WebGIS Applications For The New Generation Of Hydrological Models

Claudia Vitolo (Imperial College and Lutra Consulting) with Peter Wells (Lutra Consulting)

This presentation has been cancelled
Description: Developed to address the limitations of existing hydrological desktop applications, OASIS-WS utilises the latest advances in web technologies and data availability to inform key decision makers of the impacts of land management on water resources and flood risk, through a robust and consistent web-based approach.
Abstract:

Knowledge about the potential impact of land management scenarios on water resources and flood risk is essential for decision-making amongst various governmental organisations and the private sector. Environmental models are typically applied to assess impacts of future water shortages & flood risk and to simulate and analyse various scenarios to inform key decision makers. The practice of environmental modelling is hindered by a number of technological characteristics of the current generation of models. These models are typically stand-alone applications developed for desktop computers with little standardisation of input & outputs and with limited possibility of automatic coupling with other models. This is problematic for various reasons. Not only does it limit computational power, but it also hinders the construction of complex workflows, ensemble modelling using different model structures, and a full uncertainty analysis of the model chain. Web technologies may overcome most of the above limitations. These technologies, combined with appropriate standards, provide tools to construct analysis workflows of relevant spatial and non-spatial data available over the internet/intranets. In addition, elastic and scalable cloud services facilitate computational efforts. OASIS-WS (Web Service) is a WebGIS based tool for processing hydrological information. OASIS-WS uses recent web technologies to capture both spatial and non-spatial hydrological data, analyse and process those data, and to present results as maps and plot outputs. The data discovery element of OASIS-WS is designed to explore data availability using Open Geospatial Consortium (OGC) standards. It is based on the recently developed data exchange standard, WaterML2, which enables the storing and exploring of various hydrological data sources through a web interface. Data collected through the discovery stage are stored in a PostgreSQL / PostGIS database. An algorithm was developed to filter and select appropriate data for analyses of specific scenarios. Based on a user scenario (e.g. flooding or low flow analysis), OASIS-WS can select an appropriate hydrological model from the multi-model Framework for Understanding Structural Errors (FUSE), developed by Clark et al. (2008). FUSE consists of an ensemble of over 1200 conceptual rainfall-runoff models. It is widely used for local to national scale hydrological analyses. FUSE has been implemented in R with some portions implemented in C++ to optimise processing times. It is an integrated part of the RHydro package (publicly available from R-Forge). The functionality of FUSE is exposed to OASIS-WS as a web service using the OGC Web Processing Service (WPS) standard. Users interact with OASIS-WS through its web interface which features a map panel and modelling dashboard. Visualisation and interaction are facilitated using the OGC Web Map Service (WMS) and Web Feature Service (WFS). The exploration of input data and model results is implemented using the Sensor and Observation Service (SOS). OASIS-WS is fast, reliable and able to provide the most appropriate simulation engine for a given user scenario thanks to the modular nature of the underlying hydrological models. As OASIS-WS is web-based, it is also scalable and easily accessible.