Wood-based materials encompass a broad range of products, including engineered wood, pulp and paper products as well as biocomposites derived from cellulose. These materials are attractive in many ways but the predictability of their properties is obstructed by their complex and hierarchical material structure with features on several length scales that affect the properties, for instance its combination of nanometer and micrometer sized pores [1]. Another challenge is the diversity and variability of the wood material. To understand and improve important processing steps and properties such as moisture transport, delignification upon pulping or the effect moisture has on mechanical properties, a characterization and modeling framework that considers mechanisms spanning different length scales and strategies to take diversity and variability into account must be employed [2]. Here, an approach to build such a framework is suggested in which computational modeling is applied for modelling the material at different hierarchical length scales. Outcome from molecular dynamics simulations at atomistic scale is combined with finite element multiphysics continuum models to capture macroscopic properties such as moisture diffusivity, delignification mechanisms and hygromechanics.