Know more

Work Package Number 3

Work Package Title

Wood data integration and analysis

Activity Type

(e.g. Research, Training, Management, Communication, Dissemination…)

Research and Training

Participant Short Name

INRA

INTA

MADERA PLUS

Person-months per Participant:

3

9

3

Objectives

Integration and covariation analysis of complex wood data in joint model-species

Description of Work (possibly broken down into tasks), lead participant and role of participants and seconded staff

Leader of the WP: V. Segura, INRA

3.1     Variation and covariation mining in complex wood data within and across species (A. Martinez Meier, INTA). NIRS spectrum and microdensity profiles are complex data sets that contain much more information than that is used by conventional data analysis methods. Thanks to powerful statistical analysis methods, experienced seconded French and Argentinean staff will more completely explore the variability enclosed. 

3.2     Correlation of wood basic properties and wood hydraulic functions within and across species (J. Gyenge, INTA). Preliminary results suggest that the relationships between wood basic and hydraulic properties are complex, variable and poorly understood. Seconded French and Argentinean staff will take advantage of the unique data set that is going to be acquired in the frame of TOPWOOD to study these relationships.

3.3     Training to statistical methods for complex wood data analysis

3.3.1  Partial Triadic Analysis (J. P. Rossi, INRA). PTA has been recently used for the first time on microdensity profiles (Rossi, Jean-Pierre, Maxime Nardin, Martin Godefroid, Manuela Ruiz-Diaz, Anne-Sophie Sergent, Alejandro Martinez-Meier, Luc Pâques, et Philippe Rozenberg. 2014. « Dissecting the Space-Time Structure of Tree-Ring Datasets Using the Partial Triadic Analysis. » PLoS ONE 9 (9): e108332. doi:10.1371/journal.pone.0108332.). Seconded Argentinean staff will be trained to use this statistical method and will analyze an existing Pseudotsuga menziesii data set, then publish the results.

3.3.2  Wavelets analysis (P. Rozenberg, INRA). Seconded Argentinean staff will be trained to use this promising statistical method rarely employed with microdensity profiles.

3.4 Benefit for the wood industry of early assessment of standing tree wood quality, modeling wood properties using portable tools with standing tree data for segregation at the tree and stand level for industry supply (E. Merlo, MADERA PLUS). The seconded staff will take advantage of the many wood traits measured in several species in Spain, France and Argentina to demonstrate how the knowledge of the wood quality before and immediately after tree felling can improve the profitability of the forest and wood industrial chain. The results will be immediately disseminated to the wood industry in three countries.

Deliverables (brief description and month of delivery)

18. Task 3.1 Draft of an article about “mining variation in complex wood data” (A. Martinez-Meier, INTA)

19. Task 3.2 Draft of an article about “covariation of wood hydraulic functions and wood basic properties” (J. Gyenge, INTA)

20. Task 3.3.1 Draft of an article “exploring microdensity variation with Partial Triadic Analysis in Douglas-fir” (P. Rozenberg, INRA)

21. Task 3.3.2         Training session “Advanced methods for the analysis of microdensity profiles (Wavelets analysis, threshold method), practical work with R functions” (P. Rozenberg, INRA)

22. Task 3.4 Presentation of the results of the modeling of the industry supply from standing tree wod properties in the Wood Industry meetings (E. Merlo, MADERA PLUS)