Introduction
In order for a nation to efficiently manage and preserve its natural resources using scientific methods, the detailed state of the resources and any changes that occur must be surveyed and analyzed periodically. Many countries, including South Korea, have surveyed forest resources with scientific methods suited to the environment in order to generate national statistics, support policy makers, and maintain sustainable forest management. Furthermore, various intergovernmental organizations (IGOs) and treaties, of which South Korea is a member, require the periodic submission of forest statistics and other related national reports.
The National Forest Inventory (NFI), conducted by the Korea Forest Service (KFS), aims to “survey and evaluate forests nationwide using scientific methods, in order to acquire basic national statistics on forest resources and regularly monitor trends; thereby providing fundamental data for basic forest planning and the establishment of forest policies for the purpose of implementing sustainable forest management” (KFS, 2010). Recently, South Korea expanded the range of variables investigated in field surveys from that of previous surveys, which centered on timber resources, in order to enable the evaluation of multi-purpose forest resources (KFRI, 2014). Using data from the NFI to calculate forest resource statistics and increase utilization will enhance the validity of the NFI. This requires expert knowledge about the forest resources statistics; required by the IGOs and treaties as well as case studies from overseas including countries with advances in forestry. These countries have implemented systematic national forest inventories and calculated diverse forest statistics, having adapted their survey systems to generational changes.
The purpose of this article is to describe the development of Korea’s NFI and summarize its academic uses. Comparing the survey systems and state of forest resource statistics to countries with advancements in forestry would contribute significantly to qualitative improvements to Korea’s NFI. To this end, survey methods and forestry statistics among the Unites States, Canada, Japan, China, and European countries, such as Germany and Finland, were briefly compared. The uses of NFI data were analyzed: various studies collected from journal articles, conference papers, theses, and research reports between 2008 and 2015 within domestic universities and related research institutes.
Development of NFI in South Korea
In South Korea, the Division of Forest Resources Information, Korea Forest Research Institute (the name was changed into “National Institute of Forest Science”) carried out national forest inventory under Korea Forest Service (FAO, 2007c). The main objectives of forest inventory is to estimate the present state and the changing trends of national forest resources (Kim et al., 2015). Additional purposes of forest inventory are to measure the volume, biomass, carbon stocks and characteristics of trees and stands including forest land area (Kim et al., 2008). From 1971, South Korea conducted national forest inventory 6 times with a time interval. The 2nd, 3rd, 4th and 5th inventory was done in the period of 1978 - 81, 1986 - 92, 1996 - 2005 and 2006 - 2010 (Table 1). Last one is done from 2010 - 2015 (FAO, 2014).
At the fifth NFI, a change in inventory system was created (FAO, 2014): At the 4th NFI, forest inventory was conducted for 10 years, but from the fifth inventory (2006 - 2010) 5 years interval was implemented. In new system, 20% of permanent plot through the country was inventor every year (Kim et al., 2008).
The legal basis of Korea’s NFI
The Korean NFI is conducted on the basis of Article 32 (‘Surveys of Forest Resources’), Clause 1 of the Creation and Management of the Forest Resources Act, which states that “the Minister of the Korea Forest Service shall survey forest resources on a regular basis, as prescribed by Ordinance of the Ministry of Agriculture, Food, and Rural Affairs,” and that “the Minister of the Korea Forest Service shall publish the results of [these] surveys.” Furthermore, in the Enforcement Regulations for the Creation and Management of Forest Resources Act, the following specific regulations are included:
Article 1 (National Forest Inventory)
-In accordance with the regulations in Article 32, Clause 1 of the act, the nation’s forest resources shall be surveyed every five years.
-The content of the surveys prescribed in Clause 1 includes the following: herbaceous vegetation survey, shrub/sapling survey, survey of forest damage, including dead and felled trees, survey of timber resources, soil survey, and other articles considered necessary by the Minister of the Korea Forest Service.
-Requirements for the survey prescribed by Clause 1, including survey methods, shall be determined by the Minister of the Korea Forest Service.
