Angular observations, or observations lying on the unit circle, arise in many disciplines and require special care in their description, analysis, interpretation and visualization. We provide methods to construct concentric circular boxplot displays of distributions of groups of angular data. The use of concentric boxplots brings challenges of visual perception, so we set the boxwidths to be inversely proportional to the square root of their distance from the centre. A perception survey supports this scaled boxwidth choice. For a large number of groups, we propose circular quartile plots. A three-dimensional toroidal display is also implemented for periodic angular distributions. We illustrate our methods on datasets in (1) psychology, to display motor resonance under different conditions, (2) genomics, to understand the distribution of peak phases for ancillary clock genes, and (3) meteorology and wind turbine power generation, to study the changing and periodic distribution of wind direction over the course of a year.
Following the implementation of the Shanghai Clean Air Act, this study investigates the evolution of air pollution in central Shanghai (Putuo District) by analyzing continuous monitoring data (2016–2020) and chemical speciation of particulate matter (2017–2018). The results confirm a transition toward a “low exceedance rate and low background concentration” regime. However, short-term exceedance episodes persist, generally occurring in winter and spring, with significantly amplified diurnal variations on exceedance days. Distinct patterns emerged between PM fractions: PM<sub>10</sub> exceedances were characterized by a single morning peak linked to traffic-induced coarse particles, while PM<sub>2.5</sub> exceedances showed synchronized diurnal peaks with NO<sub>2</sub>, suggesting a stronger contribution from vehicle exhaust. Source apportionment revealed that mineral components (21.61%) and organic matter (OM, 21.02%) dominated in PM<sub>10</sub>, implicating construction and road dust. In contrast, PM<sub>2.5</sub> was primarily composed of OM (26.73%) and secondary inorganic ions (dominated by nitrate), highlighting the greater importance of secondary formation. The findings underscore that sustained PM<sub>2.5</sub> mitigation requires targeted control of gasoline vehicle emissions and gaseous precursors.
Modern weather forecasting has increasingly transitioned from numerical weather prediction (NWP) to data-driven machine learning forecasting techniques. While these new models produce probabilistic forecasts to quantify uncertainty, their training and evaluation may remain hindered by conventional scoring rules, primarily MSE, which ignore the highly correlated data structures present in weather and atmospheric systems. This work introduces the signature kernel scoring rule, grounded in rough path theory, which reframes weather variables as continuous paths to encode temporal and spatial dependencies through iterated integrals. Validated as strictly proper through the use of path augmentations to guarantee uniqueness, the signature kernel provides a theoretically robust metric for forecast verification and model training. Empirical evaluations through weather scorecards on WeatherBench 2 models demonstrate the signature kernel scoring rule's high discriminative power and unique capacity to capture path-dependent interactions. Following previous demonstration of successful adversarial-free probabilistic training, we train sliding window generative neural networks using a predictive-sequential scoring rule on ERA5 reanalysis weather data. Using a lightweight model, we demonstrate that signature kernel based training outperforms climatology for forecast paths of up to fifteen timesteps.
Laura Thapa, Marybeth Arcodia, Elizabeth A. Barnes
We discuss the utility of applying clustering as a preprocessing step for identifying subseasonal to seasonal forecasts of opportunity of coastal sea level using convolutional neural networks (CNNs). Clustering leverages potential covariance among points along the same coastline or in the same ocean basin. To evaluate the utility of clustering for reliably identifying forecasts of opportunity, we compare CNNs trained to predict sea level probability distributions in three ways: over the whole Northeast Pacific Coast simultaneously, over predetermined clusters within this coastline, and at individual gridpoints near tide gauges. All CNN prediction tasks (Whole Coast, Cluster, Point), outperform climatology by a similar margin at Week 3 when the entire test set is used to evaluate CNN skill. However, when comparing the skill of each tasks' 20% most confident predictions, we find the skill of the Cluster and Point tasks to be on par with each other and substantially more skillful than the Whole Coast task. Of the Cluster and Point task, the Cluster task represents all gridpoints in the Northeast Pacific Coast with minimal tunable parameters. Throughout this exercise we learned that clustering gridpoints as a pre-processing step is the preferred approach between the three for making S2S predictions of coastal sea level.
