This session focuses on advances in catchment hydrology process understanding across scales from national/regional, to catchment or hillslope, that are made possible by the analyses of large or diverse datasets, or by an integration of top-down and bottom-up modelling approaches.

In recent years, hydrologists have benefited from advances in data collection, storage and communication techniques, to share and access large volumes of data. This progress offers new opportunities to compare, contrast and combine datasets; to put local, experimental data in a larger-scale context; to quantify the information content of data for hydrological analysis; to discover behaviour and patterns across regional, national and international scales; and to understand trends and drivers of hydrological processes. Recent examples include analysis of model parameters/structures across hydro-climatic gradients as a diagnostic tool; and virtual observatories which access multiple datasets within cloud computing resources.

A synthesis of top-down and bottom-up approaches shows how catchment modelling can benefit from process knowledge at multiple scales. For bottom up approaches, one typically starts with process models describing local scale flow and transport in particular compartments such as in soils, aquifers and water bodies. To arrive at a catchment scale model, first those local process representations need to be coupled across different compartments. In a second step, regionalisation techniques are applied to obtain model inputs and parameters at catchment scales. The bottom-up approach will typically lead to a complex catchment model representing flow and transport processes in a spatially explicit manner. In contrast, for top-down approaches, catchment scale observations such as river discharges or solute concentrations will be the starting point. From these observations, one initially separates out processes explaining the variability of measurements with a minimum number of assumptions; and these identified processes need to be localised subsequently.

We encourage contributions in several areas:
• Work to set up or populate new hydrological databases
• Comparative analysis of hydrological processes across time and space
• New hydrological theory, laws or empirical relationships derived from large datasets
• Attempts to extract large hydrological datasets from unconventional sources e.g. photo libraries, crowd sourcing, etc.
• Assessment of magnitude, nature, and/or impact of uncertainty in large hydrological datasets
• Catchment studies where top-down and bottom-up paradigms were brought together
• Case studies involving both top-down and bottom-up modelling approaches.