Sampling design of Korea’s NFI
The 5th NFI, which began in 2006, resulted in data collection from sample points that were allocated by hierarchical clustering. The country was divided into 4 × 4 km squares and a sample point was placed at each intersection on the grid. The design included the flexibility to acquire sufficient sample points, such as using a 2 km or 1 km grid in large cities, where it is difficult to acquire a suitable sample point because of the lack of forest area. Sample points were only selected when the intersection of the grid landed on a forested region, and this hierarchical sampling method was used to allocate approximately 4,000 sample points across the entire country (FAO, 2007c).
From each sample point centered on an intersection, identical sample points were allocated at a distance of 50 m in three directions (true north, 120°, and 240°), such that each cluster consisted of a total of four sample points (Kim et al., 2008).
Measured variables in Korea’s NFI
The 5th NFI consisted of 32 measured variables. In the 6th NFI, the Forest Health Monitoring (FHM) survey was integrated at the same sample points, causing the number of measured variables to increase to 52. In 2013, additional categories of rock exposure, state of erosion, soil consistency, soil moisture, and organic layer thickness were added to the FHM survey resulting in 59 measured variables.
The Korea’s forest health monitoring
Forest health refers to the healthy and diverse maintenance of the forest ecosystem to promote its ecological function. In order to analyze and evaluate changes in forest ecosystems for healthier forest management, Korea legislated Article 19 of the Forest Protection Act in October 2010, thereby providing a legal basis for diagnosing and evaluating forest health and vitality.
The purpose of the FHM is to collect objective data regarding the current state of and potential trends in the health of the Korean forest ecosystem, and to explore countermeasures based on the analysis and evaluation of patterns of weaknesses in forest health.
As mentioned above, the FHM has shared some of sample points with the 6th NFI since 2011, which implies that the survey is being conducted systematically across the whole country. However, only the central sample point (S1) was investigated in the FHM (Fig. 1).
Five variables are measured for only assessment of FHM, constituting 12 indicators in four categories of ‘tree health,’ ‘vegetation health,’ ‘soil health,’ and ‘atmospheric health.’ Lower strata vegetation was surveyed in a 2 × 2 m survey area. The coordinates of target standing trees were measured within the central sample point (S1).
The state of NFI in other countries
United States
Led by the USDA, the Forest Inventory and Analysis (FIA) is conducted from four regional centers across the United States with the goal of conducting surveys and statistical analysis for the nation’s forest resources. The US National Forest Inventory is repeated every five years, reporting on the current state of forest resources, previous usage, future prospects, and supply and demand of renewable resources (Table 2).
US congress established national forest inventory by the McSweeney-McNary Act of 1928 for estimating the national and regional timber supply (Zeng et al., 2015). FIA research program has been in continuous existence since mandated by Congress in 1928. FIA program has been monitoring and reporting on status, condition, and trends in the nation’s forests for over 70 years and FHM program for the last 11 years (Smith, 2002). First 50 years forest inventory by FIA estimated the timber resources (Van-Hooser et al., 1992; Powell et al., 1993).
Since the 1930s, FIA has provided 225 state wide forest inventories; seven national assessments (USDA, 1958, 1965, 1973, 1982; Waddell et al., 1989; Powell et al., 1993; Smith et al., 2001); two national biomass studies (USDA, 1981; Cost et al., 1990); two national private forest land ownership studies (Birch, 1996); a national satellite forest cover map of the USA (Zhu and Evans, 1994); hundreds of primary mill, utilization, and residential fuelwood studies (Smith, 1991; May, 1998); over 300 reports on non-timber issues (Rudis, 1991).
Three phase sample design has been used: 20% of the field samples are measured annually, with a completion of a measurement cycle every 5 years. Each sample plot is randomly located inside a matrix of non-overlapping hexagons (approximately 2,400 ha per hexagon) (Magnussen et al., 2007). Common sampling approaches- double sampling for stratification are used for data collection.
In 1999, FIA and FHM fully integrated the detection monitoring plots of FHM into the FIA program. These plots became Phase 3 of the enhanced FIA program. Approximately 8,000 Phase 3 plots are located in whole country (Smith et al., 2001). Plots are measured rotationally every year. Depending on the state the remeasurement cycle is 5 - 10 years (Zeng et al., 2015).