Vedant Acharya, Abhay Pisharodi, Rishabh Mondal
et al.
Air pollution causes about 1.6 million premature deaths each year in India, yet decision makers struggle to turn dispersed data into decisions. Existing tools require expertise and provide static dashboards, leaving key policy questions unresolved. We present VayuChat, a conversational system that answers natural language questions on air quality, meteorology, and policy programs, and responds with both executable Python code and interactive visualizations. VayuChat integrates data from Central Pollution Control Board (CPCB) monitoring stations, state-level demographics, and National Clean Air Programme (NCAP) funding records into a unified interface powered by large language models. Our live demonstration will show how users can perform complex environmental analytics through simple conversations, making data science accessible to policymakers, researchers, and citizens. The platform is publicly deployed at https://huggingface.co/spaces/SustainabilityLabIITGN/ VayuChat. For further information check out video uploaded on https://www.youtube.com/watch?v=d6rklL05cs4.
David A. Lavers, Gabriele Villarini, Hannah L. Cloke
et al.
Abstract A typical question posed following an extreme precipitation event is: How does this compare to past events? This question is being asked more frequently and is of importance to climate monitoring services, such as the Copernicus Climate Change Service (C3S). Currently, the statistics extensively used for this purpose are not generally understandable to the wider public, or they are not tailored towards presenting extremes. To mitigate this situation, this article uses a modified version of the Extreme Rain Multiplier (ERM), which was developed for tropical cyclones, and applies it to precipitation events globally. For daily precipitation considered herein, the ERM is calculated by dividing the daily precipitation accumulation during an event by the mean historical annual maxima of daily precipitation (RX1day), which is computed over 1991–2020. Using the European Centre for Medium‐Range Weather Forecasts ERA5 reanalysis, the calculation of the ERM is illustrated for six extreme events around the world; these included convective systems, atmospheric rivers and tropical cyclones. A maximum ERM of 4 was found during Storm Daniel, in Greece, and in Tropical Cyclone Jasper in Australia, implying that four times the mean RX1day precipitation occurred. The ERM will be useful in C3S reporting activities because it can objectively identify extreme precipitation events. Furthermore, after extracting the number of precipitation events per year at each grid point that had an ERM exceeding 1, a trend analysis was undertaken to ascertain if the frequency of extreme events had changed with time. Results showed that the most widespread increasing trends in the ERM were in the tropics, but these trends are thought to be questionable in ERA5. There were few clear trends in other regions. In conclusion, the ERM can communicate the level of extreme precipitation in a clear manner and can be used in climate monitoring activities.
Abstract The urban heat island effect is influenced by radiation from sidewalks and streets, which alters the apparent air temperature near the surface. Therefore, urban dwellers who are close to the ground (children, pets, etc.) should have higher heat exposure, increasing vulnerability. However, it is not well known how heat health risk varies in the near-surface atmosphere given a high surface radiative temperature. To investigate this problem, wet-bulb globe temperature (WBGT), air temperature, humidity, and wind speed were measured with a Kestrel 5400 at two levels, 0.5 and 1.5 m, in two nearby locations over 4 summer days and within 2 h (1300–1400, 1500–1600) and related to forward-looking infrared (FLIR) images of the underlying sidewalks in a hot neighborhood in Charleston, South Carolina. WBGT was consistently higher at 0.5 m than at 1.5 m, and this difference was larger than differences based on location or time of the day. Cumulative distribution functions between 0.5- and 1.5-m WBGT showed the largest differences at values well above the highest defined heat stress conditions of “black flag.” Air and dewpoint temperature differences between these heights were not significantly related to differences in WBGT, but wind speeds were. Infrared surface temperature appears to have little contemporaneous relationship with air temperature at 0.5 and 1.5 m. However, WBGT at both heights was significantly positively related to the maximum (and average) infrared temperature in the sidewalk images. The potential health impacts on vulnerable children and pets should motivate mitigation measures to reduce radiation coming off urban surfaces. Significance Statement The purpose of this study is to first determine whether there is a difference in experienced heat stress between a height of 0.5 and 1.5 m above urban sidewalks and second whether surface infrared temperatures are related to any differences noted. At two locations in the afternoon hours over 4 days, the 0.5-m heat stress was always greater than the 1.5-m heat stress, and these differences were most pronounced when the health danger was extreme. As the sidewalk became warmer, the heat stress at 0.5 and 1.5 m both increased equally. Our results have particular significance for children, pets, and anyone that spends time close to the ground in an urban environment.
