This study presents a new research direction for predicting rainfall amount using cloud image data. Herein, we employ a convolutional neural networks (CNNs) to develop an image-value model from cloud image data collected from 20 May 2020 to 24 October 2020 using IoT sensors installed at two research locations in Seoul, Republic of Korea. First, we refine the dataset using data preprocessing in three steps: 1. day/night discrimination, 2. ratio adjustment, and 3. image augmentation. Second, we construct a binary classification model using one-hot encoding for the existence of rainfall. This reduces no-rain data instances and increases model performance, thereby enabling the model to extract image features. Finally, we develop a CNN-based image-value model for rainfall prediction with a well-organized model configuration. Rainfall existence results derived from the binary classification model used for model input as preprocessed cloud image data. Proposed rainfall prediction model exhibited 85.59% accuracy on cloud images with an average mean squared error (MSE) of 3.05 for observation data under 3 mm/h. In particular, single application of the function that divides boolean input by the standard deviation of the dataset within each characteristic resulted in a 17% increase in predicted rainfall accuracy. To the best of our knowledge, this is the first study to train CNN model to predict value (rainfall) with matched image data (cloud), which could be denoted as CNN-based image-value model. Notably, the proposed model can be further extended into other image datasets, including rain streaks with various backgrounds under different climatic conditions.

This study estimated raindrop size distribution (DSD) and rainfall intensity with an infrared surveillance camera in dark conditions. Accordingly, rain streaks were extracted using a k-nearest-neighbor (KNN)-based algorithm. The rainfall intensity was estimated using DSD based on a physical optics analysis. The estimated DSD was verified using a disdrometer for the two rainfall events. The results are summarized as follows. First, a KNN-based algorithm can accurately recognize rain streaks from complex backgrounds captured by the camera. Second, the number concentration of raindrops obtained through closed-circuit television (CCTV) images had values between 100 and 1000 mm⁻¹ m⁻³, and the root mean square error (RMSE) for the number concentration by CCTV and PARticle SIze and VELocity (PARSIVEL) was 72.3 and 131.6 mm⁻¹ m⁻³ in the 0.5 to 1.5 mm section. Third, the maximum raindrop diameter and the number concentration of 1 mm or less produced similar results during the period with a high ratio of diameters of 3 mm or less. Finally, after comparing with the 15 min cumulative PARSIVEL rain rate, the mean absolute percent error (MAPE) was 49 % and 23 %, respectively. In addition, the differences according to rain rate are that the MAPE was 36 % at a rain rate of less than 2 mm h⁻¹ and 80 % at a rate above 2 mm h⁻¹. Also, when the rain rate was greater than 5 mm h⁻¹, MAPE was 33 %. We confirmed the possibility of estimating an image-based DSD and rain rate obtained based on low-cost equipment during dark conditions.

Accurate predictions are crucial for addressing the challenges posed by climate change. Given South Korea’s location within the East Asian summer monsoon domain, characterized by high spatiotemporal variability, enhancing prediction accuracy for regions experiencing heavy rainfall during the summer monsoon is essential. This study aims to derive temporal weighting functions using hybrid surface rainfall radar-observation data as the target, with input from two forecast datasets: the McGill Algorithm for Precipitation Nowcasting by Lagrangian Extrapolation (MAPLE) and the KLAPS Forecast System. The results indicated that the variability in the optimized parameters closely mirrored the variability in the rainfall events, demonstrating a consistent pattern. Comparison with previous blending results, which employed event-type-based weighting functions, showed significant deviation in the average AUC (0.076) and the least deviation (0.029). The optimized temporal weighting function effectively mitigated the limitations associated with varying forecast lead times in individual datasets, with RMSE values of 0.884 for the 1 h lead time of KLFS and 2.295 for the 4–6 h lead time of MAPLE. This blending methodology, incorporating temporal weighting functions, considers the temporal patterns in various forecast datasets, markedly reducing computational cost while addressing the temporal challenges of existing forecast data.

Accurate estimation of snowfall intensity is critical for effective winter weather management, transportation safety, and hydrological forecasting. Traditional approaches predominantly rely on ground-based sensors and radar systems, which are often spatially sparse and costly to install and maintain. In this study, we propose a novel convolutional neural networks (CNNs)-based framework for estimating snowfall intensity using images captured by closed-circuit television (CCTV) cameras, which are gaining attention as prominent IoT sensing devices. This approach capitalizes on the extensive availability of CCTV infrastructure, enabling high-frequency and localized monitoring of snowfall patterns. The proposed model is trained using matched datasets comprising snowfall intensity values obtained from PARSIVEL, a type of disdrometer capable of measuring particle information at ground observation stations, and CCTV data captured simultaneously. The study area, Daegwallyeong in Gangwon Province, South Korea, is highly suitable for snowfall observations, with an average of more than 10 snowy days per month during the winter season from December to February. A notable feature of this framework is its ability to estimate snowfall intensity values from CCTV data by leveraging convolutional neural networks. Furthermore, a dedicated preprocessing step was implemented to extract snowfall particles from the original images, thereby enhancing the accuracy of snowfall intensity estimation. Experimental results demonstrate that the CNNs-based framework developed in this study is highly effective for estimating snowfall intensity using CCTV data. Moreover, the incorporation of snowfall particle extraction during preprocessing significantly improved estimation accuracy compared to scenarios where particle extraction was not applied.AcknowledgementThis work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (RS-2024-00334564), Korea Meteorological Administration Research and Development Program under Grant RS-2023-00243008, and Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (RS-2023-00272105).

This study focuses on the reliability assessment of precipitation data calculated from drop size distribution (DSD) based on disdrometer data observations according to wet-bulb temperature (T_w). Three distinct quality control (QC) methods based on fall velocity were implemented and validated against measurements from tipping-buckets and weighing rain gauges collected from January 2020 to February 2024. The analysis indicated that all QC methods exhibited high reliability (correlation coefficient (CC) > 0.98) for rainfall conditions when T_w was above 5 °C, with a mean absolute percentage error (MAPE) of approximately 8.5 %. However, the precision of precipitation measurements exhibited a notable decline when T_w was below 2 °C, as indicated by a CC of less than 0.6 and MAPE exceeding 30 %. This reduction in accuracy can primarily be attributed to the outcomes of the QC methods, which rely on the falling velocity, given that raindrops and solid particles were observed within the specified T_w range. When considering the melting of snow particles at T_w ranging from 0 °C to 2 °C, the CC approached 0.9, suggesting enhanced measurement reliability. The findings of this study indicate that T_w is a more effective variable than air temperature (T_air) for differentiating the precipitation types. This conclusion arises from the observation that the fall velocity of hydrometeors does not reach the terminal velocity of raindrops, even within the T_air range of 1–5 °C, coupled with the broad distribution of fall velocities. The DSD shape demonstrated stability across multiple QC methods when T_w was equal to or greater than 2 °C. In contrast, considerable variations were observed at lower temperatures, where particles with diameters ranging from 1 to 2 mm exhibited irregular distribution patterns at temperatures below 1 °C. These results suggest that DSD parameters should be derived from disdrometer data obtained under conditions where T_w is above 2 °C to ensure the reliability of the findings. This study provides critical insights for improving precipitation measurement techniques and DSD analyses in regions with variable temperature conditions.