Canada
First national forest inventory in Canada was carried out in 1951 (Paille, 2013). Questionnaires based inventory was conducted before 1981. After 1981, national inventories were done by sampling methods (Gills et al., 2005). After that in 1986 NFI gave information about productivity, size and volume of a species, in 1991 NFI includes all major forest areas across the country and in 1994 NFI gave information about water bodies. Canadian Forest Inventory Committee (CFIC) create a new approach for national forest inventory. This committee applied a new design that was using plot based design consisting of permanent unit. NFI programmed to estimate the state, sustainability and health of forest in 10- to 20-year cycles (Power et al., 2001; Gills et al., 2005).
A sample design was described in A Plot-based National Forest Inventory Design For Canada (Natural Resources Canada, 1999). First, it build a systematic network of plots in Canada and then stratified the sampling plot within terrestrial ecosystem with varying strata. After that, photo plots lies on permanent plot and also estimate the ground sample plot (Gills, 2001; Gills et al., 2005). Plot size vary in shape and size but the attributes should be same and sampling intensity is 20 × 20 km (Gills et al., 2005). All the sample location are within 4 × 4 km network in whole country. National forest inventory plots are permanent but data is collected from photo and ground plot (Gills, 2001). Permanent photo plots have 2 × 2 km sample grid located in each sample plot. Photo plots provide basic data and use to estimate 1% of sample mass. Approximately 18,850 photo plots (1:53,000 ha) are present in forested area (Gills et al., 2005). Ground plot measured synchronizing with photo plot and provide detailed data. 50 ground plots are located in each ecozone (Gills, 2001).
Japan
In 1897, a forest law regarding with establishment of nationwide protection system was created. After that on 1900, a survey was conducted in 1905 to estimate the forest resources. In 1990 - 2000 total national land area was estimated (FAO, 2007b). Japan conducted preliminary monitoring of forest resources for three years from 1991 to 1993. Based on this, they have conducted a national forest inventory at five-year intervals from 1999 to the present. The Japanese National Forest Inventory is managed by the Japan Forest Technology Association (JAFTA), whereas field surveys are the responsibility of approximately 14 companies in the forest industry.
The purpose of inventory is to evaluate the various aspects of forest throughout the country (Kitahara et al., 2009). Japan used two types of forest inventory system: first inventory is conducted by forest law, require forest planning system and another is to monitoring forest resources (Hirata et al., 2006). Every five years national and sub national inventories are conducted in Japan (FAO, 2007b).
From 1999 systematic sampling design is used in Japan. Sampling design is applied with 4x4 km ground grid. Total number of sample plot is 15,675. Every sample plot is composed by three overlapping subplots which are of different radius (FAO, 2007b). Each plot consists three concentric ring with various diameter of 5.64 (0.01 ha), 11.28 (0.04 ha) and 17.84 m (0.1 ha) (Kitahara et al., 2009). Remote sensing was also used in Japan from 1960s within the forest register system (FAO, 2007b). The main purpose was to classify the forest type (Hirata et al., 2006). In 2006 remote sensing was used in NFI with a view to estimate the forest boundaries (FAO, 2007b).
The results of the National Forest Inventory are used to calculate the forest resource statistics required for the compilation of the Statistical Yearbook for the forest industry and the Montreal Process report. Japan’s Statistical Yearbook for the forest industry consists of a total of 190 variables in the eight categories of ‘national economy,’ ‘forest resources, maintenance, and preservation,’ ‘forest industry,’ ‘forest production,’ ‘timber production,’ ‘finance,’ ‘overseas forests and forest industry,’ and ‘other.’ This means they measure twice as many variables than does South Korea, with the main difference being that Japan’s Statistical Yearbook includes various sociological statistics related to the forest industry.