Satellite images have become increasingly valuable for modelling regional climate change effects. Earth surface forecasting represents one such task that integrates satellite images with meteorological data to capture the joint evolution of regional climate change effects. However, understanding the complex relationship between specific meteorological variables and land surface evolution poses a significant challenge. In light of this challenge, our paper introduces a pipeline that integrates principles from both perturbation-based explainability techniques like LIME and global marginal explainability techniques like PDP, besides addressing the constraints of using such techniques when applying them to high-dimensional spatiotemporal deep models. The proposed pipeline simplifies the undertaking of diverse investigative analyses, such as marginal sensitivity analysis, marginal correlation analysis, lag analysis, etc., on complex land surface forecasting models In this study we utilised Convolutional Long Short-Term Memory (ConvLSTM) as the surface forecasting model and did analyses on the Normalized Difference Vegetation Index (NDVI) of the surface forecasts, since meteorological variables like temperature, pressure, and precipitation significantly influence it. The study area encompasses various regions in Europe. Our analyses show that precipitation exhibits the highest sensitivity in the study area, followed by temperature and pressure. Pressure has little to no direct effect on NDVI. Additionally, interesting nonlinear correlations between meteorological variables and NDVI have been uncovered.
Filip Sabo, Martin Claverie, Michele Meroni
et al.
This paper investigated the potential of a multivariate Transformer model to forecast the temporal trajectory of the Fraction of Absorbed Photosynthetically Active Radiation (FAPAR) for short (1 month) and long horizon (more than 1 month) periods at the regional level in Europe and North Africa. The input data covers the period from 2002 to 2022 and includes remote sensing and weather data for modelling FAPAR predictions. The model was evaluated using a leave one year out cross-validation and compared with the climatological benchmark. Results show that the transformer model outperforms the benchmark model for one month forecasting horizon, after which the climatological benchmark is better. The RMSE values of the transformer model ranged from 0.02 to 0.04 FAPAR units for the first 2 months of predictions. Overall, the tested Transformer model is a valid method for FAPAR forecasting, especially when combined with weather data and used for short-term predictions.
Heavy precipitation from tropical cyclones (TCs) may result in disasters, such as floods and landslides, leading to substantial economic damage and loss of life. Prediction of TC precipitation based on ensemble post-processing procedures using machine learning (ML) approaches has received considerable attention for its flexibility in modeling and its computational power in managing complex models. However, when applying ML techniques to TC precipitation for a specific area, the available observation data are typically insufficient for comprehensive training, validation, and testing of the ML model, primarily due to the rapid movement of TCs. We propose to use the convolutional neural network (CNN) as a deep ML model to leverage the spatial information of precipitation. The proposed model has three distinct features that differentiate it from traditional CNNs applied in meteorology. First, it utilizes data augmentation to alleviate challenges posed by the small sample size. Second, it contains geographical and dynamic variables to account for area-specific features and the relative distance between the study area and the moving TC. Third, it applies unequal weights to accommodate the temporal structure in the training data when calculating the objective function. The proposed CNN-all model is then illustrated with the TC Soudelor's impact on Taiwan. Soudelor was the strongest TC of the 2015 Pacific typhoon season. The results show that the inclusion of augmented data and dynamic variables improves the prediction of heavy precipitation. The proposed CNN-all outperforms traditional CNN models, based on the continuous probability skill score (CRPSS), probability plots, and reliability diagram. The proposed model has the potential to be utilized in a wide range of meteorological studies.
