Classification of tropical cyclone containing images using a convolutional neural network: performance and sensitivity to the learning dataset

Sébastien Gardoll, Oliviér Boucher

Geoscientific model development · 16 September 2022

In one sentence. CNN classifies ERA5 and MERRA-2 fields for tropical cyclone presence, and ERA5-trained models transfer better to MERRA-2 than the reverse.

Abstract

Abstract. Tropical cyclones (TCs) are one of the most devastating natural disasters, which justifies monitoring and prediction on short and long timescales in the context of a changing climate. In this study, we have adapted and tested a convolutional neural network for the classification of reanalysis outputs according to the presence or absence of TCs. We use a number of meteorological variables to form TC-containing and background images from both the ERA5 and MERRA-2 reanalyses. The presence of TCs is labelled from the HURDAT2 dataset. Special attention was paid on the design of the background image set to make sure it samples similar location and time to the TC-containing image. We have assessed the performance of the CNN using accuracy but also the more objective AUC and AUPRC metrics. Many failed classifications can be explained by the meteorological context, such as a situation with cyclonic activity but not yet classified as TC by HURDAT2. We also tested the impact of interpolation and of mix and match the training and test image sets on the performance of the CNN. We showed that applying an ERA5-trained CNN on MERRA-2 images works better than applying a MERRA-2 trained CNN on ERA5 images.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.