Workshop on Estimation with the RDBES data model (WKRDB-EST; outputs from 2019 meeting).

Type Article
Date 2020
Language English
Author(s) ICES
Contributor(s) Campbell Andrew, Chen Chun, Christman Mary, Clarke Liz, Currie David, Dubroca Laurent, Fernandes Ana Claudia, Fuglebakk Edvin, Gerritsen Hans, Hommik Kristiina, Lankov Ain, Leijzer Twan, Meitern Richard, Mirny Zuzanna, Molla Gazi Karolina, Spegys Marojus, Stoerera Sven, Suska Marta, Teruel Gomez Josefina, Thasitis Ioannis, Vanelslander Bart, Wischnewski Julia, Zarauz Lucia
Source ICES Scientific Reports/Rapports scientifiques du CIEM (2618-1371) (ICES), 2020 , Vol. 2 , N. 5 , P. 106p.
DOI 10.17895/ices.pub.5956
Abstract

The RDBES is the new Regional DataBase and Estimation System. The RDBES is expected to replace the previous RDB and InterCatch by the end of 2021 and will bring significant improve-ments and transparency in the provision of estimates from commercial fisheries to stock assess-ment and other end-users. The developments of the RDBES meet the EU-MAP requirements of progress towards statistically sound sampling schemes. The RDBES data model and associated database are able to store, among other, sampling data alongside the elements required to de-scribe the sampling design used in data collection. Upload of data to and estimation within the RDBES will require significant adaptation of the data collection processes of national institutes in several areas, including data storage, but also sampling design, field protocols, estimation and data provision to end-users. To secure a soft transition there is a need to intensify internal plan-ning of these adaptations from 2020.

The Workshop on Estimation with the RDBES data model (WKRDB-EST) prepared data for 8 of the 13 upper hierarchies of the RDBES and developed a first set of R-scripts that handles design-based estimation in the RDBES data model. Developments and tests were positive and confirmed the usefulness of the data model for design-based estimation. These developments are publically available in the ICES GitHub (https://GitHub.com/ices-eg/WK_RDBES/tree/master/WKRDB-EST). The RDBES core group will continue the development and produce an R-package that ag-gregates a) a generic set of estimation functions and b) vignettes documenting design-based es-timation in each type of sampling hierarchy.

WKRDB-EST examined and tested version v.1.17 of the RDBES model with feedback being ob-tained from 15 countries on 8 of the 13 upper hierarchies of the RDBES. The data model can now be considered relatively stable with mostly minor issues being identified. The RDBES core group will discuss these issues and incorporate in a future data model, v.1.18.

Finally, WKRDB-EST discussed the way forward in the development of the estimation compo-nent of RDBES. It was agreed that the priority for 2020 should be finalizing the development of design-based estimators. That development should include domain estimation and post-stratifi-cation since these aspects are necessary to produce estimates at the spatial and temporal resolu-tion required by a variety of end-users. Development of script for model-assisted and model-based estimation based on the RDBES format should take place in other fora (e.g. SC-RDB-coor-dinated data workshops, WKs spawned by the ICES Working Group on Commercial Catches (WGCATCH)). A new WKRDB-EST will be suggested to SC-RDB for late 2020 where the devel-opments of design-based estimation will be finalized.

Full Text
File Pages Size Access
Publisher's official version 112 13 MB Open access
Top of the page