Semantic segmentation of ground-based all-sky images (ASIs) can provide high-resolution cloud coverage information of distinct cloud types, applicable for meteorology-, climatology- and solar-energy-related applications. Since the shape and appearance of clouds is variable, and there is high similarity between cloud types, a clear classification is difficult. Therefore, most state-of-the-art methods focus on the distinction between cloudy and cloud-free pixels without taking into account the cloud type. On the other hand, cloud classification is typically determined separately at the image level, neglecting the cloud's position and only considering the prevailing cloud type. Deep neural networks have proven to be very effective and robust for segmentation tasks; however they require large training datasets to learn complex visual features. In this work, we present a self-supervised learning approach to exploit many more data than in purely supervised training and thus increase the model's performance. In the first step, we use about 300 000 ASIs in two different pretext tasks for pretraining. One of them pursues an image reconstruction approach. The other one is based on the DeepCluster model, an iterative procedure of clustering and classifying the neural network output. In the second step, our model is fine-tuned on a small labeled dataset of 770 ASIs, of which 616 are used for training and 154 for validation. For each of them, a ground truth mask was created that classifies each pixel into clear sky or a low-layer, mid-layer or high-layer cloud. To analyze the effectiveness of self-supervised pretraining, we compare our approach to randomly initialized and pretrained ImageNet weights using the same training and validation sets. Achieving 85.8 % pixel accuracy on average, our best self-supervised model outperforms the conventional approaches of random (78.3 %) and pretrained ImageNet initialization (82.1 %). The benefits become even more evident when regarding precision, recall and intersection over union (IoU) of the respective cloud classes, where the improvement is between 5 and 20 percentage points. Furthermore, we compare the performance of our best model with regards to binary segmentation with a clear-sky library (CSL) from the literature. Our model outperforms the CSL by over 7 percentage points, reaching a pixel accuracy of 95 %.

Solar forecasting constitutes a critical tool for operating, producing and storing generated power from solar farms. In the framework of the International Energy Agency’s Photovoltaic Power Systems Program Task 16, the solar irradiance nowcast algorithms, based on five all-sky imagers (ASIs), are used to investigate the feasibility of ASIs to foresee ramp events. ASIs 1–2 and ASIs 3–5 can capture the true ramp events by 26.0–51.0% and 49.0–92.0% of the cases, respectively. ASIs 1–2 provided the lowest (<10.0%) falsely documented ramp events while ASIs 3–5 recorded false ramp events up to 85.0%. On the other hand, ASIs 3–5 revealed the lowest falsely documented no ramp events (8.0–51.0%). ASIs 1–2 are developed to provide spatial solar irradiance forecasts and have been delimited only to a small area for the purposes of this benchmark, which penalizes these approaches. These findings show that ASI-based nowcasts could be considered as a valuable tool for predicting solar irradiance ramp events for a variety of solar energy technologies. The combination of physical and deep learning-based methods is identified as a potential approach to further improve the ramp event forecasts.

Cloud base height (CBH) is an important parameter for many applications such as aviation, climatology or solar irradiance nowcasting (forecasting for the next seconds to hours ahead). The latter application is of increasing importance for the operation of distribution grids and photovoltaic power plants, energy storage systems and flexible consumers. To nowcast solar irradiance, systems based on all-sky imagers (ASIs), cameras monitoring the entire sky dome above their point of installation, have been demonstrated. Accurate knowledge of the CBH is required to nowcast the spatial distribution of solar irradiance around the ASI's location at a resolution down to 5 m. To measure the CBH, two ASIs located at a distance of usually less than 6 km can be combined into an ASI pair. However, the accuracy of such systems is limited. We present and validate a method to measure the CBH using a network of ASIs to enhance accuracy. To the best of our knowledge, this is the first method to measure the CBH with a network of ASIs which is demonstrated experimentally. In this study, the deviations of 42 ASI pairs are studied in comparison to a ceilometer and are characterized by camera distance. The ASI pairs are formed from seven ASIs and feature camera distances of 0.8…5.7 km. Each of the 21 tuples of two ASIs formed from seven ASIs yields two independent ASI pairs as the ASI used as the main and auxiliary camera, respectively, is swapped. Deviations found are compiled into conditional probabilities that tell how probable it is to receive a certain reading of the CBH from an ASI pair given that the true CBH takes on some specific value. Based on such statistical knowledge, in the inference, the likeliest actual CBH is estimated from the readings of all 42 ASI pairs. Based on the validation results, ASI pairs with a small camera distance (especially if <1.2 km) are accurate for low clouds (CBH<4 km). In contrast, ASI pairs with a camera distance of more than 3 km provide smaller deviations for greater CBH. No ASI pair provides the most accurate measurements under all conditions. The presented network of ASIs at different distances proves that, under all cloud conditions, the measurements of the CBH are more accurate than using a single ASI pair.