Environmental Science
Vol. 32, No. 2, pp. 140-148, May, 2016 http://dx.doi.org/10.7747/JFES.2016.32.2.140
Estimation of the Forest Stand Volumes from Forest Inventory Data Based on Synthetic
Estimation Method: A Case of the Economic Forest in Gangwon-do, Republic of Korea
Hwan seok Seo, Jeong mook Park and Jung soo Lee*
Department of Forest Management, College of Forest and Environmental Sciences, Kangwon National University, Chuncheon 24341 Republic of Korea
Abstract
This study aims to estimate the forest volumes of the economic forest in Gangwon Province of Republic of Korea (hereinafter referred to as Gangwon) through the synthetic estimation. To estimate the forest volume, Stratified systematic sampling method was used along with the forest type maps and the 5th National Forest Inventory data. The synthetic estimation includes sample plots of the expanded areas as well as those of the target area, and the forest volume of economic forest in every city and county throughout Gangwon. Results show that the average forest volume calculated by synthetic estimation was 159.6 m3/ha in national economic forest and 129.6 m3/ha in private economic forest. The total forest volume of the national economic forest was approximately 59.45 million m3, which was 20.18 million m3 higher than that of the private economic forest. On the other hands, the standard error of the national economic forest was approximately ±2.21 m3/ha, which was ±0.30 m3/ha lower than that of the private economic forest. The lowest standard errors was about ±3.12 m3/ha in broad-leaved forest, followed by ±4.33 m3/ha of mixed forest, and
±5.78 m3/ha of coniferous forest.
Key Words: synthetic estimation, economic forest, forest volume, the 5th National Forest Inventory, 1:5,000 forest type map
Received: January 29, 2016. Revised: April 17, 2016. Accepted: May 4, 2016.
Corresponding author: Jung soo Lee
Introduction
According to the 4th forest basic planning, approx- imately 3.5 million ha of economic forest has been planned to be promoted before 2,030 to facilitate the sustainable production of timber. Also, the forest policy for economic forest growing sites has been established to secure contin- uous production of a fixed quantity of timber up to 2,050.
However, not only the utility value of timber decreased due
to improper management practices conducted in economic forest growing sites, but the availability also decreased due to the quality degradation of domestic timber. Accordingly, the necessity of forest management and of clear grasp on the status of forest resources has been emphasized. For this rea- son, Korea Forest Service has been performing promotions to activate management of both national forests and private forests since 2002 by raising efficient forest resources for economic forests and intensive management of economic
Fig. 1. Study area.
forest growing sites. In addition, since 2005, Korea Forest Service has been operating designated economic forest growing sites for stable production and supply of better timber (National Institute of Forest Science 2011b; Korea Forest Service 2013). Chong (2007) performed a study to understand topographic characteristics of national and pri- vate forests by using Geographical Information System (GIS) and National Forest Inventory (NFI) to divide types of forest per growing site according to the representative forest types within the economic forest growing sites, and Oh et al. (2008) proposed a plan for the management of economic forest as well as a long-term vision, based on the result of examining flora and vegetation of Nonsan eco- nomic forest growing site. In addition, Kim et al. (2009) conducted research with the subjects of economic forest growing sites to provide basic information on production of timber by setting target forest type and target timber pro- duction (Kim 2012). However, the result provides in- formation only on forest area, target forest type and on flora, lacking specific statistical calculation of stand status.
On the other hand, a data-based estimation method, the method of calculating regional statistics through collected
data, has widely been applied in diverse fields as well as in the forest field. In particular, the synthetic estimation has a merit of achieving reliable estimation by utilizing the sam- ple data designed for calculating national statistics (Kim et al. 2008; Noh 2009). Yim et al. (2010) estimated forest sta- tistics with an administrative district as a unit, based on the expanded area selected by synthetic estimation. Kim and Kim (2015) estimated forest volume in an administrative district level, based on the homogeneity of influential fac- tors determining the regional characteristics of forests in the expanded area selected through synthetic estimation. How- ever, previous studies and statistical data mainly estimated the general forest statistics in a national or certain admin- istrative district level, so the basic information on the status of resources per section of economic forest is still insuffi- cient, which requires further studies on this field (Hwang 2009; Kim et al. 2010). This study aims to calculate the ba- sic forest statistics of economic forests in Gangwon prov- ince, by using synthetic estimation methods, and to conduct comparison and analysis the statistics by ownership types.
