Because Earth is a complex coupled interacting system, where no one element is independent of any other, there are both pressing scientific and societal needs to improve our understanding how various physical, biological, and hydrological processes interact in surface Earth systems. This requires the development of new and increasingly more sophisticated ways of mathematically describing these systems and improvements in software and user interfaces that will dramatically enhance our ability to simulate these systems to improve the accuracy and reliability of model predictions of weather, floods, droughts, and climate variability. Such improved models will allow researchers to make better use of available data across disciplines and improve theory and algorithms that are essential to understanding Earth system behavior. Unfortunately, at present there is significant overhead of time and effort needed for discovering, accessing, understanding, and preparing data required to populate these models, as well as long learning lead times on how to use presently available models, owing to their complexity. This research overcomes some of these limitations by developing an innovative open modeling framework that can integrate data and models easily and is easy to use so that not only the research community and operational professionals can use it, but also policy makers and other interested parties. This EAGER award allows the construction of a prototype open meta-modeling framework that significantly reduces the time and effort on the part of users in the preparatory work for data and model comparisons, model testing and validations, for making fundamental knowledge discoveries in surface and ground water hydrological systems. In this framework, components/modules interact via user-configured open interfaces that allow the addition and integration of hydrological models and data sources using a common meta-level architecture and scientific workflows. The proposed prototype is based on a recently completed modeling framework, HS-NWSRFS (Hydro-information System for improving the National Weather Service River Forecast System). It represents a collaboration between investigators from three institutions, NASA, and NWS Ohio River Forecast Center (OHRFC). The funded effort will significantly expand the present code into an open community framework prototype. Broader impacts of the work include interagency collaboration, improved hydrological forecasting for rivers, and support of a PI whose gender is under-represented in the sciences.
To improve understanding of the complex behaviors of the various processes (e.g., physical, hydrological) and their interactions involved in the Earth System, as well as the accuracy and reliability of model predictions of weather, floods, droughts, and climate variability, researchers need to be able to make good use of the available data across disciplines to improve their theories, algorithms, models, and validations. However, there are significant hurdles to quickly discover, access, understand, and prepare the data. In addition, the complexity of models necessitates a long lead time to learn their usage. Therefore, the need is urgent for the development of an open modeling framework, which can integrate data and models easily for knowledge discovery, improvement of model predictability, and policy management, not only to the research community and operational professionals, but also to policy makers and other users. The goal of this NSF EAGER project is to build a prototype of open data open modeling framework, which should significantly reduce the time and effort on the part of users in the preparatory work for various data, data fusion, data visualization, model testing, validations, comparisons, modeling coupling, and fundamental knowledge discoveries. In such a framework, components/modules interact via user-configured open interfaces, so that various hydrological models and data sources can be easily added and composed to interoperate through scientific workflows. From the EAGER project support, a prototype Open Hydrospheric Modeling Framework (OHMF) is developed. This framework’s architecture consists of a system core (called as MSM core), open data, open model, and a workflow interface with the VisTrails workflow engine (see Figure 1). The Open Data architecture allows one to incorporate any new data agents into the prototype framework for the access of new external data sources easily and efficiently. The Open Modeling architecture allows one to incorporate any new model agents for the integration of new geosciences models into the prototype framework easily and efficiently. Heterogeneous models connected to the OHMF can then be coupled with each other if there is such a need in application, without the need of each model’s source code nor the recompilation of each model’s source code. The OHMF architecture is designed in such a way that model agents can be written by geoscientists themselves easily. Models connected to the OHMF system can also access the various heterogeneous external data from different data sources (e.g., NASA, NOAA, CUAHSI data centers, USGS) in an automatic manner (see Figure 2) through the OHMF Open Data architecture. In addition, the OHMF system offers services such as data fusion, data assimilation, rescaling, re-gridding, unit conversions, etc., among others. The prototype OHMF system has been tested with promising potentials and very encouraging feedback from an end-user workshop, various meeting presentations, and student-user testing group. Figures 3 -- 6 show some of the capabilities of the OHMF system, such as data access, model simulation, and model coupling. The testing and evaluation of the prototype OHMF system has demonstrated that our design approach can minimize the need for additional software development by end users; and it supports model integration and model coupling, as well as data integration. Thus, the OHMF system has a great potential to enable the geoscience communities of end users and researchers to easily contribute their data and to make use of other data sources; to easily test their models and to conduct model coupling; and to potentially carry out model intercomparisons studies and cross-disciplinary model applications through the use of "agents" which is not only easy, user-friendly, but also can significantly reduce the time and effort of geoscientists in the preparatory work of data access from various sources and preprocessing, data to model automation, and so on. Therefore, the OHMF system can significantly facilitate interdisciplinary investigations, fundamental knowledge discoveries, and testing of different hypothesis and theories/models. For example, the OHMF system has a great potential to improving our understanding of the complex behaviors of the various interacting processes (e.g., physical, hydrological, atmospheric, biological, chemical), as well as to improving the accuracy and reliability of model predictions for natural disasters including flooding and droughts. Even though, our prototype OHMF system is developed for the geosciences community, its design principles and architecture are not constrained by any specific discipline requirements, and thus, it is suitable to be reused by and extended to other disciplines where data access and modeling are fundamental for their scientific exploration.
