• Journal of the Korean Association of Geographic Information Studies
• /
• v.24 no.3
• /
• pp.1-9
• /
• 2021

In this study, we evaluated the accuracy of deep learning-based tree species classification model trained by using high-resolution images. We selected five species classed, i.e., pine, birch, larch, korean pine, mongolian oak for classification. We created 5,000 datasets using high-resolution orthophoto and forest type map. CNN deep learning model is used to tree species classification. We divided training data, verification data, and test data by a 5:3:2 ratio of the datasets and used it for the learning and evaluation of the model. The overall accuracy of the model was 89%. The accuracy of each species were pine 95%, birch 89%, larch 80%, korean pine 86% and mongolian oak 98%.

• Korean Journal of Environment and Ecology
• /
• v.32 no.2
• /
• pp.185-202
• /
• 2018

Many ecological studies regarding forest succession, which is the process of arriving at climax forest through ecological changes, have suggested that Carpinus laxiflora is the leading tree type of the climax forest of temperate zone in South Korea. However, most of these studies on C. laxiflora forest have focused on vegetation structure and succession process in specific areas or regional physiognomical forest communities, and thus some may question whether it is rational to consider C. laxiflora forest as the climax forest in South Korea's temperate forest zone. The purpose of this study was to analyze the vegetation classification, species composition, and stratum structure and to investigate the succession tendency with the vegetation data collected from C. laxiflora forests in 75 quadrates in 17 mountains between 2014 and 2017. We used the phytosociological method to analyze the vegetation structure and the importance value to analyze the stratum structure. The results showed the floristic composition of 6 vegetation units and 9 species groups. The hierarchy of the C. laxiflora community group was represented by the Ulmus laciniata community, Corylus heterophylla community (subdivided Quercus aliena group and Corylus heterophylla typical group), and Rhododendron schlippenbachii community (subdivided Fraxinus chiisanensis group, Carpinus tschonoskii group, and Rhododendron schlippenbachii typical group). Successional tendency of C. laxiflora forest based on the importance of each layer is predicted to be the total of three types such as to maintain the stand (vegetation units 2, 3, and 6), to success the Carpinus cordata forest (vegetation units 1 and 4), and to success Quercus acuta forest (vegetation unit 5).

• The Korean Journal of Ecology
• /
• v.6 no.2
• /
• pp.128-141
• /
• 1983

The districbution of $C_3; and; C_4$ type plants in Korea were studied. In the standpoint of photosynthetic types, plant distribution in Korea is classified as $C_3; and; C_4$ type plant zones. The forest destroyed by man interference, cultivating areas, and seashore areas are characterized by the dominant of $C_4$ type plants.(Figs. 2, 3, 4, 5) According to the results of this study, $C_3; and; C_4$ type plant distribution in Korea has a great relation to the habitat of plant vegetation (Table 1). The arid areas were in high proportion of C4 flora percenntages, while the well-developed woody forests or the vegetation of humid areas were in lower proportion(Fig.8).

• Journal of Korean Society of Rural Planning
• /
• v.19 no.2
• /
• pp.121-128
• /
• 2013

The greenhouse gas emission according to the energy consumption is the cause of global warming. With various climates, it is occurs the direct problems to ecosystem. The various studies are being to reduce the carbon dioxide, which accounts for more than 80% of the total greenhouse gas emissions. In this study, estimate the carbon usage using potential biomass extracted from forest type map according to land-use by satellite image, and estimate the amount of carbon dioxide, according to the energy consumption of urban area. The $CO_2$ adsorption is extracted by the amount of forest based on the direct absorption of tree, the other used investigated value. The $CO_2$ emission in Jecheon was 3,985,900 $TCO_2$ by energy consumption. At the land cover classification, the forest is analyzed as 624,085ha and the farmland is 148,700ha. The carbon dioxide absorption was estimated at 1,834,850 Tons from analyzed forest. In case of farmland, it was also estimated at 706,658 Tons.

• Journal of the Korean Geographical Society
• /
• v.9
• /
• pp.67-75
• /
• 1974

The vegetation activity of the Korean peninsula has been monitored temporal variations through a satellite remote sensing and the vegetation index was used to set up the vegetation data map of Korea. The AVHRR data sent by the NOAA-14 satellite was collected for 8 months between April and November, 1997 to calculate the normalized difference vegetation index(NDVI) which was combined the MVC(Maximum Value Composite). Then this NDVI composite map was prepared to review the temporal variations in the vegetation activity. The NDVI has been subject to the unsupervised classification for the growing season between May and October. And the vegetation type is divided into five classes ; urban, bare soil, grass, farming land, deciduous forest and coniferous forest. The unsupervised classificaion of vegetation distribution in the Korean Peninsula shows that the urban and bare soil take 4.14% of total national area, grass 4.49%, farming land 27.54%, deciduous forest 25.61% and coniferous forest 38.22%.

• Korean Journal of Environment and Ecology
• /
• v.24 no.6
• /
• pp.630-637
• /
• 2010

This study was carried out to classify forest vegetation structure of Sobaeksan National Park from May to October in 2008 using the methodology of the ZM school's phytosociological analysis. The forest vegetation was classified into 1 community groups such as Quercus mongolica community group. Quercus mongolica community group was subdivided into 3 community such as Pinus densiflora community, Acer barbinerve community and Cornus controversa community. Pinus densiflora community was subdivided into Quercus variabilis group and Tilia amurensis group. Acer barbinerve community was also subdivided into Salix hallaisanensis group and Betula ermani group(subdivided into Taxus cuspidata subgroup and Betula ermani typical subgroup). Artificial forest type indicated 2 communities such as Larix leptolepis community and Pinus koraiensis community. Accordingly, the vegetation pattern of the surveyed areas were classified into 1 community groups, 5 communities, 4 groups, and 2 subgroups and the forest vegetation was classified into 8 units in total. The vegetation unit distributions was strongly correlated with sea level and topography in this research area.

