WeatherBench is a benchmark dataset for medium-range weather forecasting of geopotential, temperature and precipitation, consisting of preprocessed data, predefined evaluation metrics and a number of baseline models. WeatherBench Probability extends this to probabilistic forecasting by adding a set of established probabilistic verification metrics (continuous ranked probability score, spread-skill ratio and rank histograms) and a state-of-the-art operational baseline using the ECWMF IFS ensemble forecast. In addition, we test three different probabilistic machine learning methods-Monte Carlo dropout, parametric prediction and categorical prediction, in which the probability distribution is discretized. We find that plain Monte Carlo dropout severely underestimates uncertainty. The parametric and categorical models both produce fairly reliable forecasts of similar quality. The parametric models have fewer degrees of freedom while the categorical model is more flexible when it comes to predicting non-Gaussian distributions. None of the models are able to match the skill of the operational IFS model. We hope that this benchmark will enable other researchers to evaluate their probabilistic approaches.Preprint. Under review.
<p>Because the atmosphere is inherently chaotic, probabilistic weather forecasts are crucial to provide reliable information. In this work, we present an extension to the WeatherBench, a benchmark dataset for medium-range, data-driven weather prediction, which was originally designed for deterministic forecasts. We add a set of commonly used probabilistic verification metrics: the spread-skill ratio, the continuous ranked probability score (CRPS) and rank histograms. Further, we compute baseline scores from the operational IFS ensemble forecast.&#160;</p><p>Then, we compare three different methods of creating probabilistic neural network forecasts: first, using Monte-Carlo dropout during inference with a range of dropout rates; second, parametric forecasts, which optimize for the CRPS; and third, categorical forecasts, in which the probability of occurrence for specific bins is predicted. We show that plain Monto-Carlo dropout does not provide enough spread. The parametric and categorical networks, on the other hand, provide reliable forecasts, with the categorical method being more versatile.</p>
Eastward-moving upper air troughs in the subtropical westerlies, commonly known as the western disturbances (WDs) in the Asian subcontinent, are primary sources of precipitation over the northwest Himalayan (NWH) region and the northern plains of India during winter. Many simulation case studies with the help of numerical weather prediction models and a few observational case studies have been conducted to understand the spatial structure, dynamics, energy and weather associated with the WDs over the NWH in the past. However, studies using in situ observations on the impacts of the approaching WDs on various surface meteorological variables at a local scale in the high-altitude mountainous regions of the NWH are lacking. The objectives of this study are to examine the impacts of the approaching WDs on various surface meteorological variables for 11 stations in the NWH and the associated precipitation amount in a 24-hr time interval. Changes (departures) in the values of various meteorological variables are examined on the first precipitation day of the occurrence of active WDs to study the impacts of the approaching WDs and the associated precipitation amount in the 24-hr time interval which are found to depend on the altitude and geographic location of a station. The mean drop in the maximum and ambient air temperatures are found to be 2.0 • and 0.7 • C, respectively, while the mean rise in the minimum air temperature was found to be 0.8 • C. A mean drop in the surface atmospheric pressure and a mean increase in the relative humidity are found to be 0.9 hPa and 19.5%, respectively, in a 24-hr time interval. The mean precipitation amount and mean maximum precipitation amount associated with the active WDs in the 24-hr time interval are found to be 8.9 and 68.8 mm, respectively. The results are briefly discussed in the paper. The findings of this study can be useful for operational weather forecasting and a selection of precursor variables for developing a real-time local scale weather forecast model(s) for remote areas of the NWH for the winter season.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.