Meteorological Applications最新文献

Weather regimes and weather patterns, here jointly referred to as circulation types, are used to generate forecasts for a variety of applications, such as energy demand and flood risk. However, there are usually many different choices available for precisely which circulation types to use. Ideally, one would like to use circulation types that are both highly informative for the application and also highly predictable, but in practice, there is often a tradeoff between informativity and predictability. We present a simple, general framework for how to construct a circulation type forecast that optimally balances these factors by segueing between different choices of circulation types at different lead times based on information-theoretic considerations. As an example, we apply this framework to the case of forecasting energy demand in Great British winters. We compare a set of 30 weather patterns produced by the UK Met Office with the much simpler two-state framework consisting of a positive and negative North Atlantic Oscillation (NAO) regime and show how to optimally combine the two across a winter season.

Understanding human thermal comfort is essential for assessing environmental conditions and their implications for well-being, particularly in the context of global climate change. This study examines the influence of 30 climatic and ecological factors, including temperature, humidity, atmospheric pressure, solar radiation, wind dynamics, and topographical characteristics, on human thermal comfort across Iran. A multidisciplinary approach was employed, integrating principal component analysis (PCA) for feature selection, multivariate regression (MR) for impact quantification, cluster analysis (CA) for climate classification, and spatial modeling (SMA) to assess regional disparities. Furthermore, machine learning models (MLMs) and artificial neural networks (ANNs) were utilized to capture complex, nonlinear relationships in climate–comfort interactions. Based on a comprehensive data set spanning 38 years (1984–2022), the findings reveal significant spatial variations in climate sensitivity. Weighted indices such as predicted mean vote (PMV), physiologically equivalent temperature (PET), and thermal discomfort index (TDI) enhance the precision of comfort assessments. The results indicate that northern Iran, particularly the western coastal region of the Caspian Sea, exhibits the most favorable climatic conditions, whereas arid and semi-arid areas experience heightened thermal stress. These insights advance biometeorological research by linking climate variability to human physiological responses and provide practical implications for urban planning, public health policies, and climate adaptation strategies. By integrating high-dimensional climate data with advanced computational techniques, this study highlights the necessity of adaptive measures to mitigate the impacts of climate change on human thermal comfort.

An optimal transport (OT) problem seeks to find the cheapest mapping between two distributions with equal total density, given the cost of transporting density from one place to another. Unbalanced OT allows for different total density in each distribution. This is the typical setting for precipitation forecast and observation data, when considering the densities as accumulated rainfall, or intensity. True OT problems are computationally expensive, however through entropic regularisation it is possible to obtain an approximation maintaining many of the underlying attributes of the true problem. In this work, entropic unbalanced OT and its associated Sinkhorn divergence are examined as a spatial forecast verification method for precipitation data. The latter being a novel introduction to the forecast verification literature. It offers many attractive features, such as morphing one field into another, defining a distance between fields and providing feature based optimal assignment. This method joins the growing research by the Spatial Forecast Verification Methods Inter-Comparison Project (ICP) which aims to unite spatial verification approaches. After testing this methodology's behaviour on numerous ICP test sets, it is found that the Sinkhorn divergence is robust against the common double penalty problem (a form of phase error), on average aligns with expert assessments of model performance, and allows for a variety of novel pictorial illustrations of error. It provides informative summary scores, and has few limitations to its application. Combined, these findings place unbalanced entropy regularised optimal transport and the Sinkhorn divergence as an informative method which follows geometric intuition.

Statistical methods can be used to create bias correction models that learn from past forecast errors and reduce systematic errors in real-time forecasts. This study presents a machine learning (ML) approach using extreme gradient-boosted (XGBoost) trees to address biases in a numerical weather prediction (NWP) nowcast model for key meteorological parameters: 2-m temperature, 2-m relative humidity, 10-m wind speed, and 10-m wind gust. These ML models have been integrated into the Finnish Meteorological Institute's (FMI) operational nowcasting framework, Smartmet nowcast. Results show that, even with a relatively modest set of meteorological predictors, the ML bias correction method significantly improves forecast accuracy, reducing the root mean square error (RMSE) by 24%–29% compared to the direct NWP model output. The implementation of this new bias correction method not only improves the quality of FMI's short-range forecasts, but also extends the availability of bias-corrected data for longer forecast lead times, offering substantial improvements over the previously implemented bias correction method. The codebase for this machine learning bias correction is available at (https://github.com/fmidev/snwc_bc).