• Journal of agriculture & life science
• /
• v.46 no.4
• /
• pp.21-30
• /
• 2012

The objective of this study was to comprehend inter-species association and factors affecting species composition by factor analysis and to classify forest types of natural forests in Dootasan. We examined the correlation (positive or negative) of the major species by correlation analysis, the selection of three factors affecting the species composition by factor analysis, cluster analysis on the basis of factor scores, and the evaluation of the results of forest type classification by ANOVA. The outputs of correlation analysis were closely associated with those of factor analysis. The first factor affecting species composition was found to be the decline phenomenon of Pinus densiflora during forest succession process. The second and third factors were growth environments in valley and slope, respectively. The cluster analysis was carried out based on three factors affecting the species composition. The results indicated that the study area was classified into four forest types as follows: Quercus mogolica-Acer mono-Fraxinus rhynchophylla community, Q. mongolica community, Q. mongolica-Tilia amunrensis community and Pinus densiflora community. The dominant species of each community in the four classified forest types were significantly different (p<0.05).

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.10 no.1
• /
• pp.57-71
• /
• 2007

While a concept on biotop or the urgency of its classification systems have been under discussion recently, this study aims to examine outdoor biotop classification systems for environment-friendly community. To this end, the feasibility of creating a biotop in the community and application elements were generated and biotops were classified and categorized. Then, elements that can be applied in consideration of traditional Korean techniques were generated and biotop classification systems and specific components in residential areas were reviewed. As for the result of this study, based on a preliminary draft prepared through literature review, considerations for biotop classification systems were taken into account. Then, based on classification criteria for biotop formats, biotop functions and biotop types, a second-tier classification system was developed. Criteria for biotop formats included surfaces, lines and points while criteria for biotop functions were large cores, small bases, corridors, stepping stones and ecological islands. Criteria for habitat types were divided to include natural forest, developed green areas, lacustrine wetland, palustrine wetland, shrubs, grasslands, linear habitats, vacant plots and practical green areas, which were sub-categorized. As for the biotop classification system, macro-classification divided biotops into three types-space, line and point-based on biotop formats. Meso-classification had five groups and micro-classification had 21 groups based on habitat types. Future studies should focus on the ecological features of each biotop categories generated in this study and their creation and management techniques to find many practical methods to create, protect and manage outdoor biotop for environment-friendly community.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.16 no.1
• /
• pp.131-145
• /
• 2013

The purposes of this research are : to analyze biotop type and carry out conservation value assessment in study areas, Daegu Science Park national industrial complex; to supply basic data for the landscape ecological planning; and to improve the application of assessment model with the development of Biotop Value Assessment Tool (B-VAT). The result is as follows. First of all, the field survey showed 434 species of flora and 220 species of insecta. According to the result of biotop type analysis in the biotop classification system with field survey, 13 biotope groups and 63 biotope types were classified. In the map of biotop type classification, readjusted farmland biotop(FA) was prevalent and forest biotop(E) was shown to the southeast side of the site. Next, according to the first assessment with B-VAT, total 19 biotop types including natural river(BA) with abundant plants had I grade and total 16 biotop types such as vegetable garden adjacent to river(BC) had V grade. In terms of the second assessment, we analyzed total 30 areas, total 82 areas, respectively, which had special meaning for the conservation of species and biotop(1a, 1b) and which had meaning for the conservation of species and biotop(2a, 2b, 2c). This research will be a basic data, which can solve the damage problem systematically and control it landscape-friendly with biotop classification and assessment which we developed. In particular, we expect that biotop value assessment tool(B-VAT) with GIS will be a great contribution to popularity compared with the value model by complicated algorism such as adding-matrix, weight and equal distribution. In addition, this will save the time and improve the accuracy for hand-counting.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.21 no.2
• /
• pp.44-55
• /
• 2018

The importance of establishing a greenhouse gas inventory is emerging for policymaking and its implementation to cope with climate change. Thus, it is needed to establish Approach 3 level Land Use, Land-Use Change and Forestry (LULUCF) matrix that is spatially explicit regarding land use classifications and changes. In this study, four types of spatial information suitable for establishing the LULUCF matrix were analyzed - Cadastral Map, Land Cover Map, Forest Map, and Biotope Map. This research analyzed the classification properties of each type of spatial information and compared the quantitative and qualitative characteristics of the maps in Boryeong city. Drawn from the conclusions of the quantitative comparison, the forest area showed the maximum difference of 50.42% ($303.79km^2$) in the forest map and 46.09%($276.65km^2$) in the cadastral map. The qualitative comparison drew five qualitative characteristics: data construction scope difference, data construction purpose difference, classification standard difference, and classification item difference. As a result of the study, it was evident that the biotope map was the most appropriate spatial information for the establishment of the LULUCF matrix. In addition, if the LULUCF matrix is made by integrating the biotope, the forest map, and the land cover map, the limitations of each spatial information would be improved. The accuracy of the LULUCF matrix is expected to be improved when the map of the level-3 land cover map and the biotope map of 1:5,000 covering the whole country are completed.