A model of sediment retention by vegetation for Great Britain: new methodologies & validation

Abstract

AbstractSoil erosion is an substantial environmental concern worldwide. It has been historically and is of increasingly concern currently. Next to natural processes, over 2 million hectares of soil are at risk of erosion through intensifying agriculture in the Great Britain (England, Wales, Scotland and their territorial islands). Predictive soil erosion models, in the form of Ecosystem Service tools, aid in helping to identify areas that are vulnerable to soil erosion. Yet, no predictions for erosion or sediment retention by vegetation based on local data have been developed for Great Britain or the United Kingdom as a whole.Here we develop an erosion retention model using the InVEST platform, which is based on the RUSLE mathematical framework. We parameterise the model, as far as feasible, with GB specific input data. The developed model estimations are validated against suspended solids concentrations (sediments) in throughout England and Wales.Next to presenting the first GB wide estimate of erosion and erosion retention using the InVEST SDR module, we test three approaches here that differ from more widely applicable RUSLE model inputs, such as created for Europe as a whole. Here, we incorporate (1) periodicity to allow erosion to potentially fluctuate within years; (2) GB-specific cover periodic management factors estimates, including a range of crop types, based on observed satellite NDVI values (3) soil erosivity under heavy rainfall following GB estimates for 2000-2019.We conclude that both the GB created erosivity layer as the added periodicity do not seem to be provide substantial improvement over non-periodic estimated created with more widely available data, when validated against this set of suspended solids in rivers. In contrast, the observed cover management factors calculated from NDVI are a good improvement affecting the ranking order among catchments. Therefore, the generating of cover management factors using NDVI data could be promoted as method for InVEST SDR model development and in more general for developing RUSLE-based erosion estimates worldwide.