EN
Forest area in Poland is annually evaluated as a part of public statistics research. However, this information is based on land use resulting from Land and Property Register (LPR). Delays in the reclassification of afforested land, as well as the natural expansion of trees on abandoned agricultural lands, observed over the last decades, caused that information from LPR becomes unreliable. In many countries forest area is assessed within the National Forest Inventories (NFIs) based upon systematic grid of sample plots. NFI in Poland has been performed since 2005, in 4×4 kilometre grid with clusters consisting of five plots. Until 2014 measurements were made only on these sample plots which were located in forest according to LPR records. Within the 3rd NFI cycle (2015−2019) the areas fulfilling the criteria of forest definition, but located on non−forest land referring to LPR, has become the object of study. The aim of this work is to present statistical approach for evaluation of forest area using NFI cluster sampling data. Additionally, results from two year measurements (2015−2016) were analysed and compared with LPR data. Attributes of NFI plots allow to apply national forest criteria as well as the land use and land cover thresholds recommended by Food and Agriculture Organization of the United Nations or Kyoto Protocol. Our research shows that forest cover in Poland is in the range of 32.8−33.0%, depending on forest definition used, and is almost 3% higher than official LPR data (30.1%). The standard error of forest cover, based on two years NFI data is 0.44. Thus, with 95% probability the true value of this parameter lies between 31.9 and 33.7%, while country thresholds of forest definition are used. Additionally it was assessed that using the entire NFI cycle data the standard error of forest cover should be lower i.e. less than 0.3. The National Forest Inventory seems to be an appropriate tool for monitoring forest area in Poland.