A call for a fundamental shift from model-centric to data-centric approaches in hydroinformatics

Abstract

Over the years, data-driven models have gained notable traction in water and environmental engineering. The adoption of these cutting-edge frameworks is still in progress in the grand scheme of things, yet for the most part, such attempts have been centered around the models themselves, and their internal computational architecture, that is, the model-centric approach. These endeavors can certainly pave the way for more tailor-fitted models capable of producing accurate results. However, such a perspective often neglects a fundamental assumption of these models, which is the importance of reliability, correctness, and accessibility of the data used in constructing them. This challenge arises from the prevalent model-centric paradigm of thinking in the field. An alternative approach, however, would prioritize placing data at the focal point, focusing on systematically enhancing current datasets and devising frameworks to improve data collection schemes. This suggests a paradigm shift toward more data-centric thinking in water and environmental engineering. Practically, this shift is not without challenges and necessitates smarter data collection rather than an excessive one. Equally important is the ethical and accurate collection of data, making it available to everyone while safeguarding the rights of individuals and other legal entities involved in the process.

Keywords

• hydroinformatics
• computational intelligence
• artificial intelligence
• data-centric approach