Through its subsidiary The Weather Company, the computing titan has partnered with the University Corporation for Academic Research (UCAR) and the National Center for Atmospheric Research (NCAR) to move beyond today's regional-scale forecasting to anticipate weather at the local level...and aspire to introduce the first model that covers the whole globe. The model this IBM collaboration wants to build would account for the influence smaller events (like thunderstorms) have on local weather. IBM will join with UCAR to co-design a computational solution that runs on the former's POWER9-based systems, which are set to launch at the end of the year. The Weather Company will use this computing powerhouse to adapt NCAR's community-weather model to a global scale and refine the longer-term predictions to make more accurate forecasts weeks or months out.
The legislation empowers the National Oceanic Atmospheric Administration (NOAA) to boost its ability to predict major weather-related events, such as hurricanes, droughts, floods and wildfires. Using faster, more powerful computers and more detailed data of weather patterns could increase the accuracy, Seitter says. Businesses have been able to access accurate, customizable weather forecasting online only in the last decade or so, says Bill Gail, chief technology officer at private forecaster Global Weather Corporation. Xcel Energy, who uses Gail's firm to anticipate wind energy production, improved its wind forecasting accuracy by nearly 35% from 2009 to 2015.
Efficient demand forecasting, which predicts future demand for products and parts based on past events and prevailing trends, is a key component to after-sales service success. In after-sales service organizations, it's all too common for service parts planners to lack visibility into supply or demand between different stocking locations, leading to forecast accuracy issues and problems maintaining a balanced inventory, and most notably, increased costs. When a manufacturer is introducing a new product, machine learning can use algorithms and analytics to track and determine the launch's success, incorporating data from sales, social media chatter and web traffic, among other sources. Machine learning is the next stage in supply chain business intelligence, specifically for after-sales service.
Both government and private weather forecasting companies are approaching the point where they get tomorrow's high temperature right nearly 80 percent of the time. Both government and private weather forecasting companies are approaching the point where they get tomorrow's high temperature right nearly 80 percent of the time. It was 66 percent 11 years ago, according to ForecastWatch, a private firm that rates accuracy of weather forecasts. It was 66 percent 11 years ago, according to ForecastWatch, a private firm that rates accuracy of weather forecasts.
In 2014, VELCO began work with IBM, Vermont's distribution utilities, and other partners to develop VWAC, which integrates IBM's precise weather forecasting with Vermont's customer load data and output data from the state's renewable generators to turn this mass of data into actionable information using leading-edge analytics. Finally, VWAC provides greater visibility to potential demand response events based on demand forecasts, from the substation to the distribution service territory to the statewide level. Linking VWAC output to VELCO's energy management system will improve core grid reliability and more accurately integrate weather-dependent generation in it. Reduced uncertainty of production from behind-the-meter (BTM) solar will result in better commitment and dispatch decisions by ISO New England, reducing production costs and avoiding unnecessary supplemental commitments and fuel consumption.
The National Oceanic and Atmospheric Administration is predicting an above-normal 2017 Hurricane Season, with five to nine hurricanes -- two to four of them Category 3 (winds at least 111 mph) or stronger. The weakness or absence of storm-suppressing El Niño climate conditions, combined with above-normal ocean surface temperatures and average or weaker vertical wind shear across the Caribbean and Atlantic Coast are factors pointing to an active hurricane season, said Ben Friedman, acting NOAA administrator. They expect that three named storms will make landfall in the U.S. April's Tropical Storm Arlene was a rare preseason storm, but it was also an indication of an active season ahead, Friedman said Thursday during a news conference at the NOAA Center for Weather and Climate Prediction in College Park, Md. In the 25 years since Category 5 Hurricane Andrew hit South Florida, forecasting accuracy has improved 65%, said Mary Erickson, deputy director at the National Weather Service.
This week's Featured Blog Friday comes from our Reykjavik University student intern, Guðbjörn Einarsson aka Mannsi, who has been working closely with our Data Scientist, Agnes Jóhannsdóttir, to implement Machine Learning technology into our AGR software. As always, if you have any questions or comments regarding this blog post, feel free to comment on this blog post, tweet us @AGRDynamics, or contact us here. AGR Dynamics is certain Machine Learning will play a big role in the future of our business. If you haven't already read through the other Machine Learning blog posts on Recommender Systems and Introduction To Machine Learning you really should, as they are great. Another area where Machine Learning can be applied is sales forecasting. Here we would like to briefly explain how that works and go through the pros and cons. The most common approach is to use a method called Neural Network. Neural Networks are designed to mimic how the human brain operates and learns and is one of ...
A prototype of the earthquake early warning system gave researchers in San Francisco about eight seconds of warning before shaking began. A prototype of the earthquake early warning system gave researchers in San Francisco about eight seconds of warning before shaking began. President Trump's budget would eliminate federal funding for an earthquake early warning system being developed for California and the rest of the West Coast, which if enacted would likely kill the long-planned effort. The budget proposal for the year ending in September 2018 also seeks to eliminate U.S. funding for critical tsunami-monitoring stations in oceans and reduce funds for a next-generation weather forecasting system.
As the White House considers its choice to lead the National Oceanic and Atmospheric Administration (the parent agency of the weather service), many climate scientists and weather experts want to make sure that the agency maintains its quest to catch up to the Europeans. They also don't want NOAA climate research and data collection--which make weather models work better--lose any more money. On Tuesday, the White House's budget proposal suggested slicing 26 percent from NOAA's Office of Oceanic and Atmospheric Research, which supports data collection, basic climate and oceanographic science, and research into more accurate weather forecasting models. Mass and others argue that cutting basic research into the oceans, atmosphere, and climate--the taxpayer-funded research done by NOAA and NWS--will lead to less reliable weather modeling by private firms like AccuWeather as well as federal models like the Global Forecast System.