The PARADISO campaign: Description and first results

The PARADISO campaign (PAyerne RAdar Disdrometer and ISOtop) was a joint venture between the Environmental Remote Sensing Laboratory (LTE) of the Swiss Federal Institute of Technology in Lausanne, the Atmospheric Dynamics group of the Institute for Atmospheric and Climate Science of the Swiss Federal Institute of Technology in Zurich (ETHZ)and the Radar, Satellite and Nowcasting and Atmospheric Data Divisions of MeteoSwiss. The main objectives of the campaign were firstly to investigate aspects of the microphysics of clouds and precipitation, the precipitation cycle and the atmosphere dynamics and secondly to validate and improve several remote sensing calibration and retrieval techniques. The measurement period took place between end of March and beginning of July 2014 in the area surrounding Payerne, in the Swiss plateau. Several instruments were deployed specifically for the campaign: 2 X-band Doppler Polarimetric weather radars (DX50 from MeteoSwiss and MXPol from EPFL), a mobile wind profiler, 5 Parsivel disdrometers, a 2D video disdrometer and a cavity ring-down spectrometer. In addition to that, operational models and data from MeteoSwiss, such as data from the weather stations of the SwissMetNet network, the operational radio sounding at Payerne, the C-band weather radar network and the COSMO numerical weather prediction (NWP) model were collected. The instruments were deployed in 4 different measurement sites. The DX50 and one PARSIVEL disdrometer were set at the Payerne military airport, the MXPol, 2 PARSIVEL disdrometers , the video disdrometer and the spectrometer were set at the MeteoSwiss headquarters in Payerne, 4 km southeast of the airport. Another instrumented site, with the wind profiler and one disdrometer, was placed 9 km northeast at Avanches. Finally a disdrometer was set 22 km northeast of the Payerne airport, at Bellechasse, almost aligned with the instruments at Avanches. In this way a transect of 3 measurement locations, from southwest to northeast, the typical flow of precipitating systems in the plateau was monitored. The scanning strategy of the 2 radars consisted of 3 RHIs over each of the instrumented sites, a set of PPIs at different elevations and a vertically pointing scan, meant for the retrieval of drop size distribution DSD and differential reflectivity calibration, which were repeated every 5 minutes. During the RHIs and the vertically pointing scan the Doppler spectra was recorded. The scans were synchronized so that when one radar was performing the vertically pointing scan the other radar was performing the RHI over the radar site. The two radars would also simultaneously perform the RHI scans over the other 2 instrumented sites and the set of PPIs. This facilitates the inter-comparison of the data and allows the study of attenuation effects since both radars were measuring at the same instrumented sites from slightly different angles. A total of 26 days with significant rainfall or convection activity have been identified. Data from both X-band radars are available for 10 of them. At the moment, a thorough radar data quality analysis has been performed with satisfactory results. A comparison between the Doppler velocity from the radar system and the radial component obtained from the windprofiler has shown an excellent agreement. Also, a comparison between rainfall rate estimation from the DX50 radar (retrieved with a simple Z-R relation with reflectivity corrected for attenuation using a basic algorithm) and measurements on the ground from disdrometers and rain gauges has been performed. As expected, the results are extremely satisfactory when measurements are performed at low altitude (i.e. where the radar volume contains rain) and close to the radar system but the performance degrades with measurement altitude and distance due to uncertainties in the attenuation correction and the fact that the radar is measuring in a snow or melting snow region. Currently under way are studies to check the performance of an hydrometeor-based quantitative precipitation estimation algorithm, wet radome and precipitation-induced attenuation characterization and DSD retrieval techniques. Results of these studies as well as more details on the data quality and comparison between sensors will be provided during the conference.