China
The first inventory was carried out from 1973 - 1976 in China with the help of Ministry of Forestry. The purpose was to estimate the micro change in forest land and also the volume of forest. After 1973 NFI was carried throughout the country with five years interval (FAO, 2007a). Until now eight NFI was conducted in China: the periods are 1973 - 1976 (first), 1977 - 1981 (second), 1984 - 1988 (third), 1989 - 1993 (fourth), 1994 - 1998 (fifth), 1999 - 2003 (sixth), 2004 - 2008 (seventh) and 2009 - 2013 (eighth). China includes three stages of forest inventory systems: the National Forest Continuous Inventory (NFCI); the Forest Management Planning Inventory (FMPI); and the Forest Operation Design Inventory (FODI) (Xie et al., 2011).
From first inventory, China used a systematic sampling design (FAO, 2007a) and a parameter field plot. Commonly two dimensional sample grid is used but sample grid varies from province to province (Zeng et al., 2015). Plot size is varied from 2 × 2 km to 8 × 8 km and approximately 415,000 fixed ground plot was established in 7th NFI across the whole country (FAO, 2007a). Plots shape also varies such as square, circular, and rectangular. More than 90 % of the provinces used square plots, but Tibet used circular plot, while rectangular plots are used in inner Mongolia and in one sub-population of Heilongjiang (Zeng et al., 2015). Sample plots are systematically laid out on the grid dots of topographic maps (scale 1:50,000) (FAO, 2007a).
Remote sensing is used in forest inventory at the middle of 1950s (Zeng et al., 2015). Remote sensing is using to know the geo-spatial distribution of forest and also to prevent biased estimation of field measurement (FAO, 2007a). 1:500,000 scale and aerial photos of small scale were used to determine forest area (Zeng et al., 2015). Remote sensing plot is set with equal distance systematic sample and the distance is mostly less than 2 × 2 km (FAO, 2007a).
Europe
In Europe first forest inventory was conducted on 14th and 15th century. In 19th century forest inventory was started to use as a tool for forest management. On that time only data was gathered but there was no statistical analysis. The purpose of national forest inventory was to provide information regarding timber resources, forest land and type for forest authorities, timber users, and planners (Zeng et al., 2015). Different countries have different purposes for forest inventory.
Lund (2009) reported that Norway started its first national forest inventory on 1910s. Whereas Sweden started its NFI on 1923 (Fridman et al., 2014). In the late 1940s in the German Democratic Republic; in 1958 in France; in the 1960s in Austria; and in the 1980s in Switzerland national inventory was conducted (Zeng et al., 2015). The Austrian National Forest Inventory (AFI) started in at 1961 with a systematic grid and a period of 10 years (Schadauer et al., 2005). European national forest inventory has been carried out since 1993 (Bastrup-Birk, 2010). It has five regional office (Spain, Germany, France, Croatia and Sweden) and three projects center. The time interval for inventory depending on the countries requirements (Zeng et al., 2015).
Germany’s first systematic survey of the nation’s forest resources was conducted in West Germany, before unification, from 1986 to 1990. A national forest inventory with statistical samples taken from the entirety of Germany was conducted for two years from 2001 to 2002, and will be conducted every 10 years thereafter, with the 3rd National Forest Inventory recently completed. The German National Forest Inventory is led by the Federal Research Institute for Rural Areas, Forestry, and Fisheries and state forestry research institutes. The Federal Research Institute gathers data that have been collected from surveys across 16 states, analyzes these data, and compiles the National Forest Inventory Report.
The German National Forest Inventory Report is characterized by a large number of statistical variables related to forest changes, which allows them to make forecasts about future changes. A total of 168 detailed statistics are reported for 16 categories, with 51 of the detailed statistics relating to change in area, timber stocks, harvest-related variables, and forest damage.
Finland’s 1st National Forest Inventory began in 1921 (Tomppo et al., 1999), and in 2011, they began the 11th National Forest Inventory. The Finnish Forest Research Institute (METLA) is the lead organization responsible for field surveys and statistical analysis. Based on these data, the ‘Forest Resources’ section of the Finnish Statistical Yearbook of Forestry is revised every year.
