This study aimed to examine the applicability of Self-organizing Map(SOM) method for synoptic climatology studies over the Republic of Korea through a case study on classification of winter atmospheric pressure patterns. When SOM was conducted using spatial range 10~70°N, 80~180°E, 4×3 map size, linear initialization, batch training algorithm and neighborhood function ep(Epanechnikov function) for the case study, classified atmospheric pressure patterns showed the best result. The classified atmospheric patterns were similar to surface weather charts of Korea Meteorological Administration(KMA). Also all input vectors were classified into one atmospheric pressure pattern(node) so it is possible to analyse time series and characteristics of climatic elements according to atmospheric pressure patterns. Therefore, SOM method might be appropriate for the studies on synoptic climatology over the Republic of Korea.
In this study we have produced daily maximum temperature, daily mean temperature, daily minimum, and daily precipitation data with 1 km horizontal resolution in daily time scale for 10 years from 2001 to 2010 using MK(Modified Korean)-PRISM(Parameter-elevation Regressions an Independent Slopes Model). Using the data produced in this study, we have investigated topographic effect and data histogram, and we have verified MK-PRISM using cross validation and grid point validation.
Daily temperature and precipitation estimated by MK-PRISM have topographic effect which shows an decrease of daily mean temperature and an increase of daily precipitation at higher altitude. Histogram analysis result for daily mean temperature shows the distribution shift toward colder temperature due to temperature lapse rate with altitude at mountain area compared to observed daily mean temperature. On the other hand, no precipitation and light rain less than 1mm per day occur less frequently than observation due to rain shadow effect, while precipitation over 1mm per day occur more frequently than observation due to topographic effect.
In this study, long term simulations for current(1979-2010) and future(2019-2050) climate over North-East Asian region were performed using RegCM4 driven by HadGEM2-AO provided by the KMA(Korean Meteorological Administration). The spatial resolution is 12.5㎞ and two RCP scenarios(4.5, 8.5) are applied for the future climate simulation. The RegCM4 simulates well the spatial and temporal variations of air temperature of current climate. However, it weakly simulates the intensity of rain band associated with the seasonal march of the East Asian summer monsoon and thereby significantly under simulates the summer rainfall and fails in reproducing the seasonal and spatial variations of precipitation. The 20-year averaged differences between current(1986-2005) and future(2031-2050) simulations showed that the temperature increases are generally greater in RCP8.5(1.93K) than RCP4.5(1.86K) and all the changes are statistically significant at 1% level. In general, the temperature increases are greater in the northern region(autumn, winter) than in the southern region(spring, summer) irrespective of RCP scenarios. The temperature over South Korea is expected to increase by 1.53K to 1.87K irrespective of RCP scenarios and seasons. The precipitation changes vary significantly according to the season and region irrespective of RCP scenarios and the changes of spring and autumn in the certain regions are statistically significant at 5% level. The precipitation over South Korea is expected to increase by +0.42㎜/day in RCP4.5 during spring but it is expected to significantly decrease during autumn(-0.35㎜/day) and in RCP 8.5 during summer(-0.30㎜/day).
The CMDPS(COMS Meteorological Data Processing System) provides 16 meteorological products by processing satellite data obtained from 1 visible and 4 infrared channels. It is expected that the COMS(Communication, Ocean and Meteorological Satellite) products will contribute to numerical weather prediction and climate change monitoring. For more active utilization of the products, we first need a scheme for the quality management based on outlier detection. This paper describes a method for detecting spatial and temporal outliers from the COMS products using Moran scatterplot and wavelet transform. The applicability of our method was made sure through the tests for seven products such as sea surface temperature, land surface temperature, aerosol optical depth, total precipitable water, upper tropospheric humidity, cloud top height, and cloud top temperature. Their quality was estimated very good in terms of our spatio-temporal statistical approach. For future work, we need more comprehensive study on the outlier detection, particularity focusing on the threshold setting in accordance with physical characteristics of each product.
This study conducted monitoring of plant species vulnerable or sensitive to Meteorological change—pine trees, azaleas, royal azaleas, sugar maple trees, and Japanese larch trees in Yeongnam region of South Korea for bioclimatological research, and predicted and managed changes in the growth environment of plants in advance, thereby presenting measures to stably preserve plant genetic resources. GPS coordinates of each of the species in each area were marked and changes in their time of blossoming, blossoms falling, leaves unfolding, leaves reddening, and leaves falling were recorded. They were examined once per week from late March 2010 to late November 2011. Indicator plants were analyzed and the time of blossoming tended to be later this year than the previous year and mostly in gravel areas their time of blossoming became later. Their blossoms falling time had a similar pattern to that of their blossoming time. There were no obvious differences in the pattern of leaves unfolding. The time of leaves unfolding of all the investigated species except for pine trees was latest in gravel areas. Sugar maple trees’ blossoms falling time became later in most areas, without regularity, and azaleas, royal azaleas, and Japanese larch trees were observed earliest in Gyeongnam Arboretum. The time of leaves falling became later in most areas. As the temperature became higher in late 2011 relative to 2010, the time of leaves reddening, and leaves falling became later and the growth period of all the species became longer except for pine trees.
In this study, the ultra-high resolution ground information database (30m × 30m), such as elevation map, facet map, coastal map, land cover map, was constructed over Korean Peninsula. ASTER GDEM with 30m resolution was used to generate elevation map, facet map, coastal map, and the accuracy of GDEM was validated using DEM constructed with 1:25,000 digital map. The facet map was generated with 8 direction and flat area using GDEM. The coastal map with 6 categories was generated by buffering of the distance from coast line, additionally considered with elevation. The land cover map was generated with Landsat ETM+ 24 scenes (around 2000’s) by supervised classification, the land cover classes was composed with urbanization, agriculture, green field, forest, tidal flat, bare land, water area. The file format of ground information database is 8-bit or 16-bit unsigned Geotiff, the image size is 27,331 × 40,858 pixels, and the file size is 1.04GB or 2.08GB. The coordinate system composed of UTM projection and WGS84 ellipsoid was applied to the database for the equal grid resolution. This ultra-high resolution ground information database will be able to provide a basis for regional climate modelling and forecasting accuracy enhancement.