Blog Posts Kari T Korhonen Forestry

According to an article recently published in the esteemed science journal Nature, the annual harvested forest area in Europe has increased by 49 per cent from the comparison years of 2011–2015 to the years 2016–2018. The article claims that the amount of harvested biomass has increased by as much as 69 per cent during the same time period. According to the article, fellings have increased, in particular, on the Iberian Peninsula and in the Baltic and Nordic countries.

In Finland, the felling volumes have indeed increased during the periods under review due to global economic trends and the resulting growth in the demand of wood. However, according to statistics from the Natural Resources Institute Finland (Luke), the increase in felling volumes in both Finland and Europe is considerably less than what is stated in the article.

According to Luke’s statistics (, the removals amounted to an average of 64 million cubic metres of stem wood with bark in 2011–2015, and 74 million cubic metres in 2016–2018. There was an increase of 10 million cubic metres, or 15 per cent. In 2019, the felling volumes began to clearly decrease. Statistics from the Food and Agriculture Organization (FAO) of the United Nations show that in Europe (during the period under review in 28 EU member states), the harvested amounts were 463 and 488 million cubic metres of stem wood without bark during the corresponding time periods. The increase in the EU countries was thus 25 million cubic metres, or 5 per cent. In Sweden, felling volumes even decreased slightly during the period under review.

Removals in 2007–2019 in EU-28 and Finland. Source: and FAOSTAT.

Why is the Nature article overestimating the amount of harvesting to such an extent?

The researchers who wrote the article published in Nature used Hansen’s global forest cover maps, which are based on an interpretation of satellite imagery, in their estimate. The maps have not been made in an attempt to interpret felling, but the changes in canopy cover that have taken place in picture elements 30 m × 30 m in size.

In the past couple of decades, significant development has taken place in the Landsat satellite imagery material used, and the authors of the maps warn that maps depicting change should only be compared with each other with caution. The authors of the Nature article have not followed this instruction. In addition, it is clear that if the goal is to detect harvesting in the satellite imagery, this should be carried out as direct change detection, taking into account the changes in the satellite imagery material and the coverage of cloud-free images.

Nor does the article make it clear how forest damage has been detected and taken into account in the research. The article only states that an attempt has been made to distinguish forest fires from harvesting with the help of forest fire maps. All other damage, such as insect and storm damage in Central Europe, is presumably considered as harvesting in the research.

Why is the Nature article causing concern?

The authors of the article are researchers from the Joint Research Centre (JRC), operating under the European Commission. The researchers present as their conclusion that the increase in harvesting is unsustainable in terms of climate change mitigation and maintaining biodiversity. The researchers have used, among other things, statistics from EUROSTAT, the statistical office of the European Union, when attempting to convert the estimated harvested areas into biomass – but have forgotten the statistics when critically evaluating the results of their research. The statistics are not fully free of errors either but, for instance, in Finland, Sweden and the Baltic countries, the harvested volumes are also estimated in National Forest Inventories (NFIs), and their results are consistent with the statistics at least in Finland and Sweden.

The published article and the methods used is raising international criticism, for example International research teams are preparing responses that will draw attention to the defects in the article published in Nature and the sources of error in the conclusions drawn based on the research. This is an important part of the functioning of the scientific community and of the process of producing and assessing scientific information.


  1. dear Kari thanks for the comment, which I see only now (sorry the late reply).
    Beyond a general warning when comparing results of the paper with statistics (see my comment to Aleksi) , because the risk is comparing apple with oranges, here I just note that the fact that “The authors of the Nature article have not followed the instruction of the dataset developers” is utterly wrong. You probably took this info from other sources. As a matter of fact, the study closely followed the available documentation, including any existing warning on temporal inconsistencies of the original dataset. Of course, the study could cannot guess undocumented inconsistencies. A clarification on these points, including a discussion on other (partly legitimate) criticisms will be soon published in Nature

    1. Dear Giacomo
      thank you for your comment.
      It is interesting to see you writing that your annual harvesting estimates are not comparable with the reality. If so, you should have been more cautious in making conclusions. Reading your article indicates that you have read the warnings of the Global Forest Change map producers but concluded that the data are still usable for estimating trends. I am convinced that in due time the scientific community will publish the real reason for your findings. While waiting that I recommend that the European decision makers base their judgement on the official statistics of Eurostat and FAO.

