• Journal of the Korean Association of Geographic Information Studies
• /
• v.7 no.4
• /
• pp.57-64
• /
• 2004

In order to decrease the area damaged by forest fires and to prevent the occurrence of forest fires, the forest fire danger rating system was developed to estimate forest fire risk by means of weather, topography, and forest type. Forest fires occurrence prediction needs to improve continually. Logistic regression and spatial analysis was used in developing the forest fire occurrence probability model. The forest fire danger index in accordance to the probability of forest fire occurrence was used in the classification of forest fire occurrence risk regions.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.18 no.4
• /
• pp.199-207
• /
• 2016

This study was conducted to develop a forest fire occurrence model using meteorological characteristics for practical forecasting of forest fire danger rate by reflecting the climate change for the time period of 2000yrs. Forest fire in South Korea is highly influenced by humidity, wind speed, temperature, and precipitation. To effectively forecast forest fire occurrence, we developed a forest fire danger rating model using weather factors associated with forest fire in 2000yrs. Forest fire occurrence patterns were investigated statistically to develop a forest fire danger rating index using times series weather data sets collected from 76 meteorological observation centers. The data sets were used for 11 years from 2000 to 2010. Development of the national forest fire occurrence probability model used a logistic regression analysis with forest fire occurrence data and meteorological variables. Nine probability models for individual nine provinces including Jeju Island have been developed. The results of the statistical analysis show that the logistic models (p<0.05) strongly depends on the effective and relative humidity, temperature, wind speed, and rainfall. The results of verification showed that the probability of randomly selected fires ranges from 0.687 to 0.981, which represent a relatively high accuracy of the developed model. These findings may be beneficial to the policy makers in South Korea for the prevention of forest fires.

• Fire Science and Engineering
• /
• v.21 no.4
• /
• pp.52-58
• /
• 2007

The effect of stand-growing-stock characteristics of thinning area and non-thinning area on forest fire was studied in this work. 14 spots were selected from 3 counties such as Yangyang, Injae, and Gapyeong and on-the-spot investigations were performed to evaluate the effect of forest fire. The stand-growing-stock characteristics on the spots were analyzed through the height of tree, breast height diameter, clear length, mortality of branch, forest tree standing crop density, degree of closure, and shrub and grass cover degree. The relation between forest fire and the risk of spread of forest fire were analyzed from the analysis of the stand-growing-stock characteristics. It is considered from this work that the possibility of forest fire is decreased on the thinning area compared to the non-thinning area because of higher clearlength, lower number of tree, lower mortality of branch and higher shrub and grass cover degree.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.21 no.3
• /
• pp.189-204
• /
• 2018

This study has developed a forest fire occurrence probability model for inaccessible areas such as North Korea and Demilitarized Zone and we have developed a real-time forest fire danger rating system that can be used in fire-related works. There are limitations on the research that it is impossible to conduct site investigation for data acquisition and verification for forest fire weather index model and system development. To solve this problem, we estimated the fire spots in the areas where access is impossible by using MODIS satellite data with scientific basis. Using the past meteorological reanalysis data(5㎞ resolution) produced by the Korea Meteorological Administration(KMA) on the extracted fires, the meteorological characteristics of the fires were extracted and made database. The meteorological factors extracted from the forest fire ignition points in the inaccessible areas are statistically correlated with the forest fire occurrence and the weather factors and the logistic regression model that can estimate the forest fires occurrence(fires 1 and non-fores 0). And used to calculate the forest fire weather index(FWI). The results of the statistical analysis show that the logistic models(p<0.01) strongly depends on maximum temperature, minimum relative humidity, effective humidity and average wind speed. The logistic regression model constructed in this study showed a relatively high accuracy of 66%. These findings may be beneficial to the policy makers in Republic of Korea(ROK) and Democratic People's Republic of Korea(DPRK) for the prevention of forest fires.

• Journal of Korean Society of Forest Science
• /
• v.85 no.1
• /
• pp.15-23
• /
• 1996

Preparing the era of forest resources management requires studies on forest fire. This study attempted to develop forest fire occurrence model using meteorological characteristics for the practical purposes of forecasting forest fire danger rate. To accomplish this goal, the relationships between forest fire occurrence and meteorological characteristics are estimated. In the process, the forest fire occurrence pattern of the study region(Taegu-Kyungpook) is categorized by employing qualification IV method. The study region was divided into three areas such as, Taegu, Andong and Pohang area. The meteorological variables emerged as affective to forest fire occurrence are relative humidity, longitude of sunshine, and duration of precipitation. To estimate the probability of forest fire danger, forest fire occurrence of three areas are regressed on the time series data of affective meteorological variables using logistic and probit model. The effectiveness of the models estimated are tested and showed acceptable degree of goodness. Those models developed would be helpful to increase the efficiency of forest fire management such as detection of forest fire occurrence and effective disposition of forest fire fight equipments.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.13 no.3
• /
• pp.148-156
• /
• 2011

