Incorrect figures on harvested forests in Nature article

Nyhet - 03 July 2020

The scientific journal Nature have published a paper reporting on dramatically increased levels of forest harvesting, which is not consistent with Swedish statistics and is based on dubious methods.

In a scientific paper in the journal Nature, a research group report that the area of harvested forest in Europe increased dramatically after year 2015. They report, among other things, that the harvested area within the EU has increased by 49% during the period 2016-2018, compared with the period 2011-2015, and that Sweden and Finland contribute substantially to this development.

This reported dramatic increase in forest harvesting is not consistent with Sweden’s national statistics. On the contrary, statistics from The Swedish Forest Agency and the Swedish University of Agricultural Sciences show that the harvested forest area has decreased within the studied time period.

"When we compare the period 2011-2015 with 2016-2018, we calculate that the area harvested annually has decreased by approximately 8 per cent" says Jonas Fridman, Head of Programme of The Swedish National Forest Inventory at the Swedish University of Agricultural Sciences.

According to the National Forest Inventory, a survey conducted in the field with objective statistical methods, the area of harvested forest in Sweden has been around 200 000 hectares per year during the past decade, while the volume of harvested wood has increased steadily during the same period.

Both the volume of harvested wood and the Swedish wood consumption have increased by 4% on average annually, comparing the period 2011-2015 with 2016-2018. This is far from the dramatic increase reported in Nature, according to Jonas Paulsson, statistician at The Swedish Forest Agency.

Several problems in the methodology

The article in Nature has been written by a group of researchers at the EU Joint Research Centre (JRC) in Ispra, Italy. They have used global maps over harvested forests, which have been compiled through analysis of satellite images by the Global Land Analysis and Discovery group at the University of Maryland (UMD GLAD).

“Both the satellite images, and the methods used to analyse them, have improved over the years, which means that more harvested forests are identified in the images in the latter years of this data set. I have had confirmation from researchers at UMD GLAD that this data cannot be used in the way JRC Ispra have used it,” says Håkan Olsson, professor in forest remote sensing at the Swedish University of Agricultural Sciences.

The article even contains other doubts, such as attempts to disregard the effects of certain types of forest damages but not others.

Contacts

Jonas Paulsson, statistician, Swedish Forest Agency, tel. +46 (0)19-44 64 07, jonas.paulsson@skogsstyrelsen.se

Svante Claesson, Head of Unit, Swedish Forest Agency, tel. +46 (0)36-35 93 68, svante.claesson@skogsstyrelsen.se

Jonas Fridman, Head of Programme of The Swedish National Forest Inventory, Swedish University of Agricultural Sciences, tel +46 (0)90-7868473, jonas.fridman@slu.se

Håkan Olsson, professor in forest remote sensing at the Swedish University of Agricultural Sciences,
tel +46 (0)90-7868376, hakan.olsson@slu.se