The NFI principle
NFI data is collected through repeated fieldwork on sample plots located throughout the Czech Republic. The ultimate NFI output corresponds to estimates of predefined parameters (attributes of the Czech forests and landscape). These estimates can be produced for the level of the whole Czech Republic or its predefined subregions (e.g. NUTS regions or natural forest regions). Accuracy of NFI results is expressed by means of standard errors and confidence intervals.
The origin and characteristics of the information provided by the NFI
Within the Czech NFI, a representative sample of locations (sample plots) is surveyed in a field in order to obtain necessary data which is subsequently evaluated by methods of survey sampling. Data from a field assessment can be combined with variables obtained by remote sensing (RS), which aims to increase the accuracy of estimates. The NFI evaluation methodology, regardless of we use RS or not, guarantees unbiasedness of results coming from the unbiasedness of field data.
The field survey of the second cycle of the Czech NFI (NFI2, 2011–2015) was carried out on more than 23 thousand sample plots, most of which were classified as the forest land category.
The selection of sample plots and the time frame
The NFI2 data collection was carried out simultaneously in two sampling grids, determining locations of sample plots throughout the CZ. A sampling grid is defined so as to enable a spatially even distribution of sample plots, resulting in higher accuracy (reliability) of NFI estimates.
Within the NFI2, a repeated field survey was carried out on sample plots of the NFI1 grid (two sample plots in a square of 2 × 2 km) established ten years ago (2001–2004). This enabled, amongst others, the evaluation of dynamic variables such as the amount of felling, increment and changes of growing stock. In addition, the repeated assessment of the NFI plots made possible to analyse the causes of the forest cover change.
A new sampling grid was established within the NFI2 to address the need for combining field survey information (one point in a 2 × 2 km square) and photogrammetric (photo-) interpretations (a densified grid, one point in a 500 × 500 m square). Estimates of many target parameters can be published more frequently, or for smaller areas (accuracy enhancement effect) due to this combined approach. Such an approach is advantageous for continuous NFIs, particularly with regard to the annual provision of selected NFI outputs for which only a fraction of the full sample size is available.
During the NFI2, the photo-interpretation and field survey schedule was strictly set – data was collected in sub-grids, covering the entire territory of the Czech Republic in each (spring and autumn) season of each survey year. This ensured that the NFI2 data, processed for different parts of the Czech Republic, remained comparable (i.e. referring to the same period). It is significant mainly as far as dynamic variables are concerned, such as estimates of felling, an increment and changes of growing stock or of a forest area, i.e. when it is desirable to maintain an equal time interval between successive surveys at particular sample plots.
The landscape aspect
In terms of the Czech NFI, the land cover as a whole is analysed first, and only then it is possible to focus the sample survey on the forest land category, the occurrence of which is not pre-specified in any way (e.g. by the land registry or a status according to the forest management plans). Valuable information for the sector of environment protection and landscape engineering can thus be provided. Quite recently, many NFIs (particularly in Europe) extended their scope beyond the forest land category in order to address information needs existing at the whole landscape level.
The above principles were implemented into the Czech NFI photo-interpretation where, in addition to sample points, landscape transects have also been included. With a slight increase in photo-interpretation costs, the scope of information provided by the Czech NFI has been extended markedly.
NFI data administration and evaluation
All data, obtained within the Czech NFI (both the field and photogrammetric survey), is managed in the PostgreSQL database (DB) with the PostGIS extension (storage, analysis and management of geographic data). Extensive functionality has been created in the database, which ensures the correct sequence and logical coherence of all activities within the collection, quality assurance, storage and evaluation of NFI data.
There are also client applications, connecting to the central NFI database. They have been designed and, except the photo-interpretation app (by Topol Software, s. r. o.), also implemented by ÚHÚL employees. Designated NFI experts manage these applications daily. Key components of the whole system are either FOSS (Free and Open Source Software) or ÚHÚL's own specific applications (e.g. the application for field survey data registration and exports of data for NFI field checks; the application for checking field survey data; the application for checking and registering GPS coordinates, etc.).
The evaluation of NFI outputs is also carried out in the PostgreSQL database. In this DB, many functions were implemented that pre-process the NFI data and pass it to a software library est4nfi calculating statistical estimates (results) of the Czech NFI. The est4nfi library is currently implemented in the C++ language and has been developed by ÚHÚL, at a specialised workplace (the Methodological Centre of the Czech National Forest Inventory) located in Kroměříž since 2006.