Korea Institute of Atmospheric Prediction Systems

This study aims to improve the accuracy of weather predictions by discovering spatial correlations between Earth observations and atmospheric states. Existing numerical weather prediction (NWP) systems predict future atmospheric states at fixed locations, which are called NWP grid points, by analyzing previous atmospheric states and newly acquired Earth observations. However, the shifting locations of observations and the surrounding meteorological context induce complex, dynamic spatial correlations that are difficult for traditional NWP systems to capture, since they rely on strict statistical and physical formulations. To handle complicated spatial correlations, which change dynamically, we employ a spatiotemporal graph neural networks (STGNNs) with structure learning. However, structure learning has an inherent limitation that this can cause structural information loss and over-smoothing problem by generating excessive edges. To solve this problem, we regulate edge sampling by adaptively determining node degrees and considering the spatial distances between NWP grid points and observations. We validated the effectiveness of the proposed method (CloudNine-v2) using real-world atmospheric state and observation data from East Asia, achieving up to 15% reductions in RMSE over existing STGNN models. Even in areas with high atmospheric variability, CloudNine-v2 consistently outperformed baselines with and without structure learning.

<strong class="journal-contentHeaderColor">Abstract.</strong> Ice surface temperature (IST) is critical for representing surface energy exchange in Arctic forecasts, yet its initialization in operational numerical weather prediction (NWP) systems remains overly simplified&mdash;often inherited from background states or prescribed as spatially uniform values&mdash;due to the scarcity of reliable, spatiotemporally continuous observations. This study examines the forecast impact of realistic IST initialization using the Korean Integrated Model (KIM), a global NWP system that performs well at mid- and low-latitudes but shows limited forecast skill over the Arctic, particularly during winter. In its operational configuration, KIM uniformly initializes a fixed IST value of 271.35 K (&minus;1.8 &deg;C) over all sea ice-covered regions, making it a suitable testbed for this investigation. We generate a physically consistent, gap-free IST dataset using a standalone sea ice model nudged with satellite-retrieved ISTs and initialize it into KIM. Numerical experiments for the 2021&ndash;2022 Arctic winter show that the control run, despite being initialized with unrealistically warm IST, exhibits a slight cold bias in the lower troposphere&mdash;indicating an inherent cooling tendency in KIM that counterbalances artificial heating from the surface boundary. The experimental run, initialized with realistic IST, further amplifies this cold bias. Although random errors are reduced by 3&ndash;5 % in the lower atmosphere, the intensified bias ultimately degrades overall forecast performance. These results demonstrate that while realistic IST initialization influences short-range Arctic forecasts, its benefits are limited without concurrent improvements in underlying model bias. The findings underscore the need for parallel improvements in internal model processes to fully realize its benefits, thereby offering guidance for achieving meaningful gains in Arctic forecast accuracy within KIM.

Abstract. Ice surface temperature (IST) is critical for representing surface energy exchange in Arctic forecasts, yet its initialization in operational numerical weather prediction (NWP) systems remains overly simplified—often inherited from background states or prescribed as spatially uniform values—due to the scarcity of reliable, spatiotemporally continuous observations. This study examines the forecast impact of realistic IST initialization using the Korean Integrated Model (KIM), a global NWP system that performs well at mid- and low-latitudes but shows limited forecast skill over the Arctic, particularly during winter. In its operational configuration, KIM uniformly initializes a fixed IST value of 271.35 K (−1.8 °C) over all sea ice-covered regions, making it a suitable testbed for this investigation. We generate a physically consistent, gap-free IST dataset using a standalone sea ice model nudged with satellite-retrieved ISTs and initialize it into KIM. Numerical experiments for the 2021–2022 Arctic winter show that the control run, despite being initialized with unrealistically warm IST, exhibits a slight cold bias in the lower troposphere—indicating an inherent cooling tendency in KIM that counterbalances artificial heating from the surface boundary. The experimental run, initialized with realistic IST, further amplifies this cold bias. Although random errors are reduced by 3–5 % in the lower atmosphere, the intensified bias ultimately degrades overall forecast performance. These results demonstrate that while realistic IST initialization influences short-range Arctic forecasts, its benefits are limited without concurrent improvements in underlying model bias. The findings underscore the need for parallel improvements in internal model processes to fully realize its benefits, thereby offering guidance for achieving meaningful gains in Arctic forecast accuracy within KIM.

ABSTRACT Soil moisture (SM) has been recognized as one of the essential climate variables in Earth science. Incorporating spatiotemporal information from satellite SM observations into land surface models via land data assimilation (LDA) has emerged as a promising method to improve continuous SM modeling and climate extreme monitoring. Since the reliable satellite SM error dynamics are crucial for successful LDA applications but often assumed to remain static over time, this study presents a novel framework that combines triple collocation analysis (TCA) with advanced machine learning techniques, such as Light Gradient Boosting Machine (LGBM) and Deep Neural Networks (DNN), to accurately quantify spatially and temporally continuous satellite-based SM characteristics on a global scale. In this study, the stand-alone TCA-based time-variant SM error prediction models, which rely exclusively on data used for the Soil Moisture Active Passive (SMAP) retrieval, were developed and comprehensively evaluated. These models successfully recover error information over areas where SMAP SM error data were previously unusable due to uncertain error characteristics. In addition, the stand-alone SMAP-based model demonstrates superior performance compared to model that relies on external datasets, such as the Global Land Data Assimilation System (GLDAS). These findings provide valuable insights into the dynamic nature of satellite-based SM error under various environmental conditions and present a novel way to improve LDA. Furthermore, the proposed methodology can be extended to predict error dynamics for other satellite-based geophysical datasets, broadening its potential applications beyond SMAP.

