Doo Ahn Kwak

Estimating Crown Variables of Individual Trees Using Airborne and Terrestrial Laser Scanners

In this study, individual tree height (TH), crown base height (CBH), crown area (CA) and crown volume (CV), which were considered as essential parameters for individual stem volume and biomass estimation, were estimated by both an airborne laser scanner (ALS) and a terrestrial laser scanner (TLS). These ALS- and TLS-derived tree parameters were compared because TLS has been introduced as an instrument to measure objects more precisely. ALS-estimated TH was extracted from the highest value within a crown boundary delineated with the crown height model (CHM). The ALS-derived CBH of individual trees was estimated by k-means clustering method using the ALS data within the boundary. The ALS-derived CA was calculated simply with the crown boundary, after which CV was computed automatically using the crown geometric volume (CGV). On the other hand, all TLS-derived parameters were detected manually and precisely except the TLS-derived CGV. As a result, the ALS-extracted TH, CA, and CGV values were underestimated whereas CBH was overestimated when compared with the TLS-derived parameters. The coefficients of determination (R2) from the regression analysis between the ALS and TLS estimations were approximately 0.94, 0.75, 0.69 and 0.58, and root mean square errors (RMSEs) were approximately 0.0184 m, 0.4929 m, 2.3216 m2 and 13.2087 m3 for TH, CBH, CA and CGV, respectively. Thereby, the error rate decreased to 0.0089, 0.0727 and 0.0875 for TH, CA and CGV via the combination of ALS and TLS data.

Forest Cover Classification by Optimal Segmentation of High Resolution Satellite Imagery

This study investigated whether high-resolution satellite imagery is suitable for preparing a detailed digital forest cover map that discriminates forest cover at the tree species level. First, we tried to find an optimal process for segmenting the high-resolution images using a region-growing method with the scale, color and shape factors in Definiens® Professional 5.0. The image was classified by a traditional, pixel-based, maximum likelihood classification approach using the spectral information of the pixels. The pixels in each segment were reclassified using a segment-based classification (SBC) with a majority rule. Segmentation with strongly weighted color was less sensitive to the scale parameter and led to optimal forest cover segmentation and classification. The pixel-based classification (PBC) suffered from the “salt-and-pepper effect” and performed poorly in the classification of forest cover types, whereas the SBC helped to attenuate the effect and notably improved the classification accuracy. As a whole, SBC proved to be more suitable for classifying and delineating forest cover using high-resolution satellite images.

Estimation of carbon storage based on individual tree detection in Pinus densiflora stands using a fusion of aerial photography and LiDAR data

The objective of this study was to estimate the carbon storage capacity of Pinus densiflora stands using remotely sensed data by combining digital aerial photography with light detection and ranging (LiDAR) data. A digital canopy model (DCM), generated from the LiDAR data, was combined with aerial photography for segmenting crowns of individual trees. To eliminate errors in over and under-segmentation, the combined image was smoothed using a Gaussian filtering method. The processed image was then segmented into individual trees using a marker-controlled watershed segmentation method. After measuring the crown area from the segmented individual trees, the individual tree diameter at breast height (DBH) was estimated using a regression function developed from the relationship observed between the field-measured DBH and crown area. The above ground biomass of individual trees could be calculated by an image-derived DBH using a regression function developed by the Korea Forest Research Institute. The carbon storage, based on individual trees, was estimated by simple multiplication using the carbon conversion index (0.5), as suggested in guidelines from the Intergovernmental Panel on Climate Change. The mean carbon storage per individual tree was estimated and then compared with the field-measured value. This study suggested that the biomass and carbon storage in a large forest area can be effectively estimated using aerial photographs and LiDAR data.

