As most of the forest fires in South Korea are related to human activity, socio-economic factors are critical in estimating their probability. To estimate and analyze how human activity is influencing forest fire probability, this study considered not only environmental factors such as precipitation, elevation, topographic wetness index, and forest type, but also socio-economic factors such as population density and distance from urban area. The machine learning Maximum Entropy (Maxent) and Random Forest models were used to predict and analyze the spatial distribution of forest fire probability in South Korea. The model performance was evaluated using the receiver operating characteristic (ROC) curve method, and models’ outputs were compared based on the area under the ROC curve (AUC). In addition, a multi-temporal analysis was conducted to determine the relationships between forest fire probability and socio-economic or environmental changes from the 1980s to the 2000s. The analysis revealed that the spatial distribution was concentrated in or around cities, and the probability had a strong correlation with variables related to human activity and accessibility over the decades. The AUC values for validation were higher in the Random Forest result compared to the Maxent result throughout the decades. Our findings can be useful for developing preventive measures for forest fire risk reduction considering socio-economic development and environmental conditions.

Bursaphelenchus xylophilus, the pine wood nematode (PWN) which causes pine wilt disease, is currently a serious problem in East Asia, including in Japan, Korea, and China. This paper investigates the hyperspectral analysis of pine wilt disease to determine the optimal detection indices for measuring changes in the spectral reflectance characteristics and leaf reflectance in the Pinus thunbergii (black pine) forest on Geoje Island, South Korea. In the present study, we collected the leaf reflectance spectra of pine trees infected with pine wilt disease using a hyperspectrometer. We used 10 existing vegetation indices (based on hyperspectral data) and introduced the green-red spectral area index (GRSAI). We made comparisons between non-infected and infected trees over time. A t-test was then performed to find the most appropriate index for detecting pine wilt disease-infected pine trees. Our main result is that, in most of the infected trees, the reflectance changed in the red and mid-infrared wavelengths within two months after pine wilt infection. The vegetation atmospherically resistant index (VARI), vegetation index green (VIgreen), normalized wilt index (NWI), and GRSAI indices detected pine wilt disease infection faster than the other indices used. Importantly, the GRSAI results showed less variability than the results of the other indices. This optimal index for detecting pine wilt disease is generated by combining red and green wavelength bands. These results are expected to be useful in the early detection of pine wilt disease-infected trees.

To assess which data type is more effective for spatial modeling in the Republic of Korea, we conducted geostatistical analysis based on frequency, intensity, and spatial autocorrelation using two types of forest fire occurrence data: that collected through field survey of the Korea Forest Service (KFS) and satellite active fire data of Moderate Resolution Imaging Spectroradiometer (MODIS). The maximum entropy (MaxEnt) model was used with environmental factors in the spatial modeling of fire probability to compare the accuracy of the two data types based on 10 years of historical data. The results showed a clear difference in fire frequency and similar fire intensity patterns. The spatial autocorrelation between the fire frequency and intensity of the two data types was analyzed using a semi-variogram. Fire intensity was significantly correlated, with the MODIS data having a higher correlation than the KFS data. Examination of the spatial autocorrelation and related factors by fire source also indicated that MODIS data had higher spatial autocorrelation, with remarkable distinction found in climate factors. In spatial the modeling, MODIS data showed a similar outcome to that of hotspot analysis, with higher accuracy and better model performance attributable to high spatial autocorrelation. Even though the KFS data were collected from post-fire surveys, they resulted in low spatial autocorrelation and reduced model accuracy owing to the wide distribution of data. MODIS had many detection errors. With spatial filtering, however, the model accuracy can be improved with relatively high spatial autocorrelation.

Afforestation of forests in South Korea may provide an example of the benefit of afforestation on precipitation storage and erosion control. In this study, we presented the effects of afforestation on water supply and soil loss prevention. A spatio-temporal simulation of forest water yield and soil loss was performed from 1971–2010 using InVEST water yield and SWAT models. A forest stock change map was produced by combining land cover data and National Forest Inventory data. The forest water yield increased about twice with changes in forest stock and climate from 1971–2010 and showed a spatially homogeneous water supply capacity. In the same period, the soil loss decreased more than three times, and the volatility of soil loss, in the 2010s, was smaller than before. The analysis of the change in forest stock without considering climate change showed an increase of 43% in forest water yield and a decrease of 87% in soil loss. An increase in precipitation increased the water yield, but also increased the soil loss volume. A change in forest stock led to positive changes in both. This study presents functional positive effects of the afforestation program in South Korea that can be useful in various afforestation programs in other countries.

The applicability of deep learning to remote sensing is rapidly increasing in accordance with the improvement in spatiotemporal resolution of satellite images. However, unlike satellite images acquired in near-real-time over wide areas, there are limited amount of labeled data used for model training. In this article, three kinds of deep learning applications--data augmentation, semisupervised classification, and domain-adapted architecture--were tested in an effort to overcome the limitation of insufficient labeled data. Among the diverse tasks that can be used for classification, rice paddy detection in South Korea was performed for its ability to fully utilize the advantages of deep learning and high spatiotemporal image resolution. In the process of designing each application, the domain knowledge of remote sensing and rice phenology was integrated. Then, all possible combinations of the three applications were examined and evaluated with pixel-based comparisons in various environments and city-level comparisons using national statistics. The results of this article indicated that all combinations of the applications can contribute to increase classification performance, even though the uncertainty involved in imitating or utilizing unlabeled data remains. As the effectiveness of the proposed applications was experimentally confirmed, enhancement in the applicability of deep learning was expected in various remote sensing areas. In particular, the proposed applications would be significant when they are applied to a wide range of study areas and high-resolution images, as they tend to require a large amount of learning data from diverse environments, owing to high intra-class heterogeneity.

Habitat changes in temperate forests are more vulnerable to climate change than tropical or boreal forests. This study assessed forest habitat suitability and diversity to determine the impact of climate change on the Korean Peninsula. We used the MaxEnt (Maximum Entropy) species distribution model, three key climate indices, and two representative climate change scenarios, using short and long-term data. Two of the three key climate indices related to temperature were more capricious than the precipitation-related index in the future. In the baseline prediction, both statistical and qualitative validation using the actual vegetation map showed excellent results. Regarding forest habitat suitability, northward migration and substantial increase were definitely distinctive in warm temperate evergreen forest. On the other hand, subalpine forest areas decreased significantly due to climate change; the suitable area for Representative Concentration Pathways (RCP) 8.5 2070s decreased by more than half. With regard to forest habitat diversity, regions with high diversity declined due to climate change. In the RCP 8.5 scenario, areas where all three forest types are suitable no longer appeared; however, in the case of RCP 4.5 2050s, suitable areas for two forest types increased, which implies climate change is not only negative in terms of diversity. As this negative prediction of future change is discouraging, active mitigation and adaptation are required to prevent these changes. The sustainability of future ecosystems is still dependent on our efforts.