QA4SM-AVTB-EXP

Starting point / motivation

The current standards for the validation of higher-level, biogeophysical data products from remote sensing satellites were established about 20 years ago. They are based on manual and effortful ground validation campaigns carried out by independent institutions around the world, and typically take about 1-3 years to complete after the initial product generation.

This makes it difficult to transfer validation procedures between products, versions, and groups, thus limiting inter-comparability. This is in sharp contrast to the production processes that were developed to industrial standards over the last ten years by the Copernicus and ESA CCI programmes.

For example, an operational production chain for soil moisture (SM) was recently established by Copernicus C3S and is now generating Intermediate Climate Data Records (ICDRs) in near-real-time (every ten days) and Thematic CDRs (TCDRs) every six months.

Contents and goals

The Quality Assurance for Soil Moisture (QA4SM) project aims at the standardisation and automation of quality control (QC) and independent validations. We propose to bring the known procedures together into a single concept, and to build an Automatic Validation Testbed (AVTB) for a systematic study towards an automatic validation service. Feasibility assessments shall be made at methodological and system levels.

Methods

This will be achieved by developing an executable architecture and prototyping core components of the AVTB, and using it for experiments with candidate verification & validation (V&V) techniques. Requirements are the scaling up from point (in situ) scale to 100s of meters (Sentinels) to 10s of kilometres (operational coarse resolution), and to have validation results traceable and reproducible from validation reports.

The concept shall support Copernicus, ESA, NASA and EUMETSAT missions and services (incl. Sentinel-1, ASCAT, SMOS, SMAP, ESA CCI, and Copernicus land and climate services) and global satellite assessments by WCRP GEWEX. Although applicable to biogeophysical variables in general, we will prove it for SM because of the maturity and variety of the validation methods available for this variable.

The AVTB shall run on cloud computing architectures and have web service interfaces for configuration, execution and monitoring of validation tasks. The used and generated data, artefacts and traceability information shall be under version control.

Expected results

The successful completion of the project will involve R&D in several specialisation areas and will require the overcoming of the differences in understanding of concepts and vocabulary between them. It will harness best practices of individual fields into a combined concept.

A toolbox of appropriate techniques for automatic V&V of retrieving systems and data products will be assembled and proved by the AVTB, which itself will be evaluated and evolved over time. The team and the EO community will acquire the right amount of knowledge and confidence so that a later development to an operational validation service will become feasible.

The team will strive for acceptance of the concept and AVTB by the international and European organisations, CEOS, Copernicus, ECMWF, ESA, European Commission, GCOS, NASA, WCRP GEWEX, and the national delegations.

Coordinator

AW Software und Technologie GmbH

Project partner

• EODC Earth Observation Data Centre for Water Resources Monitoring GmbH
• Vienna University of Technology Dep. of Geodesy and Geoinformation

AW Software und Technologie GmbH
Dr. Alexander Boresch
Mariahilfer Str. 47/3/1
A-1060 Vienna