). The latter paper describes applications to relatively simple (up to 1000-dimensional) highly nonlinear systems that previously needed hundreds to thousands of model integrations, and now only of the order of 20 model integrations. This research proposal explores the possibilities of combining 4DVar ensembles with ideas from Particle Filtering for the next generation numerical weather prediction. A simple and exciting idea is to use 4DVar solutions as particles in the Particle Filter, and this is one of the directions we will investigate. But we will also investigate other ways to generate 4DVar ensembles that are meaningful in nonlinear systems. A strong point is that we will have direct access to the operational ECMWF system, allowing us to efficiently operate between relatively simple academic models and the operational world.