A New Rainfall Runoff Software Library

Lumped-conceptual rainfall-runoff models are a ‘stock in trade’ tool for those working in the water industry and allied fields. However, there are many alternative models available, and sometimes multiple implementations of the same model. The consequence of this is a plethora of models with different operational features and different access arrangements. Similarly, there are a variety of optimisation tools being used to calibrate such models, and they too vary greatly in terms of their style and access arrangements. To address this issue we have developed a software application currently called the ‘rainfall-runoff library’ (RRL) and have made this available to end-users via the World Wide Web. The RRL includes some popular lumped conceptual rainfall-runoff models, including AWBM, SimHyd and Sacramento. It also includes popular model-independent and multi-objective parameter optimisation routines, including the Pattern Search, Shuffled Complex Evolution and Genetic Algorithm methods. In addition, a custom parameter optimisation method for AWBM is provided. The RRL has powerful visualisation and data handling features. It is expected that end-users will be more willing to trial a range of models and parameter estimation methods on their particular modelling problem as the ‘cost’ of moving from one method to another is relatively minor given the common interface elements. The tools embedded within the current version of the RRL are a sample only and we expect the software to grow significantly over the coming years. For this reason, we have designed the RRL as an extensible system that simplifies the task of adding models or optimisation methods, by using object introspection and metadata. This paper explores the functional analysis, technical design and implementation of the RRL.