The hydroclimatology science community has been utilizing large datasets and quantitative methods for many decades, such as data that represents land-atmosphere parameters at multiple spatial and temporal resolutions and scales. These have been used to reveal complex processes and interactions between these parameters and have been used in a variety of applications such as near real-time flood monitoring. Two common approaches for investigating land-atmosphere interactions and hydroclimatological processes are physical models and data-driven models. Due to the fact that hydroclimatological processes are extremely complex, more recently data-driven models have attracted much research attention and are considered complementary or as a potential alternative to physically-based models.
This attention has been driven by: the availability of large-volumes of data from satellites, climate models and various non-traditional sources such as social media; the development and use of advanced data science algorithms; and high-performance computing facilities. Together these are helping to improve hydroclimatological understanding and modeling, and development of applications. There are already review articles, research papers, and projects focused on aspects of big data science, machine learning (including deep learning) and artificial intelligence in flood modeling, precipitation-runoff modeling, urban runoff and flooding, precipitation forecasting, model validation and comparison, etc. There is also large interest in the commercial sector, such as Google's recent release of its deep-learning informed regional flood prediction model. Therefore, the time is right to create a Research Topic on big data science in hydroclimatology research and applications. This topic emphasizes the use of big data methodologies in hydroclimatology research, including:
1) High performance and cloud computing platforms
2) Data mining
3) Machine learning (particularly deep learning)
4) Model parameterization and optimization.
Submission to this Research Topic may include literature reviews, original research papers, and applied research case studies. This topic represents an emerging research direction with diverse opportunities for understanding and approaches. Therefore, our intention is to create a forum for new research and users to share views and results, to generate consensus of related issues, and help to promote a research agenda within the next 5 years.
The hydroclimatology science community has been utilizing large datasets and quantitative methods for many decades, such as data that represents land-atmosphere parameters at multiple spatial and temporal resolutions and scales. These have been used to reveal complex processes and interactions between these parameters and have been used in a variety of applications such as near real-time flood monitoring. Two common approaches for investigating land-atmosphere interactions and hydroclimatological processes are physical models and data-driven models. Due to the fact that hydroclimatological processes are extremely complex, more recently data-driven models have attracted much research attention and are considered complementary or as a potential alternative to physically-based models.
This attention has been driven by: the availability of large-volumes of data from satellites, climate models and various non-traditional sources such as social media; the development and use of advanced data science algorithms; and high-performance computing facilities. Together these are helping to improve hydroclimatological understanding and modeling, and development of applications. There are already review articles, research papers, and projects focused on aspects of big data science, machine learning (including deep learning) and artificial intelligence in flood modeling, precipitation-runoff modeling, urban runoff and flooding, precipitation forecasting, model validation and comparison, etc. There is also large interest in the commercial sector, such as Google's recent release of its deep-learning informed regional flood prediction model. Therefore, the time is right to create a Research Topic on big data science in hydroclimatology research and applications. This topic emphasizes the use of big data methodologies in hydroclimatology research, including:
1) High performance and cloud computing platforms
2) Data mining
3) Machine learning (particularly deep learning)
4) Model parameterization and optimization.
Submission to this Research Topic may include literature reviews, original research papers, and applied research case studies. This topic represents an emerging research direction with diverse opportunities for understanding and approaches. Therefore, our intention is to create a forum for new research and users to share views and results, to generate consensus of related issues, and help to promote a research agenda within the next 5 years.
