{"id":11085,"date":"2024-05-24T08:49:47","date_gmt":"2024-05-24T00:49:47","guid":{"rendered":"https:\/\/www.1ai.net\/?p=11085"},"modified":"2024-05-24T08:49:47","modified_gmt":"2024-05-24T00:49:47","slug":"%e5%be%ae%e8%bd%af-start-%e5%a4%a9%e6%b0%94%e9%a2%84%e6%b5%8b%e6%a8%a1%e5%9e%8b%e5%86%8d%e8%bf%9b%e5%8c%96%ef%bc%8c%e5%8f%af%e9%a2%84%e6%b5%8b%e4%bd%95%e6%97%b6%e5%87%ba%e7%8e%b0%e4%ba%91%e5%b1%82","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/11085.html","title":{"rendered":"Microsoft's Start weather forecast model evolves to predict when clouds and precipitation will appear"},"content":{"rendered":"

Microsoft<\/a> The Start team announced today via a blog post on the official Bing blog that the AI Weather prediction models<\/a>In addition to significant improvements that allow for more accurate forecasting of weather conditions for the next 30 days.Now predict when clouds and precipitation are likely to occur.<\/strong><\/p>\n

\"e231180a-6db9-4173-ad8c-efacca0b3ebf\"<\/p>\n

In the blog post, the Microsoft Start team says it has begun work on improving its worldwide \"precipitation forecasting\" system by the end of 2021.<\/p>\n

The model will not only call on data collected by the radar device, but will also combine it with satellite data and analyze it through AI algorithms to more accurately predict precipitation.<\/p>\n

However, the Microsoft Start team also admits that the prediction model has shortcomings, with satellite weather data being received on a daily time scale of 85% to 95%, depending on the source of the data and acceptable delays.<\/p>\n

Microsoft said:<\/p>\n

Evidence suggests that a separate decoder is needed for each task and a separate discriminator for each forecast channel, so Weather from Microsoft Start builds a model that is four times larger than the previous model that predicted only analog radar reflectivity. Finally, the new model jointly predicts satellite and analog radar reflectivity, allowing its predictions to fill gaps in data availability.<\/p>\n

According to Microsoft, the radar channel model was given six times the weight of the satellite model in the AI training because the precipitation information provided by the radar was considered more important than the satellite mission. The end result was that the new precipitation and cloud cover models \"showed significant improvement in F1 scores\" compared to the base model that used only radar forecasts.<\/p>","protected":false},"excerpt":{"rendered":"

In a post published today on the official Bing blog, the Microsoft Start team announced another major improvement to its AI weather forecasting model, which now predicts when clouds and precipitation are likely to occur, in addition to more accurately forecasting the weather for the next 30 days. In the post, the Microsoft Start team said it has been working on improving its \"precipitation forecasting\" system worldwide since the end of 2021. The model will not only use data collected by radar devices, but will also incorporate satellite data, which will be analyzed by AI algorithms to more accurately predict precipitation. However, the Microsoft Start team also admits that the forecasting model has shortcomings, as satellite weather data is received on a daily basis at a rate of 85% to 95%, depending on the source of the data and acceptable delays. Microsoft said<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[2504,280],"collection":[],"class_list":["post-11085","post","type-post","status-publish","format-standard","hentry","category-news","tag-2504","tag-280"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/11085","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=11085"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/11085\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=11085"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=11085"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=11085"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=11085"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}