Speaker Bio & Abstract
University of Canterbury
BiographyProfessor in Spatial Information and Director of the Geospatial Research Institute Toi Hangarau, University of Canterbury, Christchurch, New Zealand. Primary research interests include hydrology, flood risk and water resources under climate change, incorporating remote sensing, computational modelling and Geographical Information Systems. Currently working on the implementation of federated spatial data infrastructures to facilitate real-time access to critical data and enabling the rapid production of information for hazard management.AbstractMatthew Wilson, Jiakai Li | University of Canterbury A web application is presented which enables the statistical analysis of rain gauge data in near-real time. The primary aim is to enable river or water managers to quickly assess the current state of rainfall, with reference to the historical context through the estimation of event likelihoods, allowing them to quickly ascertain whether a particular rainfall event is significant. The application is demonstrated for the Canterbury region of New Zealand. Data are automatically obtained from gauges linked with telemetry and added to a database, then statistics are calculated including the current rainfall observation probability and the derivation of Intensity-Duration-Frequency (IDF) curves, isohyet (rainfall likelihood) and standardised precipitation index (SPI) maps. Users are presented with a map of current rainfall observations, represented as points at gauge locations, scaled and coloured according to the magnitude and likelihood of the rainfall event. Thus, high-magnitude, low-likelihood events are highlighted to users as potentially flood-inducing. Users may also select to view IDF curves for each individual gauge, and associated isohyetal and SPI maps. The toolbox is developed using the R statistical language and a MongoDB database with the application developed with OpenCPU. Planned future developments include the inclusion of weather model and satellite rainfall data.