Understanding the characteristics of the pubertal growth spurt in Korean children and adolescents can serve as crucial foundational data for researching puberty and growth-related disorders. This study aims to estimate the key parameters of pubertal growth, specifically the age and magnitude of the pubertal growth spurt, utilizing longitudinal data from a cohort of Korean children and adolescents. This study used mixed longitudinal height data from a cohort of Korean elementary, middle, and high school students aged 7-18 years. The Superimposition by Translation and Rotation (SITAR) model, a shape-invariant growth curve model, was utilized to estimate a reference height velocity curve for the entire dataset and individual curves via random effects to evaluate pubertal growth parameters. Altogether, 3,339 height measurements (1,519 for boys and 1,820 for girls) from 270 individuals (123 boys and 147 girls) were analyzed. The average age of growth spurt onset in Korean boys was 10.17 ± 0.61 years (mean ± SE), with peak height velocity occurring at 12.46 ± 0.69 years of age (9.61 ± 1.26 cm/year). Korean girls, contrarily, experience their growth spurt at an earlier age (8.57 ± 0.68 years), with peak height velocity occurring at 10.99 ± 0.74 years of age (8.32 ± 1.09 cm/year). An earlier onset of puberty in both sexes is associated with a shorter growth spurt duration (0.63 years for boys and 0.58 years for girls) and a higher peak height velocity (1.82 cm/year for boys and 1.39 cm/year for girls). These associations were statistically significant for both sexes (all p < 0.0001). This study is the first to use the height velocity curve from the SITAR model to examine the pubertal growth spurt of Korean children and adolescents. The estimated timing and magnitude of the pubertal growth spurt, and their relationships can be useful data for clinicians and researchers.

https://doi.org/10.1007/s00382-024-07210-5. In this article 17 references, which were wrongly moved to the Supplementary Information, need to be added. The list of added references as given below: The original article has been corrected

Subseasonal to seasonal (S2S) prediction of droughts and floods is one of the major challenges of weather and climate prediction. Recent studies suggest that the springtime land surface temperature/subsurface temperature (LST/SUBT) over the Tibetan Plateau (TP) can be a new source of S2S predictability. The project “Impact of Initialized Land Surface Temperature and Snowpack on Subseasonal to Seasonal Prediction (LS4P)” was initiated to study the impact of springtime LST/SUBT anomalies over high mountain areas on summertime precipitation predictions. The present work explores the simulated global scale response of the atmospheric circulation to the springtime TP land surface cooling by 16 current state-of-the-art Earth System Models (ESMs) participating in the LS4P Phase I (LS4P-I) experiment. The LS4P-I results show, for the first time, that springtime TP surface anomalies can modulate a persistent quasi-barotropic Tibetan Plateau-Rocky Mountain Circumglobal (TRC) wave train from the TP via the northeast Asia and Bering Strait to the western part of the North America, along with the springtime westerly jet from TP across the whole North Pacific basin. The TRC wave train modulated by the TP thermal anomaly play a critical role on the early summer surface air temperature and precipitation anomalies in the regions along the wave train, especially over the northwest North America and the southern Great Plains. The participant models that fail in capturing the TRC wave train greatly under-predict climate anomalies in reference to observations and the successful models. These results suggest that the TP LST/SUBT anomaly via the TRC wave train is the first order source of the S2S variability in the regions mentioned. Furthermore, the TP surface temperature anomaly can influence the Southern Hemispheric circulation by generating cross-equator wave trains. However, the simulated propagation pathways from the TP into the Southern Hemisphere show large inter-model differences. More dynamical understanding of the TRC wave train as well as its cross-equator propagation into the Southern Hemisphere will be explored in the newly launched LS4P phase II experiment.

The Korean Integrated Model (KIM) is a global numerical weather prediction (NWP) system developed by the first phase project of the Korea Institute of Atmospheric Prediction Systems (KIAPS) and has been used as the operational NWP system at the Korea Meteorological Administration (KMA) since April 2020. The second phase project of KIAPS aims at developing a next-generation NWP system to seamlessly predict from very short-range to extended medium-range. To improve the extended medium-range forecast, one of the main goals of KIAPS is to develop the ensemble prediction system with coupling to land, ocean, and sea ice. The production of extended medium-range reforecast data is necessary to understand the climatological characteristics and model biases of KIM. KIAPS developed an initial version of reforecasting system based on the KIM atmopheric model. The system has a spatial resolution of 50 km (NE090NP3) and consists of 91 vertical layers. We produce reforecast of the cold season cases for 20 years (from 2001 to 2020) and perform the diagnosis and verification of reforecast data. A suite of sensitivity experiments are also performed to investigate the impact of initial perturbations on the ensemble prediction system.

This study explores the ensemble size effect on subseasonal-to-seasonal (S2S) forecasts of the European Center for Medium-Range Weather Forecasts (ECMWF) model. The ensemble forecast skill and its sensitivity to the ensemble size are assessed for the troposphere and stratosphere, and compared with theoretical estimates under the perfect model assumption. The degree of skill improvement in ensemble-mean forecasts with increasing ensemble size agrees well with theoretical estimates in the troposphere. However, in the stratosphere, increasing the ensemble size does not yield as much of the skill improvement as expected. Decomposition of the mean square skill score reveals that the weak ensemble size effect in the stratosphere is primarily caused by a large unconditional bias, which exhibits no apparent decrease with increasing ensemble size. Removing such bias significantly improves the S2S forecast skill and ensemble size effect, suggesting that bias correction is crucial for S2S forecasts, especially in the stratosphere.