An increasing body of evidence indicates that publics want more probabilistic information included in their weather forecasts. However, more guidance on incorporating probability information into weather risk communication is needed. The National Hurricane Center (NHC) recently developed prototype forecast graphics that include probabilistic values of intensity at landfall when landfall is possible. The goal of this research was to develop those prototypes into a forecast product that expresses technical uncertainty in an intensity forecast in a manner that is understandable and effective to various publics. In Study 1, an online survey among Florida residents was conducted. Quantitative analysis of the survey data showed few significant differences between the prototypes and the currently operational forecast track graphic, commonly referred to as the cone of uncertainty (COU). Analysis of the responses to open-ended questions in the survey and feedback from focus group participants consisting of NHC partners working in hurricane-prone areas guided revisions to improve the prototypes. In Study 2, the modified prototypes produced an improvement in understanding of certain aspects of the intensity forecast. Promisingly, most people surveyed preferred the additional probabilistic information in the prototypes to the status quo COU message. In fact, nearly 90% of respondents indicated that they preferred at least some percentage values in their weather forecasts as opposed to forecasts with words only. This suggests that further development of a probabilistic landfall intensity product might be warranted.

The availability of precipitation data from in situ stations faces various challenges including quality, temporal resolution, irregular spatial distribution, and scarcity in many regions. This is particularly true for the West Bank. Hence, the need to identify alternatives sources is a priority as high quality precipitation estimates are essential for accurate hydrological applications. This study assesses the reliability of four satellite precipitation products (IMERG Final Run, PDIR-Now, CCS-CDR, CMORPH) against 442 in situ rainfall stations across Israel (354) and Palestine (88). These four satellite products, with spatial resolutions ranging from 4 to 10 km, were evaluated at the daily timescale to maximize the number of in situ stations available. The analysis reveals that IMERG outperforms the other products, with a mean $��R^2�$ of 0.33 and a Probability of Detection (POD) of 0.7, without any adjustments. The study also examined the influence of elevation on satellite performance, noting that while IMERG consistently excels in most indices, PDIR has lower Mean Absolute Errors at lower elevations. The results highlight a disparity in performance between the Israeli and Palestinian in situ stations. Overall, IMERG emerges as the most reliable satellite-based estimate for the Levant region, proving effective across different elevations, climatic zones, and rainfall intensities.

Windstorms are a significant natural hazard in Europe and Norway, and while many national meteorological agencies issue warnings for severe storm events, studies estimating their impacts are rare. It has been hypothesized that forecasting storm damages could help stakeholders make better informed decisions in the event of a storm. Using 41 years of daily municipality-level historical Norwegian insurance loss data and high resolution wind speed data from the Norwegian hindcast (NORA3), we propose a novel conceptual framework for probabilistic storm damage forecasting and we test it on the Norwegian Meteorological Institute's MetCoOp Ensemble Prediction System (MEPS). The damage forecasting is performed in two steps: first, a color-coded warning system that issues warnings based on the municipality-level probabilities of the event being a medium, high, or extreme loss event, and second, forecasting damages in monetary terms using damage functions. The color-coded warning system is implemented at the municipality level and the gridded wind speeds are weighted with population density to account for local exposure. The monetary damages are estimated on a county level using four different damage functions. The damage-informed color-coded warning system shows promising results in comparison with a more traditional wind-informed return period-based warning system, demonstrating the ability to forecast the spatial patterns of losses across different loss categories. The county-specific recorded damages lie within the range of the ensemble of damage forecasts 70% of the time for storms not used in the fitting of the damage functions. However, the proposed color-coded warning for damage forecasting is not free from false alarms but is suited to act as a decision help for skilled users.

