Interpol-IAGOS software

With their multi-decadal timescale and their geographical coverage reaching the hemispheric scale, the IAGOS measurements are an essential tool for assessing long-term simulations from chemistryclimate/ chemistry-transport models. The data are mostly gathered in the upper troposphere – lower stratosphere, a key region regarding the impact of greenhouse gases like water vapour and ozone on surface temperature, and also regarding the exchanges between the troposphere and the stratosphere. However, the IAGOS data set has still never been used in its whole ensemble to assess long-term model simulations, notably because they are not gridded and their spatio-temporal resolution is far different from the simulations outputs. For instance, in the case of the reference simulation with specified dynamics (REF-C1SD) from the Chemistry-Climate Model Initiative (CCMI) experiment, these outputs are characterized by a monthly resolution and a ~ 1 km vertical resolution at cruise altitudes. The methodology used for the Interpol-IAGOS software (https://doi.org/10.25326/81) is presented in detail in Cohen et al., 2020 (to be submitted). It consists in distributing IAGOS data on a model grid (daily or monthly), following a weighted reverse interpolation, then in deriving gridded monthly means for any available observed variable in a first step. New variables can be derived from the observed ones, like the O3/CO ratio or potential temperature. The generated data set is then named IAGOS-DM-model, where the DM acronym stands for « Distributed on the Model grid » and the « model » suffix is to be replaced by the model name. Other metrics are also calculated in order to characterize better the monthly means, as standard deviation, extreme values and an equivalent amount of involved measurement points. If the model’s potential vorticity field is available, the IAGOS data projection can be done with respect to different atmospheric layers. With the IAGOS-DM files, there are also the fields from the simulation after the application of a mask that follows the IAGOS sampling. In a second step, the final outputs from the Interpol-IAGOS software consist in monthly time series in several regions and optionally in several layers, and mean seasonal (and yearly) climatologies on every vertical grid level. All the output files are in the NetCDF format.