Spatio-temporal change in forest cover and carbon storage considering actual and potential forest cover in South Korea

This study analyzes change in carbon storage by applying forest growth models and final cutting age to actual and potential forest cover for six major tree species in South Korea. Using National Forest Inventory data, the growth models were developed to estimate mean diameter at breast height, tree height, and number of trees for Pinus densiflora, Pinus koraiensis, Pinus rigida, Larix kaempferi, Castanea crenata and Quercus spp. stands. We assumed that actual forest cover in a forest type map will change into potential forest covers according to the Hydrological and Thermal Analogy Groups model. When actual forest cover reaches the final cutting age, forest volume and carbon storage are estimated by changed forest cover and its growth model. Forest volume between 2010 and 2110 would increase from 126.73 to 157.33 m3 hm−2. Our results also show that forest cover, volume, and carbon storage could abruptly change by 2060. This is attributed to the fact that most forests are presumed to reach final cutting age. To avoid such dramatic change, a regeneration and yield control scheme should be prepared and implemented in a way that ensures balance in forest practice and yield.

Estimation of forest carbon budget from land cover change in South and North Korea between 1981 and 2010

This paper quantified carbon budget in the past 30 years (1981–2010) and identified the impact of land cover change on carbon dynamics using vegetation integrated simulator for trace gases (VISIT) model. North Korea was converted from carbon sink to source with 10.72 ± 5.18 Tg C yr−1 of net ecosystem production (NEP) in the 1980s, 3.00 ± 7.96 Tg C yr−1 in the 1990s, and −0.46 ± 5.13 Tg C yr−1 in the 2000s. NEP in South Korea was 10.55 ± 1.09 Tg C yr−1 in the 1980s, 10.47 ± 7.28 Tg C yr−1 in the 1990s, and 6.32 ± 5.02 Tg C yr−1 in the 2000s, showing a gradual decline. In North Korea, NEP was decreased by 0.52 Tg yr−1 in the 1990s due to reduction of forest, and increased by 0.36 Tg yr−1 in the 2000s due to expansion of cropland. In South Korea, it was decreased by 0.24 Tg yr−1 in the 1990s as urban and built-up area expanded, and increased by 0.04 Tg yr−1 in the 2000s with the expansion of forest. These results suggest the importance of forest and land cover management against deforestation for ensuring national carbon balance.

Evaluation for damaged degree of vegetation by forest fire using lidar and a digital aerial photograph.

The amount of vegetation physically damaged by forest fire can be evaluated using lidar (Light Detection And Ranging) data because the loss of canopy height and width by forest fire can be relevant to the number of points transmitted to the ground through the canopy of the damaged forest. On the other hand, the biological damage of vegetation caused by forest fire can be obtained from the Normalized Difference Vegetation Index (NDVI), which determines the vegetation vitality. In this study, the degree of physical damage from the lidar data was classified into serious physical damage (SPD) and light physical damage (LPD). The degree of biological damage using NDVI was likewise classified into serious biological damage (SBD) and light biological damage (LBD). Finally, the damaged area was graded into four categories: (a) SPD and SBD, (b) LPD and SBD, (c) SPD and LBD, and (d) LPD and LBD. The accuracy assessment for the area classified into four grades showed an overall accuracy of 0.74, and a kappa value of 0.61 which provides improvement over previous works.

Application of the Savitzky-Golay Filter to Land Cover Classification Using Temporal MODIS Vegetation Indices

In this study, the Savitzky-Golay filter was applied to smooth observed unnatural variations in the temporal profiles of the Normalized Difference Vegetation Index (NDVI) and the Enhanced Vegetation Index (EVI) time series from the MODerate Resolution Imaging Spectroradiometer (MODIS). We computed two sets of land cover classifications based on the NDVI and EVI time series before and after applying the Savitzky-Golay filter. The resulting classification from the filtered versions of the vegetation indices showed a substantial improvement in accuracy when compared to the classifications from the unfiltered versions. The classification by the EVIsg had the highest K (0.72) for all classes compared to those of the EVI (0.67), NDVI (0.63), and NDVIsg (0.62). Therefore, we conclude that the EVIsg is best suited for land cover classification compared to the other data sets in this study.