Predicting agricultural yields is imperative for effective planning to sustain the growing global population. Traditionally, regression-based, simulation-based, and hybrid methods were employed for yield prediction. In recent times, there has been a notable shift towards the adoption of Machine Learning (ML) methods, with Deep Learning (DL), particularly Convolutional Neural Networks (CNNs) and Long-Short Term Memory (LSTM) networks, emerging as popular choices for their enhanced predictive accuracy. This research introduces a cost-effective DL architecture tailored for corn yield prediction, considering computational efficiency in processing time, data size, and NN architecture complexity. The proposed architecture, named SEDLA (Simple and Efficient Deep Learning Architecture), leverages the spatial and temporal learning capabilities of CNNs and LSTMs, respectively, with a unique emphasis on exploring the impact of kernel size in CNNs. Simultaneously, the study aims to exclusively employ satellite and yield data, strategically minimizing input variables to enhance the model’s simplicity and efficiency. Notably, the study demonstrates that employing larger kernel sizes in CNNs, especially when processing histogram-based Surface Reflectance (SR) and Land Surface Temperature (LST) data from Moderate Resolution Imaging Spectroradiometer (MODIS), allows for a reduction in the number of hidden layers. The efficacy of the architecture was evaluated through extensive testing on corn yield prediction across 13 states in the United States (U.S.) Corn Belt at county-level. The experimental results showcase the superiority of the proposed architecture, achieving a Mean Absolute Percentage Error (MAPE) of 6.71 and Root Mean Square Error (RMSE) of 14.34, utilizing a single-layer CNN with a 15 × 15 kernel in conjunction with LSTM. These outcomes surpass existing benchmarks in the literature, affirming the efficacy and potential of the suggested DL framework for accurate and efficient crop yield predictions.
Jiangxi Province, characterized by abundant forest resources and complex topography, is highly susceptible to forest fires. This study integrated multiple factors, including topography, climate, vegetation, and human activities, and employed machine learning models, specifically random forest (RF), support vector machine (SVM), and back-propagation neural network (BPNN), to predict forest fire occurrence in Jiangxi. Using Moderate Resolution Imaging Spectroradiometer L3 fire-point data from 2001–2020, we analyzed the spatiotemporal distribution of forest fires and applied the weight of evidence (WoE) method to evaluate the correlation between forest fires and environmental factors. WoE was employed to select negative samples, which were compared with those obtained using traditional random sampling methods. The optimal model was then utilized to generate seasonal spatial distribution maps of forest fire risk throughout Jiangxi Province. The results showed that over the past two decades, the frequency of forest fires generally decreased. RF demonstrated a significant advantage over SVM and BPNN in predicting forest fires. Vegetation coverage was the most influential factor. In addition, the models trained with WoE-selected negative samples exhibited enhanced accuracy, with area under the curve values increasing from 0.946 to 0.995 for RF, 0.8344 to 0.925 for SVM, and 0.832 to 0.850 for BPNN, compared to those trained with randomly sampled negative data. Finally, forest fires were most frequent during winter, particularly in Ganzhou, Fuzhou, and Ji'an. High-risk fire zones were more dispersed in spring, whereas autumn fires were primarily concentrated in Ganzhou, and fire activity was relatively low during summer. The seasonal forest fire risk maps generated in this study offer valuable insights for guiding forest fire management in the Jiangxi Province and similar regions, providing critical practical significance for informed decision-making.
Simon James Walker, Karl Magnus Laundal, Jone Peter Reistad
et al.
Abstract The boundaries of the auroral oval and auroral electrojets are an important source of information for understanding the coupling between the solar wind and the near‐earth plasma environment. Of these two types of boundaries the auroral electrojet boundaries have received comparatively little attention, and even less attention has been given to the connection between the two. Here we introduce a technique for estimating the electrojet boundaries, and other properties such as total current and peak current, from 1‐D latitudinal profiles of the eastward component of equivalent current sheet density. We apply this technique to a preexisting database of such currents along the 105° magnetic meridian, estimated using ground‐based magnetometers, producing a total of 11 years of 1‐min resolution electrojet boundaries during the period 2000–2020. Using statistics and conjunction events we compare our electrojet boundary data set with an existing electrojet boundary data set, based on Swarm satellite measurements, and auroral oval proxies based on particle precipitation and field‐aligned currents. This allows us to validate our data set and investigate the feasibility of an auroral oval proxy based on electrojet boundaries. Through this investigation we find the proton precipitation auroral oval is a closer match with the electrojet boundaries. However, the bimodal nature of the electrojet boundaries as we approach the noon and midnight discontinuities makes an average electrojet oval poorly defined. With this and the direct comparisons differing from the statistics, defining the proton auroral oval from electrojet boundaries across all local and universal times is challenging.
