Use of remote sensing and in situ observations of the atmosphere in chemical transport models
Measuring and simulating atmospheric chemical compositions helps us to better understand the impact from long-range transport of different pollutants. Aerosols and tropospheric ozone are examples of pollutants with life times long enough to get transported between continents. Thus, observing and modelling this transport and its interaction with the surrounding atmosphere is vital to providing information about the fate of these pollutants. The main subject of this thesis is to make combined use of, and exploit the advantages of both sources of information. This is done in two ways; (i) by evaluating model results with the help of observations, and (ii) by combining observations with model results by data analysis/assimilation.
The thesis is divided into four studies, where the first study sets the stage by providing a methodology for evaluating lateral boundary conditions (LBCs) in a regional chemical transport model (CTM). The methodology is applied to ozone and carbon monoxide, and includes a direct evaluation of LBCs at the boundary of a CTM, as well as an indirect evaluation of a model run. The results show that a combined direct and indirect evaluation give a more complete picture of how well a given set of LBCs perform, compared to using only a direct or indirect comparison.
To prepare for using the methodology from study one with aerosols, we needed an aerosol optical model that can map aerosol concentration fields onto aerosol optical properties. Therefore, the second study of this thesis investigates the impact of implementing a newly developed model with a realistic description of aerosol optical properties. The results show that realistic aerosol descriptions in an aerosol optics model impact radiometric quantities and radiative forcing to the same degree as including or omitting aerosol dynamics in a CTM.
Further, the aim is also to derive LBCs based on CTM results constrained by satellite retrieved aerosol optical properties. This requires us to know how much information we can use from the retrievals to constrain model variables. The main question of the third study therefore is, how much information do extinction and backscattering measurements contain about the chemical composition of atmospheric aerosol? The information content of extinction and backscattering measurements was analysed with a singular-value decomposition and was shown to increase with the number of observations and decrease with the observation error. We also found that the model variable best constrained with these types of measurements, was PM10.
The last and fourth study invokes the new aerosol optics model and the information content analysis, to constrain LBCs obtained from a hemispheric version of the CTM MATCH by use of CALIPSO extinction retrievals at a wavelength of 532 nm. An indirect evaluation with ground based observations, shows a bias reduction for PM10.
Ground based observations
Chemical transport modelling