Materials and Methods
Study Area & Data
Economic forest sites in Gangwon were selected as the study area, and total of 86 sites from 18 administrative dis- tricts were selected as the subjects (Fig. 1). Of the 450 eco- nomic forest sites throughout Korea, Gangwon, consisting of 49 national economic forest sites and 37 private economic forest sites, is shown to include the highest number of eco- nomic forests as a single province, and in particular, Gangwon is found to have the highest forest volume which is approximate 142.5 m3/ha (National Institute of Forest Science 2011b; Korea Forest Service 2015). Utilized GIS data for this study include 1:5,000 forest type map (herei- nafter referred to as FTM), administrative map, economic forest site zone map, and National Forest Inventory 5th sampling plot data (hereinafter referred to as NSPD). In addition, FTM was used to calculate forest area, and un- stocked forest land and non-forest area were excluded. For NSPD, the average growing stock data per ha per subplot of cluster plot constructed in 4 km×4 km unit was respec- tively used as individual sampling plot data.
Method
In order to estimate the forest volumes of economic for- ests in Gangwon, initially, the average forest volume per ha per stratification based on the expanded area of NSPD was calculated by using the synthetic estimation method. Secon- dly, FTM and economic forest site zone map were used to calculate areas per stratification of economic forest. Finally, synthetic estimation was used to calculate standard error per stratification of economic forest and total stock (Fig. 2).
Selection of Expanded Area
Synthetic estimation refers to the method to estimate small targeted area only based on the sampling plot data collected in the expanded area, without any additional materials. The selection of expanded area can have an influence on the fur- ther calculation of other forest statistics, so a large area which shows characteristics similar to those of the small tar- get area is mostly selected (Gonzalez 1973; Yim et al.
2010). Accordingly, in this research, the method used by Yim et al. (2009) and by Han (2011) was adopted to select Gangwon as the expanded area since Gangwon is the place where all economic forest sites were included in the same area (Cressie 1993).
Calculation of Forest Statistics
FTM and NSPD of economic forests in Gangwon were used for calculating forest statistics, and stratification of sampling per age class and per forest type was conducted to conduct the analysis. Here in this study, forest type is com- posed of coniferous forest (C), broad-leaved forest (B), and of mixed forest (M), and their age class (1 age class=10 year) ranges from II to VI. On the other hand, in case of the age class, class I was excluded due to its low growing stock, and classes higher than VII with small forest area were con- sidered as the age class VI and included in the calculation (National Institute of Forest Science 2011a). A method to calculate the total forest volume () is as shown in Formula 1, and the area () per stratification was calculated, based on the stocked area of economic forest site but excluding unstocked forest land and non-forest area. Formula 2 was used to calculate average forest volume () per ha per stratification (Yim et al. 2010): n signifies the number of sampling plot, r signifies economic forest site, l signifies Gangwon, v signifies forest volume per ha per sampling
Table 1. Number of the 5th National Forest Inventory plots in the study area
Distribution of spatial Forest type Age class
Total Ratio (%)
Ⅱ Ⅲ Ⅳ Ⅴ Ⅵ
Gangwon province C 40 158 294 104 54 650 23
B 44 277 534 308 215 1,378 49
M 27 220 341 123 91 802 28
Total 111 655 1,169 535 360 2,830 100
Economic forest
National C 13 33 64 32 25 167 19
B 2 53 177 129 96 457 53
M 3 31 113 61 36 244 28
Total 18 117 354 222 157 868 100
Private C 11 45 93 23 5 177 31
B 14 72 90 26 5 207 36
M 14 75 75 17 6 187 33
Total 39 192 258 66 16 571 100
plot, and h signifies sampling stratum (age class per forest type).
