OPERA - Harmonizing the European Weather Radar Network

The operational weather radar network in Europe covers more than 30 countries and contains close to 200 weather radars. The radar network is heterogeneous both in hardware, signal processing, and scanning strategy, thus making it fundamentally different from the NEXRAD network, for example.

The Network of European Meteorological Services (EUMETNET) initiated the Operational Programme for the Exchange of Weather Radar Information (OPERA) in the 1990's to promote exchange of measurement data and knowledge. Since then, progress has been made in harmonization of hardware, measurements and file formats. The network-wide radar mosaic produced operationally since 2011 demonstrates progress of the harmonization efforts. The current programme phase, OPERA 4, covers years 2013-2018.

An essential part of the OPERA work is the documentation of the members’ best practices in radar siting, operation and data production. OPERA work has also involved development of algorithms as well as evaluating and reviewing existing ones. Further, joint recommendations have been authored for frequency protection and mitigation of wind farm interference.

After 2012, almost 20 of 30 OPERA members have upgraded their national radar network partially or fully to dual polarization. The dual-pol parameters are not yet exchanged, but several surveys have been executed about the best settings and of the use of the data by various stakeholders. The survey about maintenance practices revealed that it is not the radars that cause the most unavailability of radar data, but power failure and data communication failure. As to radar, the most frequent problems are due to the transmitter or the antenna controller, and consequent waiting time until spare parts arrive. One of the expected benefits of dual polarization in C-band is improved attenuation correction. However, according to a recent survey, out of 22 members with dual pol radars, only five members are using a dual-polarization based attenuation correction schema. Likewise for quantitative precipitation estimation (QPE), dual-polarization parameters are not yet routinely utilized by most of the members.

An important section of joint development in OPERA 4 has been development and assessment of tools for quality monitoring. The work package on using the Sun for radar calibration produced guidelines for determining an antenna pointing bias and stability and calibration of the receiver. Another recent OPERA study involved computing long-time accumulations of the European composite, revealing processing discrepancies and quality problems in data, but also improvement in overall quality after introduction of quality control procedures. Also a work package on comparing radar data to gauges is underway.

As to quality control, OPERA has also involved work on testing and evaluating the algorithms developed by the members. The idea is to test these algorithms in a central environment using data from across Europe, i.e. from radars of varying characteristics and settings (wavelength, scan schedule, etc.) and physical environment (climate, terrain, proximity to sea, etc.). This will yield information about the robustness and general applicability of the algorithms, leading ultimately to developing new algorithms that incorporate the best parts of the various algorithms tested.

For beam blocking, an algorithm using reflectivity and a digital elevation model has been studied, suggesting that a part of the underestimation observed in operational volumes can be corrected. Nevertheless, a better elimination of non-meteorological echoes needs to be performed. A study also on this topic is ongoing, comparing two algorithms. Vertical profile of reflectivity (VPR) is the largest single source of bias in radar-based QPE in mid-latitudes. The current OPERA rain rate composite is uncorrected and hence severely underestimated. An upcoming work package will address this issue and test solutions for correcting the composite product.

By creating a composite image of the European radars forecasters are able to track synoptic systems, cold and warm fronts. The European composite is founded on a very dense network of radars; the mean distance between radars is 126 km (80 miles). Hence, measurements often overlap and measure the same rain element with different radar which allows a better knowledge of cloud properties. In OPERA 4, a quality-based compositing (mosaicking) is under development, and is expected to yield European rainfall products with improved quality, with information about the local quality.

Open data, open code and cloud computing are global trends also affecting the radar community. Special attention has been paid on generality of algorithms as well as on portability and modularity of the developed program code. The produced composites are available for users under ECOMET licencing policy for official duty, research and education or commercial use.

Radar data is increasingly applied in numerical weather prediction (NWP) models for assimilation or verification. The OPERA User Group has contributed to this by supporting usage of radar data in assimilation and by communicating the needs of NWP community to OPERA.

The next OPERA stage 5, aimed for 2019 – 2023, is under planning in EUMETNET.