Report on first methodological improvements on retrieval algorithms and HF radar network design. JERICO-NEXT WP3 Innovations in Technology and Methodology. Deliverable D3.3. Version 1.

Abstract

Deliverable 3.3 is based on the results of Subtasks 3.2.1 and 3.2.2, i.e. on research improvements on retrieval algorithms and data quality and network design, leading to the first recommendations on the implementation of improved techniques for the JERICO-NEXT HF radar network. The work builds on previous results of WP2 (D2.1) and WP5 (D5.13), and is synergic with developments within the Copernicus Marine Environment Monitoring Service (CMEMS) project INCREASE (Service Evolution 2016). The results focuson the following three main aspects.The basic set of QC tests defined in D5.13 and in the INCREASE deliverable D3.1(http://www.cmems-increase.eu/static/INCREASE_Report_D3.1.pdf) has been further analysed and improved. Work has been performed within an extended group including scientists from the HFradar European community as well as from the US IOOS and the Australian ACORN networks. Additional QC tests with respect to the basic set identified in D5.13 have been considered and recommendationson how to include them are provided. An in-depth discussion regarding implementation methods for the tests and threshold setting is included. A detailed study has been performed on the quality of velocity retrievals, their errors and mitigation in case of high environmental variability partially resolved by radar measurement. The specific case of highly variable shallow water depth and its effects on a phased array system (WERA) are considered, but the method is general and can be applied to other cases. Methods to quantify retrieval errors are recommended and the use of innovative QC tests and multi-parametric thresholds are investigated. Methodological guidelines for the design of HF radar networks are provided. The first step consists in mapping societal needs and relevant observed variables, in order to identify areas of major interest. The choice of site locations should then minimize errors such as GDOP (Geometric Dilution Of Precision) and maximize coverage. A collaboration with Task 3.7 on the use of Data Assimilation technologies as a basis for observing system experiments is presently carried out.