The purpose of this study is to analyze the patterns of forest fire possibly related with Asian dust event and carry out a correlation analysis between forest fire occurrence and existence or not of the Asian dust event during dry seasons i.e. February to May in 2000 to 2008. To study the correlation of forest fire and Asian dust, we surveyed information of Asian dust observations, forest fire statistics, fire danger rating index, weather data such as temperature, relative humidity and wind speed of the day occurring the forest fire. As a consequence of analysis, the regional frequency of Asian dust was the highest in Gyeonggi and Chungbuk divisions. Frequencies of forest fire occurrence by the Asian dust events were the highest in the day before three days of the Asian dust event. The highest frequent regions of forest fire occurrence were district of boundary line between Gyeonggi and Western of Gangwon, Chungbuk and Gyeonbuk inland. The correlation between forest fire and fire danger rating index showed the high correlation with the day before three days and after three days of the Asian dust event. These correlation coefficients were 0.50038 and 0.53978 to 1% significance level. The result of analysis between the frequency of forest fire occurrence and wind speed had a highly negative relationship at all the Asian dust days, the day before and after three days. The correlation coefficients had been -0.58623 to -0.61245 to 1% significance level. Relative humidity showed a little of negative relationship with forest fire occurrence in -0.2568(p ${\leq}$ 0.01) for the Asian dust day and -0.35309(p ${\leq}$ 0.01) for next three days. Moreover, at the day before three days of Asian dust events, it was -0.23701 to 1% significance level. However, the mean temperature did not correlate with frequency of forest fire occurrence by Asian dust events at all.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.20 no.4
• /
• pp.348-356
• /
• 2018

Most of fires were human-caused fires in Korea, but meteorological factors are also big contributors to fire behavior and its spread. Thus, meteorological factors as well as topographical and forest factors were considered in the fire danger rating systems. This study aims to develop an advanced national integrated daily weather index(DWI) using weather data in the spring and fall to support forest fire prevention strategy in South Korea. DWI represents the meteorological characteristics, such as humidity (relative and effective), temperature and wind speed, and we integrated nine logistic regression models of the past into one national model. One national integrated model of the spring and fall is respectively $[1+{\exp}\{-(2.706+(0.088^*T_{mean})-(0.055^*Rh)-(0.023^*Eh)-(0.014^*W_{mean}))\}^{-1}]^{-1}$, $[1+{\exp}\{-(1.099+(0.117^*T_{mean})-(0.069^*Rh)-(0.182^*W_{mean}))\}^{-1}]^{-1}$ and all weather variables significantly (p<0.01) affected the probability of forest fire occurrence in the overall regions. The accuracy of the model in the spring and fall is respectively 71.7% and 86.9%. One integrated national model showed 10% higher accuracy than nine logistic regression models when it is applied weather data with 66 random sampling in forest fire event days. These findings would be necessary for the policy makers in the Republic of Korea for the prevention of forest fires.

• 한국방재학회:학술대회논문집
• /
• 2007.02a
• /
• pp.217-222
• /
• 2007

Since 1973, we attain a successful achievement of nation-wide afforestation such as a thick forest and heaped-up leaves. However, the higher of the formation density in forest, the more dangerous to be a large-scale forest fire whenever fire occurs. According to the type of forest in the country, 42% of the forest is occupied by conifer forest that are highly flammable, and the distribution of forest age is in a transition period from immature forest to mature one. And the structure is too weak to the forest fire for the occurrence and spread because there are too many scrub and shrub trees in the forest. As a matter of course, it is on the increase of the thinning-forest that can shift the forest structure from a weak on forest fire to a strong one nowaday. In other words, thinning-forest has primary purposes such as the promotion of producing forest trees, production of excellent timbers, and build-up of public forest area. Furthermore, in some reports, the reduction of ladder fuel by eliminating the vertical/horizontal fuel in a forest and ensuring spaces in the forest can decrease the occurrence of forest fire and the risk of spread of burning as by-effect. Therefore, this study is designed to clarify the relation with the risk of forest fire by an on-spat-investigation of the characteristics of forest composition on the thinning and the non-thinning area.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.9 no.3
• /
• pp.95-100
• /
• 2009

Fine fuel moisture code (FFMC), a main component of forest fire weather index(FWI) in the Canadian forest fire danger rating system(CFFDRS), indicated a probability of ignition through expecting a dryness of fine fuels. According to this code, a rising of temperature and wind velocity, a decreasing of precipitation and decline of humidity in a weather condition showed a rising of a danger rate for the forest fire. In this study, we analyzed a weather condition during 5 years in Kangwon province, calculated a FFMC and examined an application of FFMC. Very low humidity and little precipitation was a characteristic during spring and fall fire season in Kangwon province. 75% of forest fires during 5 years occurred in this season and especially 90% of forest fire during fire season occurred in spring. For developing of the prediction model for a forest fire occurrence probability, we used a logistic regression function with forest fire occurrence data and classified mean FFMC during 10 days. Accuracy of a developed model was 63.6%. To improve this model, we need to deal with more meteorological data during overall seasons and to associate a meteorological condition with a forest fire occurrence with more research results.

• Proceedings of The Korean Society of Agricultural and Forest Meteorology Conference
• /
• 2015.08a
• /
• pp.126-130
• /
• 2015