The atmosphere is chaotic. This fundamental property of the climate system makes forecasting weather incredibly challenging: it's impossible to expect weather models to ever provide perfect predictions of the Earth system beyond timescales of approximately 2 weeks. Instead, atmospheric scientists look for specific states of the climate system that lead to more predictable behaviour than others. Here, we demonstrate how neural networks can be used, not only to leverage these states to make skillful predictions, but moreover to identify the climatic conditions that lead to enhanced predictability. Furthermore, we employ a neural network interpretability method called ``layer-wise relevance propagation'' to create heatmaps of the regions in the input most relevant for a network's output. For Earth scientists, these relevant regions for the neural network's prediction are by far the most important product of our study: they provide scientific insight into the physical mechanisms that lead to enhanced weather predictability. While we demonstrate our approach for the atmospheric science domain, this methodology is applicable to a large range of geoscientific problems.
Huracan: A skillful end-to-end data-driven system for ensemble data assimilation and weather prediction
Microsoft Corporation（微软公司）；Department of Earth System Science, Ministry of Education Key Laboratory for Earth System Modeling, Institute for Global Change Studies（教育部地球系统科学部门地球系统建模重点实验室，全球变化研究机构）；Tsinghua University（清华大学）
Historical observations of severe weather and simulated severe weather environments (i.e., features) from the Global Ensemble Forecast System v12 (GEFSv12) Reforecast Dataset (GEFS/R) are used in conjunction to train and test random forest (RF) machine learning (ML) models to probabilistically forecast severe weather out to days 4--8. RFs are trained with 9 years of the GEFS/R and severe weather reports to establish statistical relationships. Feature engineering is briefly explored to examine alternative methods for gathering features around observed events, including simplifying features usin…
FengWu: Pushing the Skillful Global Medium-range Weather Forecast beyond 10 Days Lead
Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）；University of Science and Technology of China（中国科学技术大学）；Shanghai Jiao Tong University（上海交通大学）；Nanjing University of Information Science and Technology（南京信息工程大学）；The Institute of Atmospheric Physics, Chinese Academy of Sciences（中国科学院大气物理研究所）；Shanghai Meteorological Bureau（上海市气象局）
A Data-Driven Regional Model for Skillful Medium-Range Typhoon Prediction
Shanghai Typhoon Institute, and Key Laboratory of Numerical Modeling for Tropical Cyclone of the China Meteorological Administration（上海台风研究所，中国气象局台风数值模式重点实验室）；Department of Atmospheric and Oceanic Sciences and Institute of Atmospheric Sciences, Fudan University（复旦大学大气与海洋科学系和大气科学研究院）；Department of Atmospheric Sciences, University of Illinois at Urbana–Champaign（伊利诺伊大学厄巴纳-香槟分校大气科学系）
The problem of forecasting weather has been scientifically studied for centuries due to its high impact on human lives, transportation, food production and energy management, among others. Current operational forecasting models are based on physics and use supercomputers to simulate the atmosphere to make forecasts hours and days in advance. Better physics-based forecasts require improvements in the models themselves, which can be a substantial scientific challenge, as well as improvements in the underlying resolution, which can be computationally prohibitive. An emerging class of weather mode…
We present a data-driven approach for forecasting global weather using graph neural networks. The system learns to step forward the current 3D atmospheric state by six hours, and multiple steps are chained together to produce skillful forecasts going out several days into the future. The underlying model is trained on reanalysis data from ERA5 or forecast data from GFS. Test performance on metrics such as Z500 (geopotential height) and T850 (temperature) improves upon previous data-driven approaches and is comparable to operational, full-resolution, physical models from GFS and ECMWF, at least…
Extreme rainfall is one of the major causes of natural hazards (for example flood, landslide, and debris flow) in the central Himalayan region, Nepal. The performance of strategies to manage these risks relies on the accuracy of quantitative rainfall estimates. Rain gauges have traditionally been used to measure the amount of rainfall at a given location. The point measurement often misrepresents the basin estimates, because of limited density and high spatial variability of rainfall fields across the Himalayas. The Department of Hydrology and Meteorology (DHM), Nepal has planned to install a …
Colorado State University（科罗拉多州立大学）；University of Minnesota（明尼苏达大学）
