WALES lidar data from the PERCUSION field campaign for the Barbados period Version 1.2

In May 2024 the ESA/JAXA satellite mission EarthCARE was launched. The satellite combines for the first time a high spectral resolution lidar and a cloud radar with doppler capability as key instruments on one single platform. In addition, it is also equipped with a multi spectral imager and a broad band radiometer. This unique combination makes EarthCARE the most complex satellite mission to study aerosol, clouds and precipitation. To fully use these new and advanced data for science application needs validation of the measurements and data products. We have implemented an EarthCARE-like payload onboard the German research aircraft HALO (High Altitude and Long range) to prepare and validate the EarthCARE mission. This instrumentation was flown during PERCUSION (Persistent EarthCARE underflight studies of the ITCZ and organized convection) as contribution to ORCESTRA (Organized Convection and EarthCARE Studies over the Tropical Atlantic). ORCESTRA is a network of different campaigns and campaign components to better understand the organized tropical convection at the mesoscale, e.g. including the interaction of convective organization with tropical waves and air-sea interaction, and the impact of convective organization on the Earth’s climate and radiation budget. In addition, ORCESTRA helps to validate satellite remote sensing (especially EarthCARE). To achieve these objectives ORCESTRA combines several sub-campaigns taking place on the Cape Verde Islands and Barbados in August and September 2024. One of the campaigns combined in ORCESTRA is the PERCUSION campaign. It is a German initiative that uses the German research aircraft HALO and the ESA/JAXA satellite EarthCARE aiming to test factors hypothesized to influence the organization of deep maritime convection in the tropics, and the influence of convective organization on its larger-scale environment. One focus of PERCUSION was to establish confidence in the EarthCARE measurements and products. Thus, we included an EarthCARE underpass within each research flight to be used for validation purposes. HALO measurements were performed during the EarthCARE commissioning phase in August 2024 out of Sal, Cape Verde, and out of Barbados in September 2024. In addition, we performed flights out of Oberpfaffenhofen, Germany in November 2024 for validation of conditions that could not be captured in the two first campaign parts. Altogether, 33 EarthCARE underpasses were performed in different aerosol and cloud situations. Some of the flights were coordinated with in-situ measurements onboard other aircrafts (e.g. the French ATR42), with shipborne measurements onboard the German research vessel METEOR, or with ground-based radar and lidar measurements at Mindelo (Cape Verde), Barbados, and the ACTRIS stations Antikythera, Leipzig, Lindenberg and Munich. Additionally, to the EarthCARE validation measurements, four underpasses under NASA’s PACE mission were performed. This sub-dataset contains downward looking lidar profiles from the airborne demonstrator for the WAter vapour Lidar Experiment in Space (WALES) on HALO for the second part of the campaign operated out Barbados. The other parts of the dataset are published as separate zenodo datasets due to space limitations. The data set contains time series of profiles of backscatter ratio, particle depolarization, particle extinction and water vapour molecular density measured along the flight path of HALO. The first local research flight from Barbados was RF13. RF25 was the transfer flight from Barbados back to Oberpfaffenhofen, Germany. The table below lists the individual data products which were put into separate files. Please refer to the remarks section below for further details on how to interpret the data. Data Products bsrgl Backscatter ratio at 532 nm (g = green) for unpolarized detection, low sensitivity: clouds are much less often saturated, but less sensitivity to aerosol. Extinction corrected using HSRL method. tau2gl Two way atmospheric transmission due to particles at 532 nm. Molecular extinction is already corrected. bsri Backscatter ratio at 1064 nm (i = infrared) for unpolarized detection, extinction corrected using Klett’s method (no HSRL).Possibly high systematic errors for lager optical thickness of clouds. But highest sensitivity for background aerosol. bsriu Attenuated backscatter ratio at 1064 nm (i = infrared) for unpolarized detection, not corrected for aersol/cloud extinction. adepgl Aerosol-Depolarization at 532 nm from low-sensitivity channel. wv Water vapor molecular number density from 4 wavelength DIAL. lirgl Lidar ratio for 532 nm from the low-sensitivity channel. aextgl Aerosol extinction coefficient 532 nm from the low-sensitivity channel. Remarks All data, besides the water vapor product, are given on the same grid (1s x 15 m). For water vapor the horizontal grid spacing is reduced due to the necessary averaging during the processing. The water vapor, aerosol extinction and lidar-ratio products are given on a 15 m vertical grid like the other products, but the actual resolution is lower and given by the variable vertical_resolution. The corresponding averaging kernel is parabola shaped.  All data is re-gridded to a constant altitude scale above sea level. The actual flight altitude is given as a separate variable. All products contain an error variable indicating the statistical error of the main data product. To facilitate easy conversion to related quantities, airdensity and airtemperature fields from ECMWF IFS analyses interpolated in space and time to the same grid as the main data are included. The ‘description’ (not CF standard) attribute of each data field contains further information which might be valuable in interpreting the data. In contrast to older file-format versions for the WALES instrument the current format does not fill invalid or low-quality data with a fillvalue or NaN. The main data-field contains values as they are generated by the processing. To see, if a certain pixel is valid, you have to check the flags field. Only if flags is zero, the data should be considered usable. Different non-zero flag values indicate possible reasons of why the data is considered bad. Because of the last point, data filtering can now be refined or changed a posteriory: e.g. if you don’t want to  exclude saturated datapoints to generate a cloud mask you can select the good pixels by flags & ~0x0008 == 0 instead of flags == 0. Or maybe if you want to average, you should exclude low Signal to Noise Ratio (SNR) flags by testing for flags & ~0x0004 == 0 (flag codes and meanings are given in the attributes of the flags variable). Water vapor is given in units of the primary product of the DIAL as molecular number density. To convert to mixing ratio simply divide by the airdensity variable. There are a lot of processing parameters stored in the files to have a unified format also for processor internal purposes: simply ignore them…