CAS-Hydro Model

Design Overview

The CAS-Hydro model has been developed using an object orientated design in C++. This design approach resulted in a model that is simple to maintain, debug, adapt and extend. There are three key types of object in the design:

• The water object
• The cell object
• The landscape object

The modular object-orientated design means that the model components can be re-used and adapted with ease. Therefore the model can be applied to new environments and processes with minimal effort. The modular approach also allows for individual researchers to concurrently work on different components of the model.

Water Object

The water object represents a parcel of water with its associated solutes that can be moved through the environment. This water object may be split, or merged with other water objects, ensuring that the solutes are moved with the water. Due to the single object that is routed through the model environment, it is simple to add additional solutes. It is possible to evaporate the water from a water object and solutes remain.

Cell Object

The cell object represents the environmental process occuring at a point in the landscape. A collection of these cell objects are held within the model with each instance of the object representing a different section of the landscape. The cell object simulates hydrological and nitrogen cycling processes.

Landscape Object

The landscape object represents the spatial aspects of the model. This object is divided into two domains: the terrestrial and the aquatic. The terrestrial domain represents the landscape and is constructed from a collection of cell objects arranged in a grid structure. The aquatic domain represents the river system and is constructed from a collection of reach objects in a tree structure. The landscape objects routes water and solutes between the cell objects using the FD8 algorithm and from the landscape into the river system.

Summary

CAS-Hydro is a fully distributed, object orientated model that is capable of accurately simulating catchment scale hydrological and nitrogen cycling processes. The modular object orientated basis of the model enables the adaptation and extensions of the model to other environments and processes with ease.

The model has undergone a detailed sensitivity analysis and a GLUE-based multi-parameter multi-objective assessment excercise. The model has been appled to two upland UK catchments and has been shown to be capable of accurately simulating their hydrological response. The assessment of the nitrogen cycling component has been limited due to the scarcity of measured data. However, the initial results show that the model is capable of simulating the in-channel nitrate response well.