Potential Improvement of GK2A Clear-Sky Atmospheric Motion Vectors Using the Convolutional Neural Network Model

• DOI: 10.1007/s13143-023-00349-x
• Publication Date: 2024-02-08T16:02:20Z
• Authors: Hwayon Choi; Yong-Sang Choi; Hyo-Jong Song; Hyoji Kang; Gyuyeon Kim
• Keywords: Optical Flow

Abstract In this study, we propose a new approach to improve the accuracy of horizontal atmospheric motion vector (AMV) in cloud-free skies and its forecasting. We adapted optical flow convolutional neural network (CNN) framework model using two 10-min interval infrared images at water vapor channels (centered 6.3, 7.0, 7.3 $$\mu m$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>μ</mml:mi> <mml:mi>m</mml:mi> </mml:mrow> </mml:math> ) from Korean geostationary satellite GEO-KOMPSAT-2A (GK2A). Since all pixels had seamless AMVs calculated by CNN (CNN AMVs), could also predict linear regression method. The tracking performance CNN-based algorithm was validated retrieved GK2A (GK2A AMVs) estimating difference between those values ECMWF (European Centre for Medium-Range Weather Forecasts) Reanalysis v5 (ERA5) wind data over Korea 2022. showed similar or better root-mean-square differences (RMSVDs) than (12.33–12.86 vs. 15.89–19.96 m/s). RMSVDs forecasted were 2.74, 2.95, 3.41, 4.79 m/s lead times 10, 20, 30, 60 min, respectively. Consequently, our method higher production succeeded forecasting AMVs. expect that such potential improvements computational operational will contribute increased when meteorological phenomena related wind.