ReOBS: a new approach to synthetize long-term multi-variable dataset and application to the SIRTA supersite
Abstract. A scientific approach is presented to aggregate and harmonize a set of sixty geophysical variables at hourly scale over a decade, and to allow multiannual and multi-variables studies combining atmospheric dynamics and thermodynamics, radiation, clouds and aerosols, from ground-based observations. Many datasets from ground-based observations are currently in use worldwide. They are very valuable because they contain complete and precise information due to their spatio-temporal co-localization over more than a decade. These dataset, in particular the synergy between different type of observations, are under-used because of their complexity and diversity due to calibration, quality control, treatment, format, temporal averaging, metadata, etc. Two main results are presented in this article: (1) a set of methods available for the community to robustly and reliably process ground-based data at a hourly time scale over a decade is described, and (2) a single netCDF file is provided based on the SIRTA supersite observations. This file contains approximately sixty geophysical variables (atmospheric and in-ground) hourly averaged over a decade for the longest variables. The netCDF file is available and easy to use for the community. In this article, observations are "re-analyzed". The prefix "re" refers to six main steps: calibration, quality control, treatment, hourly averaging, homogenization of the formats and associated metadata, and expertise on more than ten years of observations. In contrast, previous studies (i) took only some of these six steps into account for each variable, (ii) did not aggregate all variables together in a single file, and (iii) did not offer an hourly resolution for about sixty variables over a decade (for the longest variables). The approach described in this article can be applied to different supersites and to additional variables. The main implication of this work is that complex atmospheric observations are made readily available for scientists that are non-experts in measurements. Dataset from SIRTA observations can be downloaded on http://sirta.ipsl.fr/reobs.html (tab download, no password required) under DOI http://dx.doi.org/10.14768/4F63BAD4- E6AF-4101-AD5A-61D4A34620DE .