Forest Change Survey through Hierarchical Model-Based Inference



Forests play an essential role for bio-economy and for mitigating climate change. Thus, assessments of state and change of forest biomass are becoming increasingly important. Recent developments in remote sensing techniques provide opportunities to decrease uncertainties in forest biomass assessments and also decrease the overall survey costs. Nowadays several types of remotely sensed (RS) data are acquired on annual, monthly and even weekly basis from many parts of the world. Whereas one source of RS data may not be suitable for all needs, fusion of the several data sources is often efficient. Recently introduced methods of estimating forest variables utilizing a fusion of RS data within, so called, model-based inference[1],[2] opens up possibilities for developing comprehensive statistical frameworks for forest state and change surveys in a cost-efficient manner. One such method is called hierarchical model-based (HMB) estimation. HMB estimation is proposed as a cost-efficient way of combining (e.g.): (i) wall-to-wall multispectral optical data; (ii) a discontinuous sample of laser data that are strongly correlated with forest structure; and (iii) a sparse sample of field data. Model predictions based on the strongly correlated RS data source are used for estimating a model linking the target quantity with the wall-to-wall optical RS data. With HMB uncertainties due to both modelling steps are accounted for to obtain reliable estimates. The method has been finalised for state assessment[2] of forest biomass. However, change assessment algorithms have yet to be developed. The main objective of the ForestChangeHMB project is to further develop the HMB framework so that it will be applicable for change estimation.

Goal: To develop novel statistical tools for large-scale surveys of change in forest biomass across time, by combining several sources of remote sensing data and field survey data.

Funded by: The Kempe Foundation (SMK-1847), FORMAS (FR-2019/0007), The Swedish National Space Agency (SNSA-171/19).

Duration: 2019 –


  • Methodological development of change estimation algorithms within the HMB inferential framework utilizing a fusion of RS data.
  • Testing the HMB change estimation method through case studies using field data collected in Sweden (national forest inventory data and existing RS data).
  • Comparing the HMB change estimation method through comparison with existing methods, such as Bayesian hierarchical modelling, and data assimilation based on the Kalman filter.


[1] Saarela, S., Holm, S., Grafström, A., Schnell, S., Næsset, E., Gregoire, T.G., Nelson, R.F. & Ståhl, G. (2016). Hierarchical model-based inference for forest inventory utilizing three sources of information, Annals of Forest Science, 73(4), 895-910.

[2] Saarela, S., Holm, S., Healey, S.P., Andersen, H.-E., Petersson, H., Prentius, W., Patterson, P.L., Næsset, E., Gregoire, T.G. & Ståhl, G. (2018). Generalized Hierarchical Model-Based Estimation for Aboveground Biomass Assessment Using GEDI and Landsat Data, Remote Sensing 10(11), 1832.