Estimating plot volume using lidar height and intensity distributional parameters

This study explored the feasibility of height distributional metrics and intensity values extracted from low-density airborne light detection and ranging (lidar) data to estimate plot volumes in dense Korean pine (Pinus koraiensis) plots. Multiple linear regression analyses were performed using lidar height and intensity distributional metrics. The candidate variables for predicting plot volume were evaluated using three data sets: total, canopy, and integrated lidar height and intensity metrics. All intensities of lidar returns used were corrected by the reference distance. Regression models were developed using each data set, and the first criterion used to select the best models was the corrected Akaike Information Criterion (AICc). The use of three data sets was statistically significant at R2 = 0.75 (RMSE = 52.17 m3 ha−1), R2 = 0.84 (RMSE = 45.24 m3 ha−1), and R2 = 0.91 (RMSE = 31.48 m3 ha−1) for total, canopy, and integrated lidar distributional metrics, respectively. Among the three data sets, the integrated lidar metrics-derived model showed the best performance for estimating plot volumes, improving errors up to 42% when compared to the other two data sets. This is attributed to supplementing variables weighted and biased to upper limits in dense plots with more statistical variables that explain the lower limits. In all data sets, intensity metrics such as skewness, kurtosis, standard deviation, minimum, and standard error were employed as explanatory variables. The use of intensity variables improved the accuracy of volume estimation in dense forests compared to prior research. Correction of the intensity values contributed up to a maximum of 58% improvement in volume estimation when compared to the use of uncorrected intensity values (R2 = 0.78, R2 = 0.53, and R2 = 0.63 for total, canopy, and integrated lidar distributional metrics, respectively). It is clear that the correction of intensity values is an essential step for the estimation of forest volume.

Unconstrained approach for isolating individual trees using high-resolution aerial imagery

This study outlines an algorithm that can be used for individual tree detection and crown delineation; it was applied to coniferous forest using aerial imagery. This article explains the assumptions and processes involved in the algorithm, presents the results of the applications, and discusses possible limitations. The algorithm, which adopts contextual analysis that excludes the need to specify window size, was applied to detect and delineate individual trees based on morphological and reflective characteristics. The preprocessing steps included suppression of the non-coniferous area (i.e. non-forest and leaf-off deciduous forest) and the creation of appropriately smoothed imagery using an optimal smoothing level based on accuracy index (AI); thereafter, unconstrained directional peak- and edge-finding algorithms were processed separately. To assess the tree detection and crown delineation processes, the results of the algorithms were evaluated carefully against visually interpreted crowns in six square plots using several statistical measures based on tree top correspondence, positional difference of tree top, directional crown width, and crown area assessment. The average tree top correspondence had an AI of 88.83%. The positional difference between detected and visually interpreted tree tops was measured and its average was 0.6 m. For our 0.5 m/pixel aerial imagery, the average root mean square error (RMSE) of crown width in six sample plots was found to be 2.8 m, and crown area estimation resulted in RMSE of approximately 9.23 m2 (23.25%). In general, this study highlights the potentiality of the proposed algorithm to efficiently and automatically acquire forest information such as tree numbers, crown width, and crown area.

Forest plot volume estimation using National Forest Inventory, Forest Type Map and Airborne LiDAR data

The importance of estimating forest volume has been emphasized by increasing interest on carbon sequestration and storage which can be converted from volume estimates. With importance of forest volume, there are growing needs for developing efficient and unbiased estimation methods for forest volume using reliable data sources such as the National Forest Inventory (NFI) and supplementary information. Therefore, this study aimed to develop a forest plot volume model using selected explanatory variables from each data type (only Forest Type Map (FTM), only airborne LiDAR and both datasets), and verify the developed models with forest plot volumes in 60 test plots with the help of the NFI dataset. In linear regression modeling, three variables (LiDAR height sum, age, and crown density class) except diameter class were selected as explanatory independent variables. These variables generated the four forest plot volume models by combining the variables of each data type. To select an optimal forest plot volume model, a statistical comparing process was performed between four models. In verification, Model no. 3 constructed by both FTM and airborne LiDAR was selected as an optimal forest plot volume model through comparing root mean square error (RMSE) and coefficient of determination (R 2). The selected best performance model can predict the plot volume derived from NFI with RMSE and R 2 at 50.41 (m3) and 0.48, respectively.