Ki-Hong Park

The Korea Meteorological Administration (KMA) has been operating the Korean Integrated Model (KIM), an indigenous global numerical weather prediction system, since April 2020. In May 2025, KMA began the operational service of a high-resolution KIM, which features improved horizontal resolution to enhance overall forecast performance and prediction accuracy. Despite these technical advancements, KIM revealed continuous systematic temperature biases in the high latitudes of the Northern Hemisphere (≥70°N), particularly over sea-ice areas.In this study, we evaluated the systematic errors of KIM by comparing its forecasted fields against ERA5 reanalysis data. We focused on the diagnostic analysis of surface skin temperature (Tsfc), 2m temperature (T2m) and lower-tropospheric air temperature to identify the specific characteristics and seasonal trends of these biases. Preliminary results showed that KIM underestimated surface temperatures over sea-ice areas throughout most of the year, although a transition to overestimation was observed during the spring period from mid-April to May. These biases suggest that the initialization and physical processes for surface conditions in the polar areas within KIM require further research and development.To diagnose the sources of these errors, we analyzed the relationships between temperature biases and radiation/cloud-related variables using time-series analyses. The results indicated that KIM generally underestimated surface downward and net longwave radiation, while overestimating surface net shortwave radiation. In addition, cloud fraction was overestimated, whereas total cloud condensate was generally underestimated and the fraction of cloud ice was relatively large. This study aims to examine how these discrepancies in hydrometeors and radiation variables contribute to the observed temperature biases, especially over sea-ice areas. The findings from this evaluation will provide a basis for improving physical parameterizations and surface-related processes in KIM to enhance the predictive skill of the high-resolution KIM in high-latitude areas.

Soil moisture (SM) is an indicator of the moisture status of the land surface, which is useful for monitoring extreme weather events. Representative global SM datasets include the National Aeronautics and Space Administration (NASA) Soil Moisture Active Passive (SMAP), the Global Land Data Assimilation System (GLDAS), and the European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis 5 (ERA5), but due to their low spatial resolutions, none of these datasets well describe SM changes in local areas, and they tend to have a low accuracy. Machine learning (ML)-based SM predictions have demonstrated high accuracy, but obtaining semi-real-time SM information remains challenging, and the dependence of the validation accuracy on the data sampling method used, such as random or yearly sampling, has led to uncertainties. In this study, we aimed to develop an ML-based model for real-time SM estimation that can capture local-scale variabilities in SM and have reliable accuracy, regardless of the sampling method. This study was conducted in South Korea, and satellite image data, numerical weather prediction (NWP) data, and topographic data provided within one day were used as the input data. For SM modeling, 13 input variables affecting the surface SM status were selected: 10- and 20-day cumulative standardized precipitation indexes (SPI10 and SPI20), a normalized difference vegetation index (NDVI), downward shortwave radiation (DSR), air temperature (Tair), land surface temperature (LST), soil temperature (Tsoil), relative humidity (RH), latent heat flux (LE), slope, elevation, topographic ruggedness index (TRI), and aspect. Then, SM models based on random forest (RF) and automated machine learning (AutoML) were constructed, trained, and validated using random sampling and leave-one-year-out (LOYO) cross-validation. The RF- and AutoML-based SM models had significantly high accuracy rates based on comparisons with in situ SM (mean absolute error (MAE) = 2.212–4.132%; mean bias error (MBE) = −0.110–0.136%; root mean square error (RMSE) = 3.186–5.384%; correlation coefficient (CC) = 0.732–0.913), while the AutoML-based SM model tended to have a higher accuracy than the RF-based SM model, regardless of the data sampling method used. In addition, when compared to in situ SM data, the SM models demonstrated the highest accuracy, outperforming both GLDAS and ERA5 SM data and well representing changes in the dryness/wetness of the land surface according to meteorological events (heatwave, drought, and rainfall). The SM models proposed in this study can, thus, offer semi-real-time SM data, aiding in the monitoring of moisture changes in the land surface, as well as short-term meteorological disasters, like flash droughts or floods.

Remote sensing data such as weather satellites and radars with the high spatio-temporal resolution are importantly used for severe weather monitoring. Geostationary meteorological satellites can observe clouds in a wide area, and the weather radar can obtain rainfall intensity information close to the ground truth through raindrop particle observation. These remote sensing data are collected as real-time images, which are suitable for application to image convolution-based artificial intelligence (AI) models. In this study, we developed an AI-based model for the generation of proxy radar data from the Geo-Kompsat-2A (GK2A) satellite data. As an AI model structure, an image-to-image translation technique called Pix2PixCC which is based on a conditional Generative Adversarial Network (cGAN) was used. For the training model, GK2A satellite data and composite data of radar reflectivity with Korea, Japan, and China were used. In order to solve the memory issue of AI model learning using high-performance GPU, the model was trained by dividing the image into 256x256 patch sizes. This AI model made it possible to produce real-time proxy radar data without gaps in the East Asian region, and this proxy data showed a similar shape to the real radar data compared to precipitation information from geostationary meteorological satellites in traditional methods. In future studies, we plan to improve the accuracy by applying a custom loss function that applies a high weight to the high reflectivity of radar.

