scikit-eo: a Python package for remote sensing data analysis

In recent years, a growing body of space-borne and drone imagery has become available with increasing spatial and temporal resolutions. This remotely sensed data has enabled researchers to address and tackle a broader range of challenges effectively by using novel tools and data. However, analysts spend an important amount of time finding the adequate libraries to read and process remotely sensed data.

With an increasing amount of open access data, there is a growing need to account for effective open source tools to read, process and execute analysis that contributes to underpin patterns, changes and trends that are critical for environmental studies. Applications that integrate spatial-temporal data are used to study a variety of complex environmental processes, such as monitoring and assessment of land cover changes (Chaves et al., 2020), crop classifications (Pott et al., 2021), deforestation (Tarazona et al., 2018), impact on urbanization level (Trinder & Liu, 2020) and climate change impacts (Yang et al., 2013). Other complex environmental processes that are monitored by integrating spatial-temporal data are assessments of glacier retreat (Hugonnet et al., 2018), related hydrological change (Huss & Hock, 2018), biodiversity conservation (Cavender-Bares et al., 2022) and disaster management (Kucharczyk & Hugenholtz, 2021).

To bridge the gaps in remotely sensed data processing tools, we here introduce scikit-eo, a brand-new Python package for satellite remote sensing analysis. Unlike other tools, it is a centralized, scalable, and open-source toolkit due to its flexibility in being adapted into large dataset processing pipelines. It provides central access to the most commonly used Python functions in remote sensing analysis for environmental studies, including machine learning methods. scikit-eo stands out with its ability to be used in various settings, from a lecturer room to a crucial part of any Python environment in a research project. The majority of the tools included in scikit-eo are derived from peer-reviewed scientific publications, ensuring their reliability and accuracy.

By integrating this diverse set of tools, scikit-eo allows to focus on analyzing the results of data rather than being bogged down by complex lines of code. With its centralized structure, integrated use cases, and example data, scikit-eo empowers to optimize resources and dedicate more attention to the meaningful interpretation of findings in a more efficient way.