Author: user

Visualising the intensity and progression of hurricanes on digital platforms has seen an impressive upgrade. A diligent researcher has created a set of distinctive icons that represent the categories of the Saffir-Simpson Hurricane Wind Scale. They are hosted on a GitHub repository called hurricane-net and can be found here.

These icons, interestingly, are not just static images. They are imbued with Generative AI tech called Midjourney to create dynamic backgrounds. Midjourney was chosen over other generative models like Stable Diffusion and Dall E due to its superior performance in generating diverse yet cohesive imagery.

The choice of the Saffir-Simpson scale as a basis for these icons was driven by its color scheme, which transitions from yellow to red as the severity increases from category 1 through 4. Category 5, signifying the most potent storms, is represented by the color purple. The researcher ingeniously crafted the backgrounds of the icons to echo these color scales, resulting in an immediate visual understanding of a storm’s intensity.

These icons serve a greater purpose than just aesthetic enhancement. They play a vital role in the fluids.ai application, where they help visualize live global tropical storms. The variability of the storm’s intensity over time, which can often be dramatic, can now be visualized effortlessly, enhancing the user experience in emergency scenarios.

The icons’ innovative design allows users to gain a quick understanding of the storm’s severity as it approaches the shore, adding a new dimension to how we perceive and react to such natural disasters.

Moreover, the source code is readily available to developers who wish to contribute or adapt these icons for their projects, exemplifying the spirit of open-source and collaborative problem-solving.

Indeed, the intersection of art, technology, and meteorology, as demonstrated in this project, offers a fresh perspective on disaster visualization, making information more digestible and impactful for all.

Recorded Presentation
Recording

• 11B.2: A Deep Neural Network to Globally Forecast the Track and Intensity of Tropical Cyclones

• Boston Convention and Exhibition Center
• – 156A

• Hammad Usmani
  • Georgia Institute of Technology
    Atlanta, GA, USA
• Aadil Habibi
  • Univ. of Central Florida
    Orlando, FL, USA
• Daanish Habibi
  • Univ. of Central Florida
    Orlando, FL, USA

Tropical cyclones are the most devastating weather phenomenon and the IPCC’s 2018 “Special Report on Global Warming of 1.5°C” there’s evidence that extreme tropical events are likely to worsen. With machine learning, producing skillful forecasts becomes possible provided the right data. We can create a deep neural network that fully utilizes recurrent and convolutional layers using the IBTrACS database and the NCEP/NCAR Surface Temperature imagery. This artificial intelligence is accompanied by a standard web application that can be used operationally to produce forecasts. The study develops under a permissive license that allows reuse and maintains open source. The study follows the I18n internationalization format to assist with global dissemination. The model produces near-instant track and intensity forecasts for Atlantic, Indian, and Pacific Ocean tropical cyclones and has shown skill over the NHC’s statistical baseline for Atlantic tropical storms. One of the goals of the research is to provide both professional and amateur meteorologists access to extreme tropical cyclone forecasts to assist with emergency scenarios.

J1.6A A Deep Recurrent Neural Network to Forecast the Intensity and Trajectory of Atlantic Tropical Storms

More Wednesday, 9 January 2019: 9:45 AM North 124B (Phoenix Convention Center – West and North Buildings) Hammad Usmani, Georgia Institute of Technology, Atlanta, GA Recorded Presentation The National Hurricane Center (NHC) and National Oceanic and Atmospheric Administration (NOAA) provide predictions for storms trajectories, intensity, and size. They create these predictions based on models that can be classified into 3 groups: dynamical, statistical, and ensemble. Classifications also include relative compute time required to create an output grouped as either early or late and forecast parameters such as trajectory, intensity, and wind radii. The most accurate models are late models that take upwards of 6 hours to produce an output whereas models that can produce an output in seconds are called early. Early models are crucial for emergency scenarios because of their timeliness. The statistical baseline models such as OCD5 are based on multivariate regressors that can explain a significant amount of variance. The performance for these methods can be augmented by incorporating more advanced statistical methods from deep learning such as recurrent neural networks. In this study, we research and implement the domain of machine learning and deep learning into Atlantic storm forecasting for both trajectory and intensity and evaluate them against the NHC standards. Previous research into machine learning to forecast tropical Atlantic storms includes a sparse recurrent neural network (Kordmahalleh, Sefidmazgi, & Homaifar, 2016) and an artificial neural network (Jung & Das, 2013) which achieved favorable results. Tropical system models created with deep neural networks can be utilized to develop more precise emergency planning. Because of this, there is a necessity for more accurate and timely models that can help reduce the amount of loss caused by tropical storms and hurricanes. The results of this study provide a reproducible bidirectional deep recurrent neural network implementing LSTM cells (BDRNN) forecasting both the trajectory and intensity of Atlantic storms utilizing training and testing data from the Atlantic hurricane database (HURDAT2). The model forecasts were evaluated based on 2017 data and indicated skill by performing better than the statistical baseline of OCD5. The BDRNN model can be used to provide improved decision support for emergency responses because accurate forecasts can be produced on demand. The study provides a promising framework for additional research that can incorporate satellite imagery and different domains such as the North Central Pacific.

Please stay tuned for more updates on an interactive map.