Many countries are required to report emissions of long-lived greenhouse gasses (LLGHG) under international agreements or emissions trading schemes. In the near future it is hoped that such programmes will be extended to every nation on Earth under a global climate treaty, to curb the damaging effect these gasses have on the environment. Whilst national emissions are currently reported based on 'bottom-up' methods (in which emissions inventories are compiled by considering factors such as fossil fuel use), it is clear that accurate and independent verification of these estimates will be vital if such treaties are to be successful. Here I propose to address this issue through the development the first system that can determine emissions of all of the most damaging non-carbon dioxide LLGHGs: methane, nitrous oxide, chlorofluorocarbons (CFCs), hydrochlorofluorocarbons (HCFCs), hydrofluorocarbons (HFCs), halons, perfluorocarbons (PFCs) and sulfur hexafluoride (SF6). Together they account for one third of mankind's contribution to the 'greenhouse effect'. A new modelling system is proposed for the determination of LLGHG emissions using atmospheric measurements. The method will use two complementary models to maximise the information that can be extracted from the measurements by allowing emissions to be derived at high resolution close to monitoring locations and low resolution further from them. Because of the unique combination of models and techniques in this system it will be possible, for the first time, to derive emissions of all of the above LLGHGs simultaneously, allowing mutually-beneficial information to be shared between the fields being determined. The two most important non-CO2 LLGHGs, methane and nitrous oxide, will require special attention in the proposed 'inversion'. These gasses differ from the majority of LLGHGs in that they have large, but very uncertain, natural components. Their year-to-year variability has been the subject of much debate and their likely future growth is very poorly understood. To further our understanding of the behaviour of these gasses, I will employ new techniques in collaboration with partners at the Swiss Federal Institute for Forest, Snow and Landscape Research and Massachusetts Institute of Technology: parameter estimation of a state-of-the-art wetland methane emissions model, and the incorporation of the first high-frequency isotopologue measurements of nitrous oxide into the proposed model framework. The proposed research will provide a unique opportunity to answer one of the most important questions in modern atmospheric chemistry: what is the concentration and variability of the hydroxyl radical? This is a key question because the hydroxyl radical is the single largest 'cleaner' of pollutants from the atmosphere. The novel multi-species inversion outlined in this proposal will allow much tighter constraints to be placed on this quantity, providing vital information with which to test future models of atmospheric chemistry and transport, and to predict how the lifetimes of LLGHGs may change in the future. The outcomes of this proposal will be the most accurate and comprehensive emissions estimates of all non-CO2 LLGHGs currently available, an increase in our fundamental understanding of methane and nitrous oxide source partitioning and variability, and a greater knowledge of the behaviour of the hydroxyl radical. These aims are recognised as being particularly important to the evaluation of the UK Climate Change Act (2008), and as such DECC and DEFRA will provide letters of support. The understanding gained from the proposed work, and the techniques developed, will pave the way towards international treaty verification for these important greenhouse gasses.