In this study, a high-resolution daily data set of surface weather were obtained from PRIDE(PRISMbased Dynamic downscaling Error correction) model for the period of 2000 to 2017 over South Korea. The simulation data of five RCM(Regional Climate Model) were also used which are forced by the CMIP6 participating model UK-ESM as the boundary condition under historical period (2000-2014) and SSP 5-8.5 period (2015- 2017). Here we compared the RCM data and the PRIDE data with MK-PRISM data in terms of ensemble mean and ensemble spread. Results show that the PRIDE model effectively eliminates systematic error in the RCM up to 63.0% for daily average temperature, 72.2% for daily maximum temperature, 68.2% for daily minimum temperature, and 28.7% for daily precipitation when evaluated from the RMSE perspective. Overall, the ensemble spread of the PRIDE model is significantly decreased from 1.46°C to 0.36°C for daily temperature and from 2.0 mm/day to 0.72 mm/day for daily precipitation compared to the RCM ensemble spread, indicating that the largest systematic error of the RCMs is effectively removed in the PRIDE model.

In this study, we produced grid climate data sets of 1km×1km and 5km×5km horizontal resolutions based on MK (Modified Korean)-PRISM (Parameter-elevation Regressions on Independent Slopes Model), a statistical downscaling method that can estimate grid data of horizontal high-resolution using observational station data in Korea. To compare the MK-PRISM performance according to resolution, RMSEs of 1km resolution data and 5km resolution data were calculated and analyzed. The RMSEs of the two data sets were similar, but the results classified according to the elevation were different. The 1km high resolution estimated data was shown to better reflect the impact of the terrain for the daily mean temperature and daily maximum temperature, whereas the difference between the two data sets for daily minimum temperature was not statistically significant at each elevation. Furthermore, we also divided the temperature data into 9-classes based on the observed temperatures, and then compared the estimated performance of the two data sets according to elevation. For the low temperature group, performance of the 1km resolution data at high elevations outperformed that of the 5km resolution data, regardless of the season.

MK-PRISM developed for wind interpolation was applied to case studies and was verified in previous studies. Thus, some tests were necessary before the model used in order to produce wind speed maps for the whole area of South Korea. In this study, the MK-PRISM was applied to producing wind speed maps of South Korea. The result showed that sharp changes occur in wind speed distribution, despite the continuous similar topographic. The primary reason for the phenomenon was that a linear regression slope between elevation and wind speed used in interpolation process was changed rapidly in some areas. This study used the landform classification data to address this problem. The improved model controlled similarly the slope of the linear regression equation in the continuous valley, slope, and ridge. Therefore, the slope of the linear regression equation does not change dramatically in the improved model. The improved model was named MK-PRISM-Wind in this study. The wind speed was similar on the ridge continuously in the wind speed distribution produced by MK-PRISM-Wind. In addition, the wind speed was more gradually changed compared to the previous model on the plains and foothills. The results mean that MK-PRISM-Wind can produce wind speed maps more reasonable than the previous model, and it can be applied to the wind speed interpolation of South Korea. High-resolution gridded wind speed map produced by MK-PRISM-Wind is expected to be utilized for various studies.

This study investigates the theoretical background of the interpolation methods that regards the topographical effect on the climate data, such as Co-kriging, Artificial Neural Network and MK-PRISM(Modified Korean Parameter-elevation Regressions on Independent Slopes Model). Prior to applying the MK-PRISM to the interpolation of wind speed, this study has improved the model to be closer to the fundamental concept of the PRISM and verified it‘s validity. Since each method has individual advantages and disadvantages, there will be a need for comparative studies in order to select an interpolation method that is suitable for the topography of Korea. This study has added a weighted value that considers the existence of clusters at the known point, and has supplemented the digital elevation models and aspects distribution of multiple scales for application. In addition, this study has allowed the consideration of sharp changes between the known point and unknown point when calculating the topographic facet weighting. The supplement model was verified through the interpolation of rainfall in Jeju Island. The coefficient of determination and KGE(Kling and Gupta Efficiency) of the model displayed the results of 0.86 and 0.87, respectively for August 2010 monthly precipitation in Jeju Island, and the model was accordingly verified. This study is able to provide the necessary information to the researchers who wish to interpolate the observation data of wind speed. Furthermore, the supplement MK-PRISM becomes available to the research on the interpolation of wind speed.

In this study, we developed IS-HYPS(Independent Slopes Hypsometric) method in order to complement the weakness of existing MK-PRISM (Modified Korean-Parameter-elevation Regressions an Independent Slopes Model) version 1.1, and verified the IS-HYPS by applying the method to daily temperature data for recent 11 years from 2000 to 2010. Analysis Results show that IS-HYPS method is nicely applicable in regions where elevation of target grid is high and elevation of surrounding stations are not distributed evenly. Verification results also show that RMSE (Root Mean Square Error) of IS-HYPS method is about 20 % smaller than that of MK-PRISM 1.1 (0.44°C for daily mean temperature, 0.47°C for daily maximum temperature, and 0.58°C for daily minimum temperature) and have a trend to estimate temperature lower than MK-PRISM 1.1. The favorable condition to apply the method found to be the regions where the difference in standard deviation of elevation of observation stations is larger than 70m between with and without target grid inside the searching range of radius, indicating that MK-PRISM version 2.0 can produce temperature grid data reasonably well by applying the IS-HYPS selectively to the best condition.