Abstract:This study presents a cost‐effective machine‐learning model for automated cloud‐type classification using ground‐based global horizontal irradiance (GHI) measurements and a clear‐sky model. Data labeling of cloud types was obtained from all‐sky images by meteorological observers. Eight features were extracted from the GHI "signature" into a determined time window around the datetime of each all‐sky image. The XGBoost classifier on 33‐min windows was the most effective, achieving a remarkable accuracy of 0.88 and a Cohen's kappa of 0.84. Cloud observations and characterization are crucial owing to their influence on energy balance, climate, and weather. Their particular effects on radiation vary depending on different cloud parameters, such as cloud base or top height, water content, and cloud optical thickness, all of them closely related to the specific cloud type. Cloud classification therefore becomes a crucial task in meteorology, although it remains challenging for weather services worldwide owing to the intensive associated labor and cost. In this study we introduce a new low‐cost method for automating cloud classification based on a combination of ground‐based global horizontal irradiance (GHI) measurements, a clear‐sky model, and machine learning. Based on the hypothesis that different cloud types have their own GHI signatures, we trained different supervised learning algorithms using GHI data manually labeled by meteorological observers from time‐synchronized all‐sky images. Multiple time windows were extracted from each GHI series, with eight features defined in each case to characterize the sequence. The best outcome was achieved using an XGBoost model on features extracted on time windows of 33 min, obtaining an accuracy of 0.88 and a Cohen's kappa of 0.84 in a held‐out test set. The development presented in this study has the ability to provide low‐cost cloud classification from ground‐based observations, which is a challenge for weather services worldwide owing to intensive labor and cost.

Segmentation Algorithms for Ground-Based Infrared Cloud Images

Processing of Global Solar Irradiance and Ground-Based Infrared Sky Images for Solar Nowcasting and Intra-Hour Forecasting Applications

Data acquisition and image processing for solar irradiance forecasting

BenchCloudVision: A Benchmark Analysis of Deep Learning Approaches for Cloud Detection and Segmentation in Remote Sensing Imagery

Effective Cloud Detection and Segmentation using a Gradient-Based Algorithm for Satellite Imagery; Application to improve PERSIANN-CCS

Detection of clouds in multiple wind velocity fields using ground-based infrared sky images

An Improved Cloud Recognition and Classification Method for Photovoltaic Power Prediction Based on Total-Sky-images

Review of Kernel Learning for Intra-Hour Solar Forecasting with Infrared Sky Images and Cloud Dynamic Feature Extraction

Cloud detection method using ground-based sky images based on clear sky library and superpixel local threshold

Very Short-Term Forecast: Different Classification Methods of the Whole Sky Camera Images for Sudden PV Power Variations Detection

Cloud Segmentation and Matching Using Deep Learning in All-Sky Images

A novel intra-hour PV power forecasting technique based on total-sky images

Efficient Region Segmentation of PV Module in Infrared Imagery using Segnet

Computational Imaging for Long-Term Prediction of Solar Irradiance

Ground Sky Imager Based Short Term Cloud Coverage Prediction

Deeplab V3+ Based Segmentation Method for PV Panels with Aerial Orthoimages

Sky Images for Short-Term Solar Irradiance Forecast: A Comparative Study of Linear Machine Learning Models

Cloud classification through machine learning and global horizontal irradiance data analysis

Using sky-classification to improve the short-term prediction of irradiance with sky images and convolutional neural networks

End-to-end Cloud Segmentation in High-Resolution Multispectral Satellite Imagery Using Deep Learning

Hybrid Solar Forecasting Method Using Satellite Visible Images and Modified Convolutional Neural Networks