Nanthapong Chantaraprachoom, Hikari Shimadera, Katsushige Uranishi
et al.
This study utilized the Community Multiscale Air Quality (CMAQ) model to assess the impact of open biomass burning (OBB) in Thailand and neighboring countries—Myanmar, Laos, Cambodia, and Vietnam—on the PM<sub>2.5</sub> concentrations in the Bangkok Metropolitan Region (BMR) and Upper Northern Region of Thailand. The Upper Northern Region was further divided into the west, central, and east sub-regions (WUN, CUN, and EUN) based on geographical borders. The CMAQ model was used to simulate the spatiotemporal variations in PM<sub>2.5</sub> over a wide domain in Asia in 2019. The Integrated Source Apportionment Method (ISAM) was utilized to quantify the contributions from OBB from each country. The results showed that OBB had a minor impact on PM<sub>2.5</sub> in the BMR, but transboundary transport from Myanmar contributed to an increase in PM<sub>2.5</sub> levels during the peak burning period from March to April. In contrast, OBB substantially impacted PM<sub>2.5</sub> in the Upper Northern Region, with Myanmar being the major contributor in WUN and CUN and domestic burning being the major contributor to EUN during the peak months. Despite Laos having the highest OBB emissions, meteorological conditions caused the spread of PM<sub>2.5</sub> eastward rather than into Thailand. These findings highlight the critical impact of regional transboundary transport and emphasize the necessity for collaborative strategies for mitigating PM<sub>2.5</sub> pollution across Southeast Asia.
Uncertainties in projected 21st century warming were very large a decade ago, increasing the costs of climate change adaptation, especially those associated with long-lived infrastructure. Here we show that through progress in climate policy and climate science, these uncertainties have decreased dramatically over the past decade.
In this work, long-term spatiotemporal changes in rainfall are analysed and evaluated using whole-year data from Rajasthan, India, at the meteorological divisional level. In order to determine how the rainfall pattern has changed over the past 10 years, I examined the data from each of the thirteen tehsils in the Jaipur district. For the years 2012 through 2021, daily rainfall information is available from the Indian Meteorological Department (IMD) in Jaipur. We primarily compare data broken down by tehsil in the Jaipur district of Rajasthan, India.
Kai Jeggle, David Neubauer, Gustau Camps-Valls
et al.
Cirrus clouds are key modulators of Earth's climate. Their dependencies on meteorological and aerosol conditions are among the largest uncertainties in global climate models. This work uses three years of satellite and reanalysis data to study the link between cirrus drivers and cloud properties. We use a gradient-boosted machine learning model and a Long Short-Term Memory (LSTM) network with an attention layer to predict the ice water content and ice crystal number concentration. The models show that meteorological and aerosol conditions can predict cirrus properties with $R^2 = 0.49$. Feature attributions are calculated with SHapley Additive exPlanations (SHAP) to quantify the link between meteorological and aerosol conditions and cirrus properties. For instance, the minimum concentration of supermicron-sized dust particles required to cause a decrease in ice crystal number concentration predictions is $2 \times 10^{-4}$ mg m\textsuperscript{-3}. The last 15 hours before the observation predict all cirrus properties.
A better understanding of the decision context within which climate services are used is likely to be central to designing user-relevant climate services for adaptation action. As climate change presents a risk, one entry point to better understand the decision context is through an exploration of the perceptions of climate change risk. How risky climate change is perceived to be will influence whether action is taken on climate change, what decisions are made and the types of information that are used when taking action, providing valuable insights into the decision-context. This study quantifies and explores climate change risk perceptions, and its determinants, amongst policy decision influencers in east Africa. Climate change risk perceptions are found to be heightened, driven by observance of social norms, perceptions of climate change as a proximal risk, frequent experience of extreme weather events and a predominantly self-transcending (outward looking) value system among policy decision influencers. By drawing on known principles from environmental psychology, the study’s results lead to a set of suggestions about how currently available climate services could be aligned to the east African decision context to better encourage uptake and action.
Meteorology. Climatology, Social sciences (General)