BRIX is a first international collaborative initiative to inter-compare a set of biomass retrieval algorithms for P-band full-polarimetric SAR sensors. The exercise will focus on airborne campaign data acquired in the context of the preparation of the BIOMASS mission. The inter-comparison of biomass retrieval algorithms is expected to contribute to the understanding of the different uncertainty contributors and help in improving the biomass retrieval from P-band full-polarimetric SAR data.

Introduction
 

Selected as ESA's seventh Earth Explorer in May 2013, the BIOMASS mission will provide crucial information about the state of our forests and how they are changing. The data will be used to improve our understanding of the role of forests in the global carbon cycle. To prepare for the mission, several groups have started to develop Above Ground Biomass (AGB) retrieval algorithms based on P-band full-polarimetric, interferometric airborne SAR data acquired in recent campaigns. To support these activities and to bring together the groups involved in this endeavour ESA organizes BRIX, an exercise to inter-compare and learn from different biomass retrieval algorithms. The results of the exercise are expected to point out the strengths and weaknesses, as well as communalities and discrepancies of the various ideas to retrieve biomass . in order to suggest and define ways for their further improvement.

Objectives

The exercise aims at bringing together scientists working on the  development  of biomass retrieval algorithms and to  invite them to generate  AGB products under well-defined conditions. The input data will be a set of P-band full-polarimetric airborne SAR data of various test sites of tropical forest over Gabon from the recently flown ESA AfriSAR campaign. A common and harmonised inter-comparison procedure will be agreed and followed by all the participants.
 

The BRIX Objectives are:

• Provide an objective, standardized comparison and assessment of biomass retrieval algorithms developed for the Biomass mission;
• Establish a platform to involve scientists in the development of the Biomass retrieval algorithms and to exchange ideas;
• Learn about the strengths and weaknesses, as well as communalities and discrepancies of the various ideas to retrieve biomass.

The objectives shall be achieved by making available standardised test cases (based on campaign data), inviting the scientific community to develop and apply retrieval algorithms based on this test case, and finally compare, evaluate and discuss the submitted results.

Expected Outcomes

• An objective comparison and discussion of different biomass retrieval algorithms that shall be published as a joint paper in a high ranking earth observation journal;
• A protocol and coordinated plan for inter-comparing and validating different products.

Scientists benefits

• Participants have an easy and direct access to BIOMASS campaign data. Laborious (non-scientific) tasks as reading data of different campaigns which is usually formatted differently is eliminated by providing all data in an identical, easy to access format;
• Participants will learn about their algorithm through an objective and fair inter-comparison based on identical boundary conditions (identical Level-1, training and auxiliary data).
• Participants will co-author a potentially highly cited paper;
• Participants, that have so far not been involved in the BIOMASS community, will get visibility in the BIOMASS community and at ESA. This is especially the case for young scientists and graduates.

The BIOMASS Level-2 Algorithms Test Bed

For the purpose of an objective algorithm evaluation ESA has set up the BIOMASS Level-2 Algorithm Test Bed. This test bed is a virtual machine run that:

1. Provides access to all Biomass campaign data in a unified format;
2. Includes software tools that allow to implement and run algorithms in common open source scientific languages (Python, Fortran, C/C++, etc.);
3. Makes available processing resources.

Participants will be invited to upload their code to the test bed (or develop it on the test bed) and run it on the test bed using the predefined campaign datasets.

This Test Bed shall not only facilitate the inter-comparison of different algorithms but also facilitate sharing code and ideas and foster the discussion between the participating groups.

Schedule and Deadlines

After the registration deadline, ESA will grant all participants access to the test bed.

Then, the Biomass Retrieval Algorithm Inter-comparison Exercise will be performed in three phases:

1. Application of the biomass retrieval algorithms: During the time between the registration deadline and the submission deadline participants can develop and/or upload their algorithms. Until the submission deadline, the participants will apply their biomass retrieval schemes on the test cases keeping the processing parameters constant. The results will be submitted for analysis (2nd phase) to ESA;
2. Analysis of the results: ESA will analyse the submitted results from all participants and will assess the inter-comparison outputs based on a set of predefined metrics. All the results will be announced to participants prior to the BRIX workshop;
3. BRIX workshop: The obtained results will be presented and discussed. Before the obtained results are made public (in a peer reviewed journal paper) participants will have the possibility to provide feedback and to discuss the results during the BRIX workshop.

In the framework of the exercise, a Workshops will be held:

• ESA/ESRIN, Frascati, Italy 30 May 2018.