Using a regional model as intermediate between in situ airborne measurements and Sentinel 5P δD column retrievals

The isotopic composition of water vapour is a natural tracer of moisture cycling and moist processes such as rain-out efficiency in the atmosphere. The abundancy of deuterium (denoted as δD) in atmospheric water vapour can be studied using a variety of observational platforms that have a wide range of spatial and temporal resolutions. Recently, high-resolution and high-frequency δD total column retrievals from the Sentinel 5P satellite have become available, and need validation from in situ measurements. While satellite-retrieved products of δD in water vapour allow to study δD variability on spatial scales of several 1000km every 12-24h, the few available in-situ water vapour measurements of δD are at a much higher temporal resolution (seconds) and cover a more limited spatial extent (10s of km). In this study, we present a methodology for comparing datasets with different scales to each other. Thereby, it is important to take the principal time and length scales of water vapour δD features into account. To this end, we use model simulations with the isotope-enabled weather prediction model COSMO_iso as an intermediate to bridge the scales between the newly developed retrieval of water isotopologues for the Sentinel 5P satellite, based on the University of Leicester Full Physics retrieval algorithm, and in-situ vertical profiles of δD from ultralight aircraft acquired during the L-WAIVE campaign in June 2019. We illustrate that the assessment of spatial and temporal δD correlated patterns in COSMO_iso can serve as a proxy for spatial representativeness, and as such guides towards an unbiased comparison of datasets. Overall, we demonstrate that the combination of in-situ measurements and COSMO_iso simulations with satellite-retrieved δD can help to better constrain vertical δD gradients and to understand the temporal evolution of large scale δD patterns and associated moist processes. From our findings, we also derive more general recommendations for future comparison studies of in-situ measurement, satellite products, and model-simulated δD.