I attended an Environment Agency workshop in Reading yesterday on “Land Use and Flooding Research Prioritisation” when Amy Parrott, Adam Baylis and Wendy Brooks presented their work on gathering evidence on the use of land use management strategies for Flood Risk Management with a view to identifying areas where more evidence is required. The summary is in the form of a “jigsaw” or spreadsheet of questions about the impact of different land use interventions that might be made with a view to reducing flood risk. This was of great interest to me because I was involved in the Defra/EA review of the impacts of land management on flood runoff (FD2114) and a study on whether such effects can be identified in the historical record at catchment scales (FD2120). FD2114 (led by Enda O’Connell) suggested that model projections of local land management effects to the catchment scale could not yet be considered reliable, while FD2120 suggested that it was difficult to identify any significant impact in past rainfall-runoff data series given the variability in the climate forcing (and the inconsistencies in catchment data sets), except perhaps for small storms with relatively dry antecedent conditions. There is, however, a lot of evidence for such impacts at experimental plot scales, and a lot of anecdotal evidence of increased frequency of surface runoff with agricultural intensification. Others have also suggested that the effects of land use can be seen in the historical record as increased rates of rise and excess numbers of peaks (e.g Archer et al. Hydrology Research, 2010). The EA are certainly still considering a range of land management options as a potentially useful method for reducing flood risk.
The question then is how to prioritise different investment options for maximum impact. Where options expected to reduce flood runoff such as buffer strips, grip blocking, or riparian forests can be justified under other environmental support schemes for agriculture, then this will generally be useful. It is possible to envisage situations, however, where there might be disbenefits. For example, retardation of runoff in downstream parts of a catchment, so that it is delayed to coincide with the upstream peak might actually make peak discharges worse. It is also possible to envisage that the creation of wetlands and more saturated soils by grip blocking, is creating source areas that might produce runoff more quickly than in the drained condition. It is clear that in going from local measures to catchment scale effects it might still be difficult to extrapolate the evidence from small scale studies to the larger scale. In particular, it will be difficult to assess the relative costs and benefits when such extrapolations will involve significant uncertainty.
I pointed to this yesterday in one post-it comment (it was a facilitated workshop so there were lots of post-its) on the use of the the word quantify in the questions addressed in the jigsaw spreadsheet. Quantification could be useful in deciding between investment options and priorities but will involve exactly the (uncertain) extrapolations to other places, times and scales and will be difficult. It seemed to me that the EA need to give this some more thought.
They also need to give some thought to the total lack of links between this FRM initiative and other drivers such as WFD. Those links we missing but as we move towards integrated catchment management it would seem to be necessary to explore the synergies in managing for runoff, water quality and ecosystem services, especially in justifying the costs of mitigation measures. The Defra Demonstration Test Catchment studies (mostly concerned with the impact of land management on water quality) were suggested as an opportunity to put some of these things together but there still seems to be some institutional resistance to this type of integration (in practice if not in principle)……..
k