The Finnish Statistical Yearbook of Forestry contains 201 statistics in the categories ‘forest resources,’ ‘forest biodiversity and health,’ ‘silviculture,’ ‘roundwood trade,’ ‘removals and transportation of roundwood,’ ‘multiple-use forestry,’ ‘forest sector labor force,’ ‘wood consumption,’ ‘energy,’ ‘forest industries,’ ‘foreign trade by forest industries,’ ‘forest sector in Finland’s national economy,’ and ‘international forest statistics.’ This is the most diverse collection of forest statistics among countries with advanced forest industries, taking a broad interpretation of the scope of forest industry and providing statistics that treat forest-derived industries as part of the forest sector.
The forest statistics of advanced countries in forestry include diverse statistical variables related to the entire industry including forest resources. However, the Korean Statistical Yearbook of Forestry contains only half as many variables as that of Japan and Finland, and there are fewer variables related to forest resource statistics. Those forest resource statistics that are included in the yearbook primarily relate to forest area and timber stocks, but more detailed information needs to be collected.
Finland includes trends in their statistics in addition to the current state of the forest area and timber stocks, as well as statistics relating to growth volume and rate. The United States and Germany are also not limited to the current state of variables but also describe forecasts. In South Korea, because the 6th NFI is comparable with the NFI of five years earlier, it should be possible to extract useful information for establishing future forest policies by examining changes in statistical variables over time, as is done in countries with advanced forestry, and predict trends in forest resources based on these changes.
The use of national forest inventory data in academic areas
The outputs of domestic research by using NFI data between 2008 and 2015 were classified into scientific journal articles, conference papers, theses, and research reports. Furthermore, these publications were grouped by subject (Table 3).
Scientific journal articles
A total of 39 articles were published, at an average of 4 - 5 per year. The content of the articles was diverse: including timber stocks, forest growth, forest biomass, forest carbon, thematic maps, statistical methodology in small forest area, and methodology for estimating forest statistics. The majority of articles suggested efficient, scientific methods to enable the calculation of relevant forest statistics based on partial data from specific regions.
Conference papers
A total of 74 conference papers were published, at an average of 9 per year. The subjects covered were more diverse than journal articles: including timber stocks, forest growth, forest biomass, forest carbon, forest vegetation, statistic methodology in small forest area, methodology for estimating forest statistics, thematic maps, stand density, dead trees, site index, survey methodology, forest health, and forest changes.
Theses
Of the 18 total theses, master’s theses were double of those of doctorate theses. The most common subject matter was forest carbon, occurring in seven theses, followed by forest growth, timber stocks, and forest biomass, and finally methodology for estimating forest statistics and statistical methodology in small forest area was the subject of the thesis. Forest carbon and forest biomass are both related to greenhouse gas statistics, and these two fields accounted for eight of the theses, more than half. This is related to the growing interest in estimating carbon storage capacity and carbon absorption in the forest sector in relation to climate change.
Of doctorate theses, three were on the estimation of forest carbon storage capacity and uncertainty, demonstrating the urgency and importance of research in this area. The other doctorate thesis was on methodology for estimating forest statistics. Forest carbon was the most common subject among the master’s theses, occurring in three theses, followed by forest growth, forest biomass, and timber stocks, in two theses each. There was also one thesis on methodology for estimating forest statistics. In the future, if data can be accumulated from monitoring surveys at fixed sampling points, research results could be obtained for more diverse fields, such as forest change and forest health assessment. This is expected to diversify the subject matter of theses.
Research reports
Nine reports were related to NFI research projects. Each subject, such as forest growth, forest carbon, methodology for forest statistics, survey method and the change of forest, has been reported. Recently, the suggestion of utilization of NFI data and extended forestry statistics using NFI data were published.
Outputs of researches utilizing NFI data were broadly categorized into 15 areas (Table 3). Journal articles have been published in seven areas, including timber stocks, forest biomass, and forest growth; theses and research reports each deal with six areas of research. In addition, conference papers were concerned with the most diverse subjects, covering 14 out of the 15 areas, with a high number of publications. The most common research area among all publications was on forest growth, with a total of 34 papers, followed by 27 papers on forest carbon storage capacity. Timber stocks, which, together with forest area, is one of the most important items in forest resource statistics, and is covered relatively often, with a total of 18 publications among journal articles, conference papers, and theses, although not among research reports.