× ··· (1)
··· (2)
Comparison of Standard Error
Distribution of estimation (Formula 3) and standard er- ror (Formula 4) were respectively calculated to compare er- rors between the estimated of average forest volume per ha per stratification and average forest volume per ha per sam- pling plot (Cochran 1977).
··· (3)
··· (4)
Results and Discussion
Status of NFI 5th Sampling Plots
The number of NSPD of economic forests used to ana- lyze the forest volume of economic forests was 868 for na- tional forests and 571 for private forests, respectively. The total number of Gangwon’s NSPD, used in synthetic esti- mation was 2,830, which was about 3.3 times higher and 5 times higher compared to that of national economic forests and private economic forests, respectively (Table 1). The number of expanded sampling plots was shown to be di- versely distributed, considered by the administrative dis- trict level in previous research performed by Han (2011) and Yim et al. (2010), but it was approximately 5 times higher in average than that of the target district, such ratio of which was similar to that of this research. As for the ratio of forest type per NSPD of national economy forests, broad-leaved forest is found to account for approximately 53%, which was the highest distribution, followed by mixed forest with approximately 28%, and coniferous forest with approximately 19%. On the other hand, the ratio order of forest type per NSPD of private economy forests was iden- tical to that of the national economy forests, but the dis- tribution of sampling plots was approximately 31%, 36%, and 33% for coniferous forest, broad-leaved forest, and mixed forest, respectively, all of which represented a similar distribution of sampling plots.
Table 2. Estimates of mean of forest volume with forest types (unit : m3/ha)
Forest type Gangwon province
National economic forest
Private economic forest
C 180.0 200.6 156.5
B 129.2 140.7 108.5
Estimation of the Forest Volume of Economic Forests
Stratification of the Average Forest Volume per ha The average forest volume per ha was approximately 159.6 m3/ha and 129.6 m3/ha for national economic forests and private economic forests respectively. The average for- est volume per ha of national economy forests was about 12.7 m3/ha, which was higher than that of Gangwon, but 17.3 m3/ha lower than that of the private economy forests.
On the other hand, for average forest volume per ha per forest type, coniferous forest represented the highest values with approximately 200.6 m3/ha and 156.5 m3/ha for na- tional economic forests and private economic forests, re- spectively, and the rest showed similar patterns in the order of mixed forest and broad-leaved forest. However, the aver- age forest volume per ha per forest type of national econom- ic forests was estimated to be approximately 1.3 times high- er than the average forest volume per ha per forest type of private economic forests (Table 2). Seen by administrative district, while 13 districts were shown to have higher aver- age growing stock per ha in national economic forests than in private ones, such districts as Gosung, Donghae, Yanggu and Jeongseon were shown to have higher average forest volume per ha in private economic forests than in national ones (Fig. 3a). In case of the national economic forests, the highest values, approximately 192.6 m3/ha and 190.8 m3/ha, were shown in Wonju and Hwacheon, respectively, and the lowest value of 107.8 m3/ha was shown in Cheolwon. On the other hand, it is estimated that Yangyang and Taebaek have higher average forest volume per ha in private mixed forest and broad-leaved forest than those of national eco- nomic forests (Fig. 3c, d). As for the private economic for- ests, the highest value of approximately 197.9 m3/ha was re-
corded in Jeongseon, while the lowest value of approx- imately 92.0 m3/ha was shown in Samcheok.
Comparison of Standard Error
Standard error, calculated by applying synthetic estima- tion, was approximately ±2.2 m3/ha for national economic forests and ±2.5 m3/ha for private economic forests, both of which were similar. For standard error per forest type, re- gardless of ownership types, broad-leaved forest is showed to have the lowest values of approximately ±2.6 m3/ha for national economic forests and approximately ±3.6 m3/ha for private economic forests, respectively, and the rest showed similar values in the order of mixed forest and con- iferous forest. As Yim et al. (2009) and Han (2011) also considered the standard error per forest type, calculated by synthetic estimation, as the range of approximately
±3.5-10 m3/ha, which was similar to that of this study. On the other hand, standard error per stratification was within the range of ±4.3-23.1 m3/ha for national economic forests, and within the range of ±5.0-36.2 m3/ha for private eco- nomic forests. In particular, standard error of coniferous forest with age class VI was the highest, which was ±23.1 m3/ha for national economic forests and ±36.2 m3/ha for private economic forests. This result is attributed largely to the fact that the number of sampling plots in age class VI (25 for national and 5 for private economic forests re- spectively) was relatively lower than that of other age classes, and that the range of growing stock per ha was rela- tively wider (58.4-458.3 m3/ha for national forests and 98.5-273.0 m3/ha for private economic forests, respectively (Table 3).