The earth system is exceedingly complex and often chaotic in nature, making prediction incredibly challenging: we cannot expect to make perfect predictions all of the time. Instead, we look for specific states of the system that lead to more predictable behavior than others, often termed "forecasts of opportunity." When these opportunities are not present, scientists need prediction systems that are capable of saying "I don't know." We introduce a novel loss function, termed the "NotWrong loss", that allows neural networks to identify forecasts of opportunity for classification problems. The N…
Colorado State University（科罗拉多州立大学）；University of Minnesota（明尼苏达大学）
The earth system is exceedingly complex and often chaotic in nature, making prediction incredibly challenging: we cannot expect to make perfect predictions all of the time. Instead, we look for specific states of the system that lead to more predictable behavior than others, often termed "forecasts of opportunity". When these opportunities are not present, scientists need prediction systems that are capable of saying "I don't know." We introduce a novel loss function, termed "abstention loss", that allows neural networks to identify forecasts of opportunity for regression problems. The abstent…
University of Cambridge（剑桥大学）；Vector Institute（矢量研究所）；University of Toronto（多伦多大学）；Microsoft Research（微软研究院）；British Antarctic Survey（英国南极调查局）；The Alan Turing Institute（阿兰·图灵研究所）；European Centre for Medium-Range Weather Forecasts（欧洲中期天气预报中心）；Google Deepmind（谷歌DeepMind）
This popular article provides a short summary of the progress and prospects in Weather and Climate Modelling for the benefit of high school and undergraduate college students and early career researchers. Although this is not a comprehensive scientific article, the basic information provided here is intended to introduce students and researchers to the topic of Weather and Climate Modelling - which comes under the broad discipline of Atmospheric / Oceanic / Climate / Earth Sciences. This article briefly summarizes the historical developments, progress, scientific challenges in weather and clim…
Bridging short- and medium-range weather forecasting with machine learning
National Oceanic and Atmospheric Administration (NOAA)（美国国家海洋和大气管理局）；Nansen Environmental and Remote Sensing Center（南森环境与遥感中心）；Bjerknes Center for Climate Research（比耶克内斯气候研究中心）；Earth Prediction Innovation Center (EPIC)（地球预测创新中心）；Cooperative Institute for Research in Environmental Sciences (CIRES) at the University of Colorado Boulder（科罗拉多大学博尔德分校环境科学合作研究所）；University of Colorado Boulder（科罗拉多大学博尔德分校）
AtmoRep: A stochastic model of atmosphere dynamics using large scale representation learning
CERN, European Center for Nuclear Research（欧洲核子研究中心）；Forschungszentrum Jülich（于利希研究中心）；Otto-von-Guericke-Universität Magdeburg（奥托·冯·格里克马格德堡大学）；European Centre for Medium Range Weather Forecasting（欧洲中期天气预报中心）
StormDiT: A generative AI model bridges the 2-6 hour 'gray zone' in precipitation nowcasting
Shanghai Typhoon Institute（上海台风研究所）；Key Laboratory of Numerical Modeling for Tropical Cyclone of the China Meteorological Administration（中国气象局热带气旋数值模拟重点实验室）；State Key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry (LAPC), Institute of Atmospheric Physics, Chinese Academy of Sciences（中国科学院大气物理研究所大气边界层物理与大气化学国家重点实验室）；College of Earth and Planetary Sciences, University of Chinese Academy of Sciences（中国科学院大学地球与行星科学学院）；Earth System Modeling and Prediction Center, China Meteorological Administration（中国气象局地球系统模式与预测中心）；State Key Laboratory of Severe Weather Meteorological Science and Technology（severe weather meteorological science and technology 国家重点实验室）；Gaoling School of Artificial Intelligence, Renmin University of China（中国人民大学高瓴人工智能学院）；Key Laboratory of Meteorological Disaster, Ministry of Education (KLME)/Joint International Research Laboratory of Climate and Environment Change (ILCEC)/Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters (CIC-FEMD), Nanjing University of Information Science and Technology（南京信息工程大学气象灾害教育部重点实验室/气候变化与环境国际合作联合实验室/气象灾害预报评估与协同创新中心）
Modern weather and climate models share a common heritage, and often even components, however they are used in different ways to answer fundamentally different questions. As such, attempts to emulate them using machine learning should reflect this. While the use of machine learning to emulate weather forecast models is a relatively new endeavour there is a rich history of climate model emulation. This is primarily because while weather modelling is an initial condition problem which intimately depends on the current state of the atmosphere, climate modelling is predominantly a boundary conditi…
University of Minnesota, Twin Cities（明尼苏达大学双城校区）；University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；George Mason University（乔治梅森大学）
Sub-seasonal climate forecasting (SSF) is the prediction of key climate variables such as temperature and precipitation on the 2-week to 2-month time horizon. Skillful SSF would have substantial societal value in areas such as agricultural productivity, hydrology and water resource management, and emergency planning for extreme events such as droughts and wildfires. Despite its societal importance, SSF has stayed a challenging problem compared to both short-term weather forecasting and long-term seasonal forecasting. Recent studies have shown the potential of machine learning (ML) models to ad…