The Philippines is frequently affected by tropical cyclones (TCs). Among the TC-associated hazards, rainfall can lead to cascading impacts such as floods and landslides. A robust and computationally inexpensive TC rainfall forecasting method is critical in disaster preparation and risk reduction efforts. We used machine learning (ML) to develop a TC rainfall forecast model from parameters such as TC track and locale-specific characteristics. Specifically, a self-organizing map (SOM) was utilized to cluster the TC tracks, which were then fed into a random forest (RF) regression model that used TC position, intensity, translational speed, and other parameters to predict accumulated TC rainfall. The resulting artificial intelligence (AI)-based TC rainfall model was initially assessed against ground rainfall observations for calibration. Then, the model was evaluated for its prediction skill. Model interpretability of the RF model revealed insights into how the input parameters influence the model response. The RF model determined that distance to TC has the most influence on the variability of the accumulated TC rainfall, followed by TC duration, latitude of land grid, and the type of TC track as clustered by the SOM. The model produced similar rainfall distributions to calibrated satellite rainfall observations. It was able to produce rain predictions well and is particularly skillful in predicting intense rainfall events in comparison with the other statistical or dynamical weather models (i.e., WRF model). The predictive ability of the RF model, together with its low computational power requirement, makes it a potential tool to augment TC rainfall forecasting in the Philippines.

Increasing the proportion of energy generation from renewables is a necessary step towards reducing greenhouse gas emissions. However, renewable energy sources such as wind and solar are highly weather sensitive, leading to a challenge when balancing energy demand and renewable energy production. Identifying periods of high shortfall, here defined as when electricity demand substantially exceeds renewable production, and understanding how these periods are affected by weather is therefore critical. We use a previously constructed energy dataset derived from reanalysis data for a fixed electricity system to analyse the link between weather regimes and periods of high shortfall during the winter for 28 European countries. Building on previous work and following similar studies, we provide both a subcontinental and country-specific perspective. For each country, we identify days with critical energy conditions, specifically high-energy demand, low wind and solar generation, and high-energy shortfall. We show that high shortfall is more driven by demand than by production in countries with colder climates or less installed wind capacity, and is more driven by production than by demand in countries with warmer climates or more installed wind capacity. Of the six weather regimes considered here, only a subset is found to favour the occurrence of high shortfall days. This subset affects much of Europe, causing simultaneous shortfall days across multiple countries. Furthermore, if multiple countries experience shortfall days, neighbouring countries are more likely to experience shortfall days. Motivated by this result, we examine the hypothetical impact the coldest European winter of the 20th century, 1962/1963, would have had on the present-day energy system. We found that persistent blocking conditions associated with that winter, if they occurred today, would lead to higher demand and shortfall across Europe during most of the winter and would be extreme in this respect compared to other winters.

Ventilation of cities by local cold-air flows is an important measure in urban heat island mitigation and climate-resilient urban planning. We introduce a cold-air connectivity analysis to identify relevant cold-air formation areas as well as urban quarters ventilated by cold-air flows. The nocturnal cold-air flow trajectories are calculated from numerical model simulations using the single-layer cold-air drainage model KLAM_21 and the newly developed trajectory calculator KLATra. The German city of Freiburg im Breisgau is chosen to demonstrate the cold-air connectivity analysis based on trajectories calculated for two 3-hourly periods during an idealised night. Hydrological catchment boundaries and land use define eight rural cold-air formation areas as starting points for forward trajectories, whereas administrative urban district boundaries and land use data are used to define five built-up quarters potentially prone to overheating as starting points for cold-air backward trajectories. A rate of connectivity is calculated from the ratio of trajectories connecting cold-air formation areas with overheated urban quarters to the total number of trajectories. The analysis reveals the potential of cold-air formation areas to ventilate single or multiple urban quarters at connectivity rates up to 82%. The connectivity analysis therefore supports identification and assessment of the relevance of specific cold-air formation areas for urban heat island mitigation and may serve as a valuable planning tool and data basis for objective decision making.