Atmospheric chemical processing drives the removal of emitted pollutants, and leads to the formation of ozone and secondary aerosol, which are harmful to human and environmental health, and contribute to climate forcing. Quantitative understanding of such chemistry is essential for the accurate prediction of current air quality and future atmospheric composition. In the troposphere, gaseous chemical processing is critically dependent upon the abundance of nitrogen oxides (NOx, NO + NO2), which regulate atmospheric oxidising capacity, ozone formation and the major components of many aerosol particles. Globally, the dominant NOx sources are all continental (traffic, power generation, industry, soil emissions of NO); these are well understood in some locations, but are very uncertain and rapidly increasing in developing nations, particularly African megacities. Once in the atmosphere, NOx is converted to reservoir compounds such as PAN, which may release NOx after transport, and ultimately into nitric acid (HNO3) on timescales of days. Current understanding is that HNO3 is the final atmospheric sink for NOx, and is removed from the atmosphere by deposition. Consequently, at remote marine sites a number of days transit time from the coast, we would expect NOx levels to be very low, and the inorganic nitrogen budget to be dominated by unreactive transported HNO3. Recent observations challenge this understanding: surprisingly high levels of NOx species, and HONO (a NOx precursor with a lifetime of a few minutes) have been observed over the tropical Atlantic ocean. This points to a missing source of HONO and NOx. It has been hypothesised that the photolytic conversion of particle-bound nitrate to gaseous HONO and NO2 may account for these observations and form the missing NOx source - a mechanism termed "renoxification" (Ye et al., Nature 2016). We have performed proof-of-concept measurements and modelling of HONO and NOx levels at the Cape Verde observatory in the tropical Atlantic, which we have found to be consistent with this mechanism (Reed et al., ACP, 2017) - however, order of magnitude uncertainties over the rate and products of particle nitrate photolysis remain, and observational evidence for its occurrence on dominant aerosol species (dust, sulfate aerosol) is missing, meaning that impacts on the global-scale are unknown. This project aims to address these uncertainties, through integrating existing ground-based, aircraft and satellite observations with targeted new field and laboratory studies. We will focus upon a natural laboratory, the tropical Atlantic region where we will probe the emissions and evolution of nitrogen species in the outflow of polluted air from the developing regions of West Africa to the clean marine environment of the mid Atlantic (Cape Verde Observatory, CVO). Specifically, we will (1) Use the tropical Atlantic as a natural laboratory to study renoxification during different seasons and aerosol regimes, alongside laboratory studies to parameterise the particulate nitrate photolysis process; (2) integrate this new understanding into a global chemistry-transport model to evaluate the recycling and transformations of NOx during transport, and hence the impacts of these process in the tropical Atlantic ocean, and upon our understanding of atmospheric chemical processing globally.

Life on land depends upon freshwater. Mountains act as water towers, producing water by lifting moist air, and by providing temporary surface and below-ground storage of water for later release into rivers. These stores are particularly important in regions that experience seasonal droughts, as snow and ice melt can counteract reduced rainfall during dry spells. Two main natural depots of frozen water exist. Snow is a short-term store, delaying the release of water after snowfall on daily to seasonal timescales. Ice melt also releases water seasonally. However, glacier ice is a longer-term reservoir, storing water for decades to centuries. A similar behaviour can be observed in the non-frozen part of a mountain catchment. Stores such as wetlands, ponds and shallow below-ground flow provide short-term storage, while lakes and deeper groundwater show long-term release characteristics. The combination of these different processes determines the magnitude and behaviour of a mountain range's water tower function for the surrounding area. This is particularly important in the Andes, where some of the most important water towers of the globe are found. The human population in regions neighbouring the Andes depend on mountain water resources for drinking, food production and hydropower, as do animals and plant life. Unfortunately, human-induced climate change is altering the stores of water held in the Andes water towers. Greenhouse gas emissions mean that snow-bearing weather conditions are becoming less frequent, depleting the stocks of snow held in the mountains. The lack of replenishing snow, and increasing temperatures, are causing glaciers to lose the ice they store, retreating to the higher and colder portions of the mountains. In combination with climate change impacts on the rest of the catchment, this is contributing to water shortages across the Andes. Ongoing droughts are hitting high-population cities, where the concentration of people increases the demand for water. For example, the cities of Lima and Huaraz (Peru), La Paz (Bolivia) and Santiago (Chile), are all situated in catchments where snow and ice melt contribute to river flow. However, upstream rural areas, which are less adaptable to climate change, are often even more directly reliant upon snow and ice meltwater. This impacts irrigation for agriculture, stressing the food security of the region. To help manage these changes to water supplies, this project aims to achieve two things. The first is to provide better monitoring. The high altitudes of the Andes are poorly instrumented. To work out where and how fast conditions are changing, we will install more scientific instruments to measure snow, weather and river discharge. To contextualise the changes we can measure now, we need longer observational records extending back in time. Many glaciers have been retreating since 1850, leaving behind an imprint in the landscape which we will map. Using satellite imagery, we can track the retreat of these glaciers from the 1970s to their present position. We will also utilise records of past climate conditions, recorded by sailors in ships-log books and stored in the landscape in sediments. Our second goal is to project future changes, which requires computer models of climate, glacier and river processes. Such projections are required for policy makers, who need to be reliably informed of potential future change. We will combine state-of-the-art models, to simulate the changing water resources in ten Andean catchments. To assess the skill of our models at making predictions, we will test them against our observations of past conditions and current changes. Models that perform well at replicating observed conditions will be used to project a range of possible future climate scenarios. By combining these observational and model-based approaches, we will improve the approach to projecting water resource change, and help to inform water management plans.