Surface solar irradiance (SSI) is a crucial component in climatological and agricultural applications. Because the use of renewable energy is crucial, the importance of SSI has increased. In situ measurements are often used to investigate SSI; however, their availability is limited in spatial coverage. To precisely estimate the distribution of SSI with fine spatiotemporal resolutions, we used the GEOstationary Korea Multi-Purpose SATellite 2A (GEO-KOMPSAT 2A, GK2A) equipped with the Advanced Meteorological Imager (AMI). To obtain an optimal model for estimating hourly SSI around Korea using GK2A/AMI, the convolutional neural network (CNN) model as a machine learning (ML) technique was applied. Through statistical verification, CNN showed a high accuracy, with a root mean square error (RMSE) of 0.180 MJ m⁻², a bias of −0.007 MJ m⁻², and a Pearson’s R of 0.982. The SSI obtained through a ML approach showed an accuracy higher than the GK2A/AMI operational SSI product. The CNN SSI was evaluated by comparing it with the in situ SSI from the Ieodo Ocean Research Station and from flux towers over land; these in situ SSI values were not used for training the model. We investigated the error characteristics of the CNN SSI regarding environmental conditions including local time, solar zenith angle, in situ visibility, and in situ cloud amount. Furthermore, monthly and annual mean daily SSI were calculated for the period from 1 January 2020 to 31 January 2022, and regional characteristics of SSI around Korea were analyzed. This study addressed the availability of satellite-derived SSI to resolve the limitations of in situ measurements. This could play a principal role in climatological and renewable energy applications.

Evapotranspiration (ET) is a fundamental factor in energy and hydrologic cycles. Although highly precise in-situ ET monitoring is possible, such data are not always available due to the high spatiotemporal variability in ET. This study estimates daily potential ET (PET) in real-time for the Korean Peninsula, via an artificial neural network (ANN), using data from the GEO-KOMPSAT 2A satellite, which is equipped with an Advanced Meteorological Imager (GK2A/AMI). We also used passive microwave data, numerical weather prediction (NWP) model data, and static data. The ANN-based PET model was trained using data for the period 25 July 2019 to 24 July 2020, and was tested by comparing with in-situ PET for the period 25 July 2020 to 31 July 2021. In terms of accuracy, the PET model performed well, with root-mean-square error (RMSE), bias, and Pearson’s correlation coefficient (R) of 0.649 mm day⁻¹, −0.134 mm day⁻¹, and 0.954, respectively. To examine the efficiency of the GK2A/AMI-derived PET data, we compared it with in-situ ET measured at flux towers and with MODIS PET data. The accuracy of the GK2A/AMI-derived PET, in comparison with the flux tower-measured ET, showed RMSE, bias, and Pearson’s R of 1.730 mm day⁻¹, 1.212 mm day⁻¹, and 0.809, respectively. In comparison with the in-situ PET, the ANN model produced more accurate estimates than the MODIS data, indicating that it is more locally optimized for the Korean Peninsula than MODIS. This study advances the field by applying an ANN approach using GK2A/AMI data and could play an important role in examining hydrologic energy for air-land interactions.

Surface urban heat island (SUHI) impacts control the exchange of sensible heat and latent heat between land and atmosphere and can worsen extreme climate events, such as heat waves. This study assessed SUHIs over three megacities (Seoul, Tokyo, Beijing) in East Asia using one-year (April 2011–March 2012) land surface temperature (LST) data retrieved from the Communication, Ocean and Meteorological Satellite (COMS). The spatio-temporal variations of SUHI and the relationship between SUHI and vegetation activity were analyzed using hourly cloud-free LST data. In general, the LST was higher in low latitudes, low altitudes, urban areas and dry regions compared to high latitudes, high altitudes, rural areas and vegetated areas. In particular, the LST over the three megacities was always higher than that in the surrounding rural areas. The SUHI showed a maximum intensity (10–13 °C) at noon during the summer, irrespective of the geographic location of the city, but weak intensities (4–7 °C) were observed during other times and seasons. In general, the SUHI intensity over the three megacities showed strong seasonal (diurnal) variations during the daytime (summer) and weak seasonal (diurnal) variations during the nighttime (other seasons). As a result, the temporal variation pattern of SUHIs was quite different from that of urban heat islands, and the SUHIs showed a distinct maximum at noon of the summer months and weak intensities during the nighttime of all seasons. The patterns of seasonal and diurnal variations of the SUHIs were clearly dependent on the geographic environment of cities. In addition, the intensity of SUHIs showed a strong negative relationship with vegetation activity during the daytime, but no such relationship was observed during the nighttime. This suggests that the SUHI intensity is mainly controlled by differences in evapotranspiration (or the Bowen ratio) between urban and rural areas during the daytime.