Observations, lab measurements, and modeling are the three central elements in atmospheric chemistry. Progress in atmospheric chemistry is often driven by the interactions between these components and they form an integrated whole. For example, observations may tell us that more of a chemical compound exists than previously thought possible. This will stimulate new modeling and laboratory studies which will increase our scientific understanding to a level where we can explain the observations.
Observation Field observations of chemical systems are essential to understanding atmospheric processes and determining the accuracy of models. Atmospheric chemistry measurements are long term to observe continuous trends or short term to observe smaller variations. In situ and remote measurements can be made using observatories, satellites, field stations, and laboratories. Routine observations of chemical composition show changes in atmospheric composition over time. Observatories such as the
Mauna Loa and mobile platforms such as aircraft ships and balloons (e.g. the UK's
Facility for Airborne Atmospheric Measurements) study chemical compositions and weather dynamics. An application of long term observations is the
Keeling Curve - a series of measurements from 1958 to today which show a steady rise in the concentration of
carbon dioxide (see also
ongoing measurements of atmospheric CO2). Observations of atmospheric composition are increasingly made by
satellites by passive and active
remote sensing with important instruments such as
GOME and
MOPITT giving a global picture of air pollution and chemistry. Surface observations have the advantage that they provide long term records at high time resolution but are limited in the vertical and horizontal space they provide observations from. Some surface based instruments e.g.
LIDAR can provide concentration profiles of chemical compounds and aerosols but are still restricted in the horizontal region they can cover. Many observations are available online in
Atmospheric Chemistry Observational Databases Laboratory studies Laboratory studies help understand the complex interactions from Earth’s systems that can be difficult to measure on a large scale. Experiments are performed in controlled environments, such as aerosol chambers, that allow for the individual evaluation of specific chemical reactions or the assessment of properties of a particular atmospheric constituent. A closely related subdiscipline is atmospheric
photochemistry, which quantifies the rate that molecules are split apart by sunlight, determines the resulting products, and obtains
thermodynamic data such as
Henry's law coefficients. Laboratory measurements are essential to understanding the sources and sinks of pollutants and naturally occurring compounds. Types of analysis that are of interest include both those on gas-phase reactions, as well as
heterogeneous reactions that are relevant to the formation and growth of
aerosols. Commonly used instruments to measure aerosols include ambient and
particulate air samplers,
scanning mobility particle sizers, and
mass spectrometers.
Modeling Models are essential tools for interpreting observational data, testing hypotheses about chemical reactions, and predicting future concentrations of atmospheric chemicals. To synthesize and test theoretical understanding of atmospheric chemistry, researchers commonly use computer models, such as
chemical transport models (CTMs). CTMs provide realistic descriptions of the three-dimensional transport and evolution of the atmosphere.
Atmospheric models can be seen as mathematical representations that replicate the behavior of the atmosphere. These numerical models solve the differential equations governing the concentrations of chemicals in the atmosphere. Depending on the complexity, these models can range from simple to highly detailed. Models can be zero-, one-, two-, or three-dimensional, each with various uses and advantages. Three-dimensional chemical transport models offer the most realistic simulations but require substantial computational resources. These models can be global e.g.
GCM, simulating the atmospheric conditions across the Earth, or regional, e.g.
RAMS focusing on specific areas with greater resolution. Global models typically have lower horizontal resolution and represent less complex chemical mechanisms but they cover a larger area, while regional models can represent a limited area with higher resolution and more detail. A major challenge in atmospheric modeling is balancing the number of chemical compounds and reactions included in the model with the accuracy of physical processes such as transport and mixing in the atmosphere. The two simplest types of models include box models and
puff models. For example,
box modeling is relatively simple and may include hundreds or even thousands of chemical reactions, but they typically use a very crude representation of atmospheric
mixed layer. Once the reactions are chosen, the code generator automatically constructs the
ordinary differential equations that describe their time evolution, greatly reducing the time and effort required for model construction. Differences between model prediction and real-world observations can arise from errors in model input parameters or flaws representations of processes in the model. Some input parameters like surface emissions are often less accurately quantified from observations compared to model results. The model can be improved by adjusting poorly known parameters to better match observed data. A formal method for applying these adjustments is through
Bayesian Optimization through an inverse modeling framework, where the results from the CTMs are inverted to optimize selected parameters. This approach has gained attention over the past decade as an effective method to interpret large amounts of data generate by models and observations from satellites. One important current trend is using atmospheric chemistry as part of
Earth system models. These models integrate atmospheric chemistry with other Earth system components, enabling the study of complex interactions between climate, atmospheric composition, and ecosystems. == Applications ==