Abstract/Summary

Aerosols are extremely important to monitor and predict given their impact on health, radiation and visibility. Chemical transport and numerical weather prediction (NWP) models are becoming increasingly sophisticated, providing detailed forecasts of aerosol dispersion and characteristics. Consequently, the need for model evaluation and data assimilation using appropriately detailed observations is high. Automatic lidar and ceilometers (ALC) observe attenuated backscatter which contains information on aerosols in the absence of hydrometeors. However, a forward operator is needed to directly relate ALC observations to forecast aerosol characteristics. In this thesis, a flexible and computationally cheap ALC aerosol forward operator (aerFO) is developed to estimate attenuated backscatter (βm) in clear-sky conditions. Vertical profiles of dry aerosol mass mixing ratio (m) and relative humidity are inputs with default, tuneable assumptions about aerosol constituents. Parameterisations initially estimate physical characteristics, then further estimate optical properties. An extinction enhancement factor (f(RH)) is used to represent the change in optical properties due to aerosol swelling. Large components of work involve evaluating the Met Office UK variable resolution regional NWP model (UKV) (1.5 km resolution) in London. Sensitivity of aerosol characteristics to estimates of aerosol optical properties is explored. Aerosol speciation, total number concentration and dry mean radius are critical and optical properties can vary greatly with wavelength. Evaluation of UKV output over London suggests insufficient mixing in the model can lead to large errors in near-surface βm and the NWP surface scheme is important in urban areas. Additionally, βm evaluation is highly dependent on ALC data quality. Given uncertainties in optimal ALC network design, spatial patterns in βm variability over Greater London were explored in the UKV and the Met Office research London Model (333 m resolution), then clustered to produce informative maps relevant to instrument placement. Spatial patterns were strongly related to orography, wind and relative emission locations.