Monitoring and modelling the pre- and post-blocking hydrological response of moorland drains

Recent flooding in the UK has raised increasing concerns that changes in modern rural land use/management (LUM) practices may be having a detrimental effect on flooding downstream. An understanding must therefore be developed of how distributed upstream effects on runoff can affect downstream flooding, as a basis for designing efficient and safe rural-based flood prevention and mitigation measures. The central problem is understanding how effects on runoff at the small scale, where the impacts of LUM practices are relatively well understood, can aggregate and propagate downstream to have an impact on flood peak flows and volumes. A typical problem involves understanding how flow fields behave in the uplands, such as flow fields in drained moorland subject to LUM practices that include blocking open-channel drains (i.e. grip blocking). This paper describes some results from field work and numerical modelling for grip blocking at Sapling Clough (1.7 km2), a headwater of the River Hodder catchment, NW England, UK. Data from field surveys and a dense nested hydrometric monitoring network are being used to test a novel design of physically-based model that has been developed specifically to study how flow connectivity and water storage are affected by LUM changes (MURSAFF; Model for Upland Runoff Storage And Flow Fields). The dominant feature of the flow field at Sapling Clough is geometrical complexity, in the distribution of water flows and water storages in the channels and on and within the peat body. This complexity is impossible to model adequately using traditional distributed models that require the landscape to be represented using GIS squares or finite-difference grid cells or finite element triangles. Given that the geometrical complexity might play a role in the aggregation and propagation of impacts, MURSAFF uses a very flexible irregular polygon grid that can be adapted to closely match features such as grips and hillslopes.