In the previous article, we talked about GOA-UVa, the Atmospheric Optics Group of the University of Valladolid, whose facilities we had the pleasure of visiting last Friday, June 28, guided by Roberto Román.
That day, we also learned about the project they are developing: Presente. It is a project that will leave no one indifferent and is already starting to bear fruit.
Project Objectives
This project aims to set up a network of sky cameras in Valladolid, with approximately one camera every 2 km. Why do they want to build a network of cameras? To be able to predict the movement of clouds. And again, what is this for? To improve solar panel installations, leading to smarter solar energy production.
Too much to take in? This university group does all this and much more. In this article, I will explain this project step by step.
Background
First, we need to know a few things about sky cameras. As we mentioned in the previous article, a sky camera takes pictures of the sky, including clouds. With a photo showing the sky and clouds, we can determine the direction, from the central point where the camera is located, in which a cloud is found.
Sky camera. Photo by Lucas Pérez.
Knowing the direction of a cloud alone doesn’t achieve much, but if we combine two cameras and the angles of the clouds, we can start triangulating the position of the clouds using trigonometric calculations.
Image transformation. Photo by Lucas Pérez.
Development and Progress
Initially, it was planned to install 25 cameras at different points in the province of Valladolid, but so far, only 14 cameras have been installed. When placing new devices, they encounter some of the following issues:
- There are no buildings (preferably public) in the selected area.
- If there are buildings:
- They are reluctant (this rarely happens).
- It’s a bad location because it is isolated (no Internet access) or has trees that continuously cast shadows on the camera.
Map showing the locations of the different cameras. Photo by Lucas Pérez.
From the GOA-UVa website, we can see all the stations of the Proyecto Presente, as well as views from the different cameras in the city.
Sky camera at CEIP Gonzalo de Berceo with lightning. Photo by Lucas Pérez.
What You Don’t See
The university group has developed software that allows determining the height of an exact point (clicked in the view of a camera) by selecting two cameras. For example, from this software, we select a cloud seen from the camera at IESO Arroyo and choose as the second camera the one located at CEIP El Peral. This way, the software will find the same point from the view of the camera at CEIP El Peral and, using mathematical calculations, determine the height of the cloud.
Photo of the sky with a sky camera at the Faculty of Sciences on 08/28/2024 at 15:25 UTC. Photo by Lucas Pérez.
With this programming, they aim to create an artificial intelligence model capable of predicting the movement of clouds. As mentioned earlier, this will help improve solar panel installations, producing more energy since, according to the meteorological prediction model, we can know where the clouds are and what radiation will be in that location based on the following statements:
- If the clouds block the sun: not enough solar radiation reaches the photovoltaic panel.
- If the clouds surround the sun: more radiation reaches the panel, as it “bounces” off these clouds and ends up hitting the installation.
- If there are no clouds: a normal level of solar radiation reaches the panel.
General view of a cloud. Photo by Lucas Pérez.
History
They have been working with this instrument for a long time and had the idea several times, but they never managed to secure funding. When the opportunity to fund the project arose, they jumped at it and were granted the funding to build a network of cameras of this magnitude.
Roberto studied Physics and initially had little interest in Environmental Physics, preferring Material Physics. However, he later started working in Environmental Physics and liked it. He also tells us that “science is like a Ferris wheel,” meaning that “many times, the expected results don’t come, or one gets stuck without finding a solution, but other times, when interesting results are achieved, whether expected or not, it is very rewarding and a great learning experience.” He adds that science is never complete: “solving one question often leads to even more questions.”