John M Douglass, Environment Agency.
A good SCIMAP-sediment model will identify probable critical source areas of fine sediment within a catchment at a field scale. It does this by estimating relative erodability of soils and estimating hydrological connectivity across a high resolution DEM. However, land management practice often varies at a scale finer than the land use inputs to the model, and can be independent of land use. Sub-field scale land management practice may have a large effect on field scale sediment yield. For example, a heavily poached gateway may contribute large yields, even if the field it is within is low slope grassland. Is connectivity to the river a sufficiently mitigating factor, even when SCIMAP estimates the risk of a land area incorrectly, to allow accurate representation of critical source areas?
This talk will compare a SCIMAP-sediment model of the River Kennet, a tributary of the River Thames, with a two stage catchment walkover conducted in 2012.
The walkover survey identified 16 riparian and near-riparian sediment ‘hotspots’ and a further 71 less serious sediment source areas. The most common source of sediment in the most serious category was roads and tracks with large, steep arable fields being the second most common.
Comparison with SCIMAP shows a poor spatial correlation between walkover survey identified ‘hotspots’ and SCIMAP outputs. SCIMAP landscape erosion risk does not consider sub-field scale land management practice, so a lack of correlation may be expected. Neither the in-channel risk nor the in-channel risk concentration show a good relationship to walkover survey ‘hotspots’ either.
That walkover survey identified ‘hotspots’ do not align with SCIMAP critical source areas should not diminish the usefulness of either approach, rather they should be thought of as complementary, but operating on different scales. Where walkover surveys are very good at detecting local, often transitory sediment sources, SCIMAP is more useful in identifying the inherent risk in the geographical-land use setting. It is possible that repeated catchment walkover surveys may be used to inform revised relative erodability classes that incorporate farming practices.