      1. dear Kari
        thanks for your comments. Just few clarifications:
        – I did not say that our “annual harvesting estimates are not comparable with the reality”. They are/should be comparable with “reality”, I simply said that we detected clear-cuts (30x30m) only, while in many countries the majority of harvest comes from thinning or selective logging. Thus, a direct comparison with country statistics should be done with caution. While I think the abstract could have explained it better (and it was partly at the origin of the confusion), the fact that we detected clear-cuts is clearly explained in the text. Our results should be seen as a serious early warning to be further assessed, not as an attack to the EU forestry or to country statistics (as it was often wrongly misrepresented).
        – I do confirm that we have read the warnings of the Global Forest Change map producers and followed any available documentation. Many similar studies used the same method, none was criticized. More (much more!) on our rebuttal in few weeks in Nature.
        – “I recommend that the European decision makers base their judgement on the official statistics of Eurostat and FAO.”. I largely agree. Our study is a scientific paper, and should not be treated like an Impact Assessment. Results are not meant to feed directly an EU legislation (although they may inform the discussion).

  2. According to EU GHG inventory by DG Clima supported by JRC (Joint Research Centre) mean annual biomass losses were -122953.67 Kt C and -134233.58 Kt C for periods of 2011-2015 and 2016-2018, respectively for forest land. Therefore, statistics compiled by JRC to UNFCCC under convention [] show that biomass losses have increased by 9% with EU-28 between 2011-2015 and 2016-2018. Biomass losses include harvested biomass and that of natural mortality. According to Ceccherini et al. 2020 (JRC) increase in biomass loss has been 69% between periods of 2011-2015 and 2016-2018. There seems the be a substantial gap between official GHG statistics (9%) and the estimate by Ceccherini et al. 2020 (69%). Official GHG statistics have been based on national forest inventories and also on national harvesting statistics. While, the analysis of Ceccherini et al. 2020 builds on optical satellite images.

    1. dear Aleksi, thanks for the comment, which I see only now (sorry the late reply). Few quick reactions:
      1. The data you collect from the EU GHGI is not adequate for the comparison you suggest. First, not all MS report losses of C; second, losses in table 4A may mean several things, depending on the method used; third, if you make a simple conversion of official harvest stats into C, you realise that much more than 134 MtC is harvested in the EU (supporting my first two points)
      2. Ceccherini et al, as explained in the text and methods, detects clear-cuts, and misses most of small scale operations (thinnings, selective loggings). Furthermore, the method to convert areas into biomass means that our biomass data are fellings overbark. Overall, with simple calculations it can be shown that the paper captures less than half of what reported by statistics.
      Any comparison even as (%) of clear-cut felling overbark (paper) vs total removals underbark (stats) should be made with great caution! it risks to be an apple-to-orange comparison. The comparisons done in the paper (possibly not clearly enough!) were aimed to identify patterns and trends, not to compare absolute values. This does not eliminate possible discrepancies, but please be careful in the comparison! A clarification on these points will be soon published in Nature

  3. OK, forestry industry defense as usual.
    GFW clarify that 3300 square kilometers of forest is lost 2019 in Sweden. Swedish authorities claim that less than 2000 square kilometers are lost. This is the problem. Maybe, forest experts would like to explain, or rather defend this fact.
    Finnish and Swedish forest industry survive solely due to the carte blanche given by their authorities. But the truth is approaching.
    Bye, bye, devastating and destructive industry. Future belongs to sustainability. You better accomodate.

Leave a Reply

Your email address will not be published. Required fields are marked *