Using machine learning to find cloud‐base height: a didactic challenge. (25th February 2022)

• Record Type: Journal Article
• Authors: Lewis, Helena; Bowyer, Jack; Broad, Alistair Louis; Chamberlain‐Clay, Alex; Jones, Caroline; Chan, Steven; Kahraman, Abdullah; Morcrette, Cyril
• Is Part Of: Weather. Volume 77:Number 11(2022)
• Page Start: 391
• Page End: 395
• Publication Date: 2022-02-25
• Subjects: Meteorology -- Periodicals; Weather -- Great Britain -- Periodicals; 551.1
• DOI: 10.1002/wea.4163
• Languages: English
• ISSNs: 0043-1656

Abstract

Abstract : This is an example of the training data used to learn how to predict lowest cloud‐base height using a neural network. The Illustration shows 1300 columns that form part of a training data array. Each column consists of 280 rows, comprising 70 rows each of standardised temperature, humidity and pressure and 70 rows indicating the location of lowest cloud base in binary format. To benefit those wishing to begin exploring machine learning in an atmospheric science context, we provide a dataset and some example code, which can be used to train a neural network to predict the height of cloud base given profiles of temperature, humidity and pressure.