Gridded Visibility Products over Marine Environments Based on Artificial Neural Network Analysis

Abstract

The reconstruction and monitoring of visibility over marine environments is critically important because of a lack of observations. To travel safely in marine environments, a high quality of visibility data is needed to evaluate navigation risk. Currently, although visibility is available through numerical weather prediction models as well as ground and spaceborne remote sensing platforms and ship measurements, issues still exist over the remote marine environments and northern latitudes. To improve visibility prediction and reduce navigational risks, gridded visibility data based on artificial neural network analysis can be used over marine environments, and the problem can be regarded as an air quality prediction problem based on machine learning algorithms. This new method based on artificial intelligence techniques developed here is tested over the Indian Ocean. The mean error of the inferred visibility from the artificial neural network analysis is found to be less than 8.0%. The results suggested that satellite-based optical thickness and numerical model-based reanalysis data can be used to infer gridded visibility values based on artificial neural network analysis, and that could help us reconstruct and monitor surface gridded visibility values over marine and remote environments.

Machine-learning regression applied to diagnose horizontal visibility from mesoscale NWP model forecasts

“…Visibility forecasts are important to many customers and are a key requirement for all forms of transport, especially within the fields of land, aviation and shipping, though the range of adverse visibility thresholds of interest varies [1,2]. Of particular interest is the reliable forecasting of poor visibility, i.e.…”

“…Statistical supervised machine-learning (ML) techniques rely on a training period of historical data which connect forecasts and observed visibility. In this framework, machine-learning algorithms are emerging as suitable methods for detection and prediction of meteorological phenomena [2,15,16]. Some machine-learning methods have been used for low-visibility forecasting, for example artificial networks [17], multiple linear regression [18], and tree-based methods [16].…”

Machine Learning-Based Front Detection in Central Europe

“…The random forest method [35,36]-an ensemble machine learning method based on the construction of many decision trees that is widely used for many applications in meteorology [37][38][39][40], climatology [41,42], medicine [43,44], renewable energy [45][46][47], and many other fields-was used to build a model that combined meteorological parameters from the ERA5 dataset with the positions of fronts from digitized DWD maps. Since atmospheric conditions differ significantly between weather seasons in Central Europe, our analyses were performed separately for winter (DJF), spring (MAM), summer (JJA), and autumn (SON).…”