Prediction Of Hydrological Models’ Uncertainty By A Committee Of Machine Learning-Models

This study presents an approach to combine uncertainties of the hydrological model outputs predicted from a number of machine learning models. The machine learning based uncertainty prediction approach is very useful for estimation of hydrological models' uncertainty in particular hydro-metrological situation in real-time application [1]. In this approach the hydrological model realizations from Monte Carlo simulations are used to build different machine learning uncertainty models to predict uncertainty (quantiles of pdf) of the a deterministic output from hydrological model . Uncertainty models are trained using antecedent precipitation and streamflows as inputs. The trained models are then employed to predict the model output uncertainty which is specific for the new input data. We used three machine learning models namely artificial neural networks, model tree, locally weighted regression to predict output uncertainties. These three models produce similar verification results, which can be improved by merging their outputs dynamically. We propose an approach to form a committee of the three models to combine their outputs. The approach is applied to estimate uncertainty of streamflows simulation from a conceptual hydrological model in the Brue catchment in UK and the Bagmati catchment in Nepal. The verification results show that merged output is better than an individual model output. [1] D. L. Shrestha, N. Kayastha, and D. P. Solomatine, and R. Price. Encapsulation of parameteric uncertainty statistics by various predictive machine learning models: MLUE method, Journal of Hydroinformatic, in press, 2013.

A Serious Gaming Approach For Serious Stakeholder Participation

Serious games are a category of games which are designed for a specific purpose other than for pure entertainment. It is not a new concept but serious games using real data, coupled with real time modelling and combining model results with social and economic factors opens up a new paradigm for active stakeholder participation. DHI and UNEP-DHI Centre initiated a project called Aqua Republica where a virtual world is developed which allows participants to develop a river basin and visualise the consequences of their decisions. The aim of this project is to raise awareness of the interconnectivity of water and educate on integrated water resources management. Aqua Republica combines a game layer with a water allocation model, MIKE BASIN, to create an interactive, realistic virtual environment where players play the role of a catchment manager of an undeveloped river catchment. Their main objective is to develop the river catchment to be as prosperous as it can be. To achieve that, they will need to generate a good economy in the catchment to provide the funds needed for development, have a steady food supply for their population and enough energy and water for the catchment. Through these actions by the player, a meaningful play is established to engage players and to educate them about the complex relationships between developmental actions in a river basin and the natural environment as well as their consequences. The game layer also consists of a reward system to encourage learning. People can play and replay the game, get rewarded from performing the right principles and penalised from failures in the game. This abstract will explain the concept of the game and how it has been used in a stakeholder participation environment.