In terms of other areas, there were twelve publications for forest biomass, ten for small forest area statistics, and nine for methodology to estimate forest statistics. Similar to forest carbon, forest biomass is related to greenhouse gas statistics, and as such research will be conducted in this area in the future because of the significance of climate change. Methodology for estimating forest statistics dealt with overall measures for calculating forest statistics, including the calculation of small forest area statistics, calculation of annual forest statistics, re-calculation of past forest statistics, and algorithms for calculating individual forest statistics.
Concerning forest health and vitality, research was conducted in the areas of forest vegetation, forest health, and dead trees. A total of five publications, consisting of conference papers and research reports, were found on the compilation of thematic maps that provided information about the spatial distribution of forest statistics.
Although there have been five publications on forest change, because the NFI is repeated every five years at the same sampling points, we expect this area of research to become more active in the future. In particular, analysis of forest change is anticipated to provide information not only on changes in the usage of forestland and in forest type, diameter structure, and age structure, but also on changes in forest growth, including growth rate.
Application of NFI data
Because the majority of studies have typically been conducted in restricted regions, rather than providing data for the country as a whole, there are tasks that need to be completed in order to use research results in applicable forestry practice. In the case of forest growth estimating the tree diameter, tree height, growth volume, and growth rate by region is an important task; it should be possible to calculate statistics using the NFI monitoring data. At the same time, it would be possible to calculate growth volume and growth rate by region and by tree species.
In the areas of forest biomass and forest carbon, which are related to greenhouse gas statistics, various studies have been conducted on estimating forest biomass, carbon storage capacity, and carbon absorption for different types of forest, as well as employing classifications used in the estimation of timber stock, such as region, forest type, and age class. Therefore, it should be possible to calculate these statistics using NFI data, and it will also likely be necessary to calculate statistics by species.
In the area of forest vegetation, research has been conducted on vegetation distribution, the distribution of dominant species, evaluating species diversity, and classifying types of forest vegetation. If the regional and topographical distribution of forest vegetation were categorized by type, it would be possible to calculate statistics related to the distribution of forest vegetation.
The field of forest health has relatively little research. If more data were accumulated in the future, it would be possible to calculate statistics related to the change of forest health and vitality. In terms of dead trees, it should also be possible to develop dead tree models that could estimate dead trees and their distribution based on NFI sampling point data. Similarly, as described above, because repeated measure is conducted every five years at fixed sampling points, it should be possible to calculate various statistics for forest changes. It will be possible to calculate statistics that enable assessment not only of changes in forestland usage, but also of changes in timber stocks and other forest resources, such as standing tree structure and forest growth.
With regard to thematic maps, studies have been conducted using NFI data to construct maps for the distribution of timber stocks, forest biomass and forest carbon, and species diversity. Although Korea has existing thematic maps for forest type and forest soil, a more diverse array of thematic maps is required, depending on the intended use. If numerous variables could be measured regularly throughout the country, as is done in the NFI, it should be possible to construct various thematic maps that are suited to particular needs.
Conclusion
Korea conducted the national forest inventory 6 times with a time interval from 1971 with legal basis, but the 6th NFI is only comparable with the NFI of five years earlier. Outputs of researches utilizing NFI data have been broadly used in academic area which can be categorized into 15 areas. Although the forest statistics of advanced countries in forestry include diverse statistical variables, the Korean Forestry Statistics contains only half as many variables as that of advanced countries in Forestry.
There could not be overemphasized in the importance of the quality control of NFI data. Although scientists and working groups supported by Korea Forest Service have been substantial efforts to find more efficient sample designs, get consistent measurement, and implement advanced technologies to measure diverse variables, additional data for FHM since 6th NFI must be under quality control. This will be an essential part of the inventories which can provide the chronological change of forest health. Furthermore, this will be closely related to the indexes of forest ecosystem service and sustainable forest management.