Total Forest Volume
The total forest volume of economic forests in Gangwon calculated by synthetic estimation was approximately 98.71 million m3, consisting of broad-leaved forest (approximately 45%), coniferous forest (approximately 43%), and mixed forest (approximately 13%) in order. In particular, the age class IV of broad-leaved forest and coniferous forest is shown to have relatively higher ratio of the total forest vol- ume with approximately 44%, with approximately 24% of broad-leaved forest and 20% of coniferous forest. The total
Fig. 3. Comparison of forest volume with forest types and administrative district.
Table 3. Estimates of standard error for forest volume (unit: m3/ha)
Distribution of Spatial Forest type Age class
Ⅱ Ⅲ Ⅳ Ⅴ Ⅵ Mean
National economic forest C ±10.1 ±10.5 ±11.1 ±12.5 ±23.1 ±6.3 ±2.2
B ±13.7 ±6.6 ±4.3 ±5.1 ±6.0 ±2.6
M ±4.4 ±8.0 ±7.1 ±9.0 ±10.4 ±4.4
Private economic forest C ±11.4 ±9.6 ±7.6 ±16.9 ±36.2 ±5.2 ±2.5
B ±9.1 ±5.0 ±6.3 ±11.3 ±23.0 ±3.6
M ±11.3 ±6.1 ±7.2 ±18.4 ±28.4 ±4.2
Table 4. Estimates of total forest volume by forest types
(unit: 10,000 m3) Forest type National economic forest Private economic forest
C 2,186 2,033
B 2,957 1,449
M 802 445
Total 5,945 3,926
(approximately 50%), coniferous forest (approximately 37%), and mixed forest (approximate 13%) in order. On the other hand, the total forest volume of private economic forests was estimated to be approximately 1.5 times lower than that of the national economic forests, with coniferous forest (approximately 52%), broad-leaved forest (approximately 37%), and mixed forest (approximately 11%), which repre- sents a different pattern from that of the national economic forests (Table 4). In the total forest volume of economic for- ests per administrative district, the highest value of approx- imately 18.40 million m3 was shown in Hongcheon, which is approximately 41 times higher than the lowest value shown in Donghae. On the other hand, the total forest vol- ume of national economic forests, rated from the highest to the lowest, was in the order of Hongcheon＞Injae＞Pye- ongchang＞Jeongseon (Fig. 4A), and the total forest vol- ume of private economic forests was in the order of Hongcheon＞Pyeongchang＞Hoengseong＞Wonju from the highest to the lowest (Fig. 4B).
Conclusion
The study of economic forest sites in Gangwon as the
in economic forest sites by calculating forest statistics and comparing standard errors, utilizing FTM and NFI 5th. Synthetic estimation was used to calculate average forest volume per ha and total forest volume, and standard error was used to compare error of economic forests per owner- ship. The number of sampling plots was approximately 1.5 times higher for national economic forests compared to that of private economic forests, and it is found that regardless of ownership types, the national economic forest was com- posed in the order of broad-leaved forest＞mixed forest＞
coniferous forest, from the highest to the lowest proportion.
