Monitoring Algal Blooms in the Sandbox (Jupyter Lab) with Python – Excerpts
Today, 6th August 2024, GeoVille, in conjunction with Space in Africa, hosted a webinar titled “Monitoring Algal Blooms the Sandbox (Jupyter Lab) with Python ” to train users on how to navigate the Example Notebooks for facilitating sustainable algal ecosystems. Led by Johannes Schmid, IT and Operations Manager at Geoville, participants were taken through the process of accessing Sentinel data and creating applications using Example Notebooks on the DUNIA platform.
Visit here to watch the webinar.
Johannes Schmid started with a short introduction of the DUNIA platform, an all-in-one initiated and designed by the European Space Agency in collaboration with GeoVille and other partners from France, Poland, and Romania to empower the African EO community by making EO data easily accessible for experts, students, and people interested in creating EO applications for sustainable development. The platform is outfitted with the infrastructure and datasets users require to create and exchange their ideas and applications and connect with other creators and experts through the DUNIA marketplace feature.
Thus, the DUNIA platform is for experts and beginners in Africa who are interested in earth observation data for space applications for professional and personal use. The platform also constantly evolves to include more training events, functionalities, datasets, images, and bands to ensure the African DUNIA community grows. The platform also has stringent data protection laws to ensure your application and work process are only visible to you until you share them. To get started, sign up for the DUNIA platform.
Key takeaways from the training
Johannes further takes us through the process of accessing the sandbox from the platform and highlights that every new DUNIA sandbox user gets a free Mahala 1x large machine, which consists of 4 CPUs, 60 gigabytes of RAM, and 50 gigabytes of object storage. Upon sign-up, users from any GMES and Africa consortium automatically get a premium account. The platform may enable new features and resources for extra CPUs or GPUs for training models. The Algal Bloom Example Notebook is a Python-driven resource in the Jupiter Lab, which is great for presenting and developing applications.
Johannes Schmid noted that Algal blooms deplete the oxygen in water, thereby leading to the death of marine organisms; thus, they are a hazard that affects aquatic systems and often results in ecological and economic stress on local fishing communities and public health issues for human swimmers. They need to be constantly monitored, which is only possible using satellite EO data. Through EO images, scientists can see the evidence and progression of algal blooms in waters and create appropriate mitigation strategies.
During the training, Johanne went through a step-by-step process of uplinking and downlinking data on the DUNIA platform, accessing the Sentinel 3 optical and infrared data and specific datasets, preprocessing data using histogram normalisation to increase contrast on the images and detecting small changes, applying algorithms to create chlorophyll concentration results, and then visualising the data. He also explained that the chlorophyll pigment in a water body mainly indicates the severity of algal blooms in such an ecosystem.
When addressing the capability of the Algal Blooms notebook for multiple regions, Johannes noted that it is easier for users to develop the application region by region, but this is optional for the processing part of the task. An application app can also be used for the task. To confirm if your applications work for other regions, the geometry of the area and dates of interest can be changed from the notebook, and everything should be similar.
According to Johannes, the DUNIA platform is constantly updated to make workflows as seamless and easy as possible for users, and it supports any application or process if it already has the required datasets. Furthermore, features of the platform, such as data access, can be used separately and adapted to various EO-related services; as such, the developers are open to looking into how to help users integrate DUNIA into existing services or platforms.
Existing and new users already creating and building on DUNIA stand a chance to have their work promoted and published on the DUNIA and ESA websites, thereby gaining acclaim in the global EO community. Send an email to mustapha@spaceinafrica.com and ayooluwa@spaceinafrica.com to share your DUNIA success stories.
Kindly tune in to our YouTube page to watch the full session and gain a deeper understanding of the various features and capabilities available to you for sustainable EO applications on the DUNIA platform.