Tropical cyclone intensity estimation is a challenging task as it required domain knowledge while extracting features, significant pre-processing, various sets of parameters obtained from satellites, and human intervention for analysis. The inconsistency of results, significant pre-processing of data, complexity of the problem domain, and problems on generalizability are some of the issues related to intensity estimation. In this study, we design a deep convolutional neural network architecture for categorizing hurricanes based on intensity using graphics processing unit. Our model has achieved better accuracy and lower root-mean-square error by just using satellite images than 'state-of-the-art' techniques. Visualizations of learned features at various layers and their deconvolutions are also presented for understanding the learning process.

E ach year across the US, mesoscale weather events-flash floods, tornadoes, hail, strong winds, lightning, and localized winter storms-cause hundreds of deaths, routinely disrupt transportation and commerce, and lead to economic losses averaging more than US$13 billion.1 Although mitigating the impacts of such events would yield enormous economic and societal benefits, research leading to that goal is hindered by rigid IT frameworks that can't accommodate the real-time, on-demand, dynamically adaptive needs of mesoscale weather research; its disparate, high-volume data sets and streams; or the tremendous computational demands of its numerical models and data-assimilation systems.In response to the increasingly urgent need for a comprehensive national cyberinfrastructure in mesoscale meteorology-particularly one that can interoperate with those being developed in other relevant disciplines-the US National Science Foundation (NSF) funded a large information technology research (ITR) grant in 2003, known as Linked Environments for Atmospheric Discovery (LEAD). A multidisciplinary effort involving nine institutions and more than 100 scientists, students, and technical staff in meteorology, computer science, social science, and education, LEAD addresses the fundamental research challenges needed to create an integrated, scalable framework for adaptively analyzing and predicting the atmosphere.LEAD's foundation is dynamic workflow orchestration and data management in a Web services framework. These capabilities provide for the use of analysis tools, forecast models, and data repositories,

No abstract

Open science, as both a concept and a term, is increasing in popularity and usage. However, definitions, interpretations, and perceptions as to what the term "open science" means varies. Some definitions are fairly narrow and only focus on providing more open access to science as a body of knowledge. These narrow definitions place an emphasis on openly sharing scientific knowledge as early as possible in the research process (University of Cambridge, 2020). On the other hand, broader definitions of open science acknowledge that science is both a body of knowledge and a systematic method for thinking. Broad definitions place an emphasis on encouraging a culture of openness (Bartling & Friesike, 2014) that includes the entire process of conducting science (National Academies of Sciences, Engineering, & Medicine, 2018a, 2018b) and encourages open collaboration and access to knowledge (Vicente-Saez & Martinez-Fuentes, 2018). In its broadest definition, the term "open science" refers to a paradigm shift in how the methods of science are conducted. This expansive vision of open science acknowledges that rapid technology changes, primarily driven by the Internet, may enable a second scientific revolution that fundamentally changes research methods and standards across science. To complicate matters, the term "open science" is sometimes used interchangeably to represent various principles that support the broader idea of open science itself. These principles include ideas such as open data, open source software, open journal access, and reproducibility. For example, reproducibility, or the ability to verify another scientist's results, is enabled by the principles of open data, open code, and transparent methodologies, yet reproducibility itself is not equivalent to open science.While open science definitions are variable and ambiguous, the value of open science as both a concept and a paradigm change is accepted by the majority of the scientific community. Open science not only benefits the scientific endeavor itself but has also been shown to benefit individual researchers through increased citations and media attention, a larger collaborative network, and exposure to new career and funding opportunities (

No abstract