On the other hand, the standard errors of national private economic forests, regardless of ownership types, was in the order of coniferous forest＞mixed forest＞broad-leaved forest, implying that the standard errors tend to decrease as the number of sampling plots increase. The value of aver- age forest volume per ha of economic forest sites estimated by synthetic estimation, was approximately 159.6 m3/ha for national economic forests, which was approximately 30 m3/ha higher than that of private economic forests, and the dis- tribution of economic forest sites was in the order of con- iferous forest＞mixed forest＞broad-leaved forest from the highest to the lowest. On the other hand, in terms of the to- tal forest volume, while the broad-leaved forest showed the highest value of approximately 29.57 million m3 in national economic forests, the coniferous forest showed the highest value of approximately 20.33 million m3 in private econom- ic forests. Overall, the study calculates basic forest statistics of economic forests in Gangwon using FGIS and NFI 5th data and synthetic estimation. The statistics have a sig- nificance in that the forest resources in economic forest zones were estimated, instead of those in administrative dis-
Fig. 4. Distribution of total forest volume with administrative district.
only facilitate more efficient and intensive management of forests, but propose diverse methods for utilization of NFI 5th data as well. In addition, this study is expected to be sub- stantially utilized as basic information in understanding the status of forest resources per space zoning unit, which has been globally required nowadays.
References
Chong SK. 2007. Classification of forest types for timber pro- ductive forestlands using GIS. Korean Journal of Forest Measurements 10: 64-70. (in Korean with English abstract) Cochran WG. 1977. Sampling techniques. 3rd ed. John Wiley &
Sons.
Cressie N. 1993. Statistics for spatial data. Revised ed. John Wiley
& Sons.
Gonzales ME. 1973. Use and evaluation of synthetic estimates.
Proceeding of social statistics section, American Statistical Association 73: 7-15.
Han SH. 2011. Selection of methods for estimation of forest sta- tistics in small area district. Ph.D. thesis, Kookmin University, Seoul, Korea. (in Korean with English abstract)
Hwang JH. 2009. Method of forest statistics estimation using na- tional forest inventory data. Master thesis. Kookmin University, Seoul, Korea. (in Korean with English abstract)
Kim ES, Kim CM. 2015. Estimations of forest growing stocks in small-area level considering local forest characteristics. Journal of Korean Forest Society 104: 117-126. (in Korean with English abstract)
Kim ES, Kim KM, Kim CC, Lee SH, Kim SH. 2010. Estimating
the spatial distribution of forest stand volume in Gyeonggi prov- ince using national forest inventory data and forest type map.
Journal of Korean Forest Society 99: 827-835. (in Korean with English abstract)
Kim H, Kim HJ, Lee SH. 2009. Establishment of target forest type and production goal of timber at economic forest regions in Wanju, Jeonbuk province. Journal of Agricultural & Life Sciences 40: 20-25. (in Korean with English abstract)
Kim JS, Hwang HJ, Shin KI. 2008. Comparison of spatial small area estimators based on neighborhood information system. The Korean Journal of Applied Statistics 21: 855-866. (in Korean with English abstract)
Kim UD. 2012. A study on the management policy to develop in- tensive management forest in Kyongbuk province. Master the- sis, Kyungpook National University, Sangju, Korea. (in Korean with English abstract)
Korea Forest Service. 2013. Study on the improvement of the des- ignation and management system of the forestry promotion area.
(in Korean)
Korea Forest Service. 2015. Statistical yearbook of forestry. (in Korean)
National Institute of Forest Science. 2011a. The 5th national forest inventory report. (in Korean)
National Institute of Forest Science. 2011b. Study on the manage- ment methods for economic forest area and the model for its long-term management plan. (in Korean)
Noh YO. 2009. Study of application for estimator of small area.
Master thesis, Jeju National University, Jeju, Korea. (in Korean with English abstract)
Oh HK, Choi YH, Beon MS. 2008. Study on the vegetation and flora of Nonsan economical forest growing site. The Korea Society For Environmental Restoration And Revegetation
Technology 11: 37-45. (in Korean with English abstract) Yim JS, Han WS, Hwang JH, Chung SY, Cho HK, Shin MY.
2009. Estimation of forest biomass based upon satellite data and national forest inventory data. Korean Journal of Remote Sensing 25: 311-320. (in Korean with English abstract)
Yim JS, Han WS, Jung IB, Kim SH, Shin MY. 2010. Application of synthetic estimator for estimating forest growing stock vol- umes at the small-area level. Journal of Korean Forest Society 99: 285-